JP2009259215A - Road surface marking map generation method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To obtain a position coordinate of a road surface mark by generating an image including the road surface mark from an image obtained by photographing a road surface while traveling on the road. <P>SOLUTION: The road surface is photographed with a video camera while traveling on the road, and the position coordinates of each photographing point are acquired by a GPS, or the like. A computer converts each frame image of this moving image and generates an orthogonal image seen from the right overhead. It generates a linearly connected image by arranging the orthogonal image on a linear axis representing a travel distance from a reference point at a photographing time on the basis of the position coordinates of the photographing point. It recognizes the position and size of a road mark, such as a pedestrian crossing on the basis of the connected image. After that, it obtains the absolute coordinates of each mark by mapping a traveling track during photographing and the mark position to an absolute coordinate system. A road surface mark can be recognized with high accuracy in this way. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to a technique for generating a road marking map including a marking applied to a road surface.

  Electronic map data used in car navigation or the like requires various detailed data in order to realize various functions. One example is the markings on road surfaces such as pedestrian crossings, central lines, and lane boundaries. By acquiring these signs as images in advance, it is possible to provide the user with an image close to the actual road surface, and to realize guidance that is easy to understand intuitively.

As a technique for efficiently generating an image of a road surface including a sign, Patent Document 1, Patent Document 2, and the like can be cited.
Patent Document 1 discloses a technique for generating a still image including a road surface sign from an image acquired by a digital camera or the like with respect to the front or rear or side of a vehicle. In this technique, a road marking or the like is photographed with a digital camera or the like mounted on the vehicle while traveling on the target road with the vehicle. Then, each frame image constituting the moving image is converted into an orthogonal image as viewed from directly above, and arranged according to the shooting position. An orthographic image is a road image when the viewpoint is placed at an infinite point vertically above the road. By arranging a plurality of frame images, it is possible to obtain a composite image of the road surface along the trajectory of one run (hereinafter also referred to as a path).

  Patent Document 2 discloses a technique for synthesizing a wide road image by synthesizing images obtained by two passes. In this technique, first, an affine transformation is applied to an image obtained in one pass so that what is originally drawn linearly, such as a road lane boundary line, is displayed as a straight line. Then, the image of one path is affine transformed so that the coordinates of the lane boundary line or the like that are photographed in common in the two paths coincide. Also disclosed is a technique of synthesizing three or more paths by synthesizing images while performing affine transformation for each path by a similar method.

  Patent document 3 is disclosing the technique which extracts a sign from the image of a road surface. In this technique, a portion corresponding to road paint is extracted from an image such as an aerial photograph by image processing, and the type of road paint is determined based on the feature amount of the extracted portion. As the feature amount, an angle between an axis representing the longitudinal direction of the extracted portion and a link of the road network is used.

JP 2007-249103 A Japanese Patent No. 3820428 JP 2007-122665 A

Map data includes a road network in which roads are represented by nodes and links for route search, and data in which roads are represented by polygons in order to display a map. In the road network, since the road is represented by one or two links, the coordinates attached to the link do not know exactly which part of the road is represented. In the drawing data, the position coordinates of the outer periphery of the polygon representing the road are known, but the position coordinates of the points inside the road are not known.
For example, if there is map data in which the position coordinates of each point inside the road are obtained in detail, the lane change is determined by determining which lane on the road the vehicle is driving according to the current position of the vehicle It is possible to realize highly functional guidance such as providing guidance on the vehicle and warning that a pedestrian crossing is approaching the vehicle.
However, the conventional map data has only insufficient accuracy to realize such high-precision and high-performance guidance. Even if the current position of the vehicle is grasped with high accuracy, it cannot be said that detailed map data has been prepared to make use of the position information.

The road marking is suitable as an object for enriching the position coordinates on the road. For example, if the position coordinates of the pedestrian crossing and the lane boundary line are obtained, it is possible to contribute to the realization of the high-performance guidance described above.
However, all the prior arts have a main purpose of obtaining a composite image of a road surface, and have not been intended to obtain a position coordinate of a road surface sign.
For example, the technique of Patent Document 2 merely represents an image with the traveling direction of the vehicle as the X axis and the movement distance as the X coordinate regardless of whether the road is a straight line or a curve, and a direction orthogonal to the X axis. Only the image is affine transformed. Since there is no guarantee that the X coordinates determined in this manner are sufficiently coincident with each other for images obtained by a plurality of passes, the technique of Patent Document 2 cannot accurately obtain the position coordinates of road markings. .
The affine transformation can also be said to be a transformation having the effect of distorting the rectangular area of the original image into a parallelogram. Therefore, in the technique of Patent Document 2, image degradation is caused by combining images by affine transformation, and the road surface. There is also a problem of further reducing the position coordinates of the markings.
On the other hand, the technique described in Patent Document 1 only discloses processing for an image obtained in one pass and cannot sufficiently cover the entire road.

The prior art has a problem that the position accuracy of the sign cannot be sufficiently secured even when the sign is extracted from the road surface image. Although Patent Document 3 recognizes the type and position of the sign by the positional relationship with the link given by the road network, as described above, the coordinates attached to the link represent any part of the road. It is data that we do not know exactly. Therefore, even if the position of the sign is recognized with reference to the link, sufficient accuracy cannot be ensured.
Further, in Patent Document 3, the angle between the main axis and the link of the area extracted from the road image is used as a feature amount. However, the direction of the main axis depends on the extraction accuracy of the area, and is stable. It is not always required with high accuracy. If the range of the reference value when judging the type of the label is increased in consideration of the error included in the main axis, there is a possibility of erroneous recognition of the label.

  It is an object of the present invention to solve these problems, extract road surface markings from a composite image of a road surface with high positional accuracy, and generate a map including these markings.

The present invention can be configured as a road marking map generation method in which a road marking map including a marking applied to a road surface is generated by a computer.
In the present invention, first, the computer inputs image data of a continuous image obtained by photographing a road surface including a sign while moving on a road to be photographed, and position coordinate data representing a photographing position of the image data.
The above-mentioned image data can be photographed by a photographing device mounted on a moving body such as a vehicle, for example. For example, a digital video camera or the like can be used as the photographing apparatus. Moreover, it is preferable that the photographing apparatus is equipped with a position measuring device that acquires position coordinate data at the time of photographing. As the position measuring device, for example, a GPS (Global Positioning System) or an inertial navigation device such as a gyro can be used alone or in combination. For convenience of processing, it is preferable to prepare a recording apparatus that inputs a photographed image and position coordinate data and records them in synchronization.

The computer converts each frame image constituting the input image data to obtain an orthographic image in a state where the road surface is viewed from directly above. The orthographic image may be generated using a part of each frame image.
Next, a linear distance axis that represents the distance from the reference point on the path, which is the movement trajectory when the photographed road is moved in a predetermined direction on the two-dimensional plane, as an axis used by the computer for processing Define The reference point may be the start point of the path or an arbitrary point provided on the path. The distance axis can be defined in any direction in the two-dimensional plane. However, it is preferable to match with the x-axis or the y-axis from the viewpoint that each processing described later is simplified.
Based on the distance from the reference point along the path, the orthographic image is arranged on the distance axis, thereby generating a linear connected image representing the road surface of each path. At this time, a part of the orthographic image may overlap. For example, the orthographic image is preferably arranged with its center line along the distance axis.
In this way, a connected image representing a straight road along the distance axis can be obtained.

In the present invention, the computer can recognize the type and position of the sign applied to the road surface based on the linearly connected image. A path is a travel locus when photographing a road surface. Since the position coordinate data is acquired while traveling, the position accuracy is sufficiently secured, and the position accuracy of the extracted sign can be secured.
Further, by arranging the orthographic image on the distance axis, it is possible to obtain a connected image extending linearly even when the road is curved. When the connected image is generated in this way, the sign given on the road exists in a certain positional relationship with respect to the distance axis. For example, pedestrian crossings and stop lines are drawn so as to be orthogonal to the distance axis, and lane boundary lines are drawn parallel to the distance axis. In this way, how to draw a sign to be extracted in the connected image is constant for each sign, so it is possible to clearly set the conditions for extracting the sign and extract the sign stably. Is possible.

In the present invention, if an orthographic image is arranged and then image data combined as a single image or image data obtained from the beginning as a single image such as an aerial photograph is used. It is very difficult to obtain a linear connected image along the distance axis. It is necessary to recognize the shape of the road from these images by image processing and then convert the image so that the road becomes a straight line. Since the shape of the road is various, the calculation required for this conversion also differs depending on the shape of the road and is difficult to generalize.
On the other hand, in the present invention, orthographic images are arranged to generate a connected image. Therefore, by arranging the orthographic image on the distance axis, there is an advantage that it is possible to relatively easily obtain a linearly connected image.

When the road width is wide, it may not be possible to photograph the entire road with one pass. In such a case, an image of the entire road may be obtained by combining images taken with a plurality of paths having different travel positions.
In this case, the image may be taken by moving a plurality of times on any lane. In the former case, the lane and the path have a one-to-one correspondence. In the latter case, the path and the lane correspond to each other in a plurality of one-to-one. The present invention is applicable to any case.
When an image is taken with a plurality of passes, the positional accuracy is not always the same for each pass. Therefore, even if the connected images of each path are arranged based on the respective position coordinate data, inconsistencies may occur in the positions of the markings. The present invention can also be used in a method for eliminating such inconsistencies and synthesizing a plurality of connected images to obtain an image of the entire road.
In order to perform the above-described composition, in the present invention, the computer first extracts a sign for each of a plurality of paths. Then, in a region photographed in common with a connected image of two or more passes among a plurality of passes, corresponding points corresponding to each other are specified based on the indication obtained with each connected image. . For example, when a pedestrian crossing is photographed in common in two paths, any corner of the striped pattern of the pedestrian crossing can be set as a corresponding point in each connected image. The corresponding point is identified by using a connected image in an absolute coordinate system in which the orthographic image is arranged at each position on the path based on the position coordinate data, not the connected image in which the orthographic image is arranged on the distance axis. Is preferred.
The computer compares the extracted sign types and positional relationships among multiple paths and identifies the corresponding correspondence of the sign so that any point on the sign is automatically identified as the corresponding point. May be. Further, the connected image and the extracted sign may be displayed on a computer display, and the operator may view the display and indicate the corresponding point.
The deviation between the corresponding points specified in this way represents an error in position coordinates.

The computer sets one of a plurality of paths as a reference path. This setting may be performed after the corresponding point is identified or may be performed before that. Then, based on the movement vector set so that the positions of the corresponding corresponding points coincide with each other, the combined image of the road surface extending over a plurality of paths can be obtained by correcting the connection image of the path other than the reference path. Generate.
This correction is performed by translating for each region constituting the connected image based on the movement vector. This area may be obtained by dividing the generated connected image once into a size corresponding to the original orthographic image or another arbitrarily set size. In addition, when generating a connected image, if the orthographic image is merely arranged without being synthesized, the position may be corrected for each orthographic image.

In this way, since a connected image representing the road surface of each path can be generated based on the position coordinate data at the time of shooting, the connected image can be obtained in a state where positional accuracy is ensured.
Then, one of the multiple paths is set as the reference path, and this path is fixed, and the position of the other paths is corrected. It is possible to eliminate the positional accuracy error.
Further, since the synthesis of each path is performed by translating the connected image for each area, the position can be corrected without adding distortion to the orthographic image of each area. Therefore, at the time of this correction, the sign on the road surface can maintain the relative positional accuracy in the orthographic image of each region.
By the above operation, according to the road marking map generation method of the present invention, a composite image including road markings can be obtained while ensuring the position accuracy.

In the road marking map generation method of the present invention, the reference path can be set by various methods such as designation by an operator.
The computer may also input evaluation data on the accuracy of the position coordinate data for each path and set a reference path based on the evaluation data. For example, among the plurality of paths, there is a method of setting a path that is evaluated as having the highest position accuracy based on evaluation data as a reference path. If a composite image is generated so as to match another path with the reference path set in this way, the highest position accuracy can be ensured.
The evaluation data may be data that directly represents the position accuracy quantitatively or may be data that can be used to calculate the position accuracy.

In the present invention, the sign is specified based on the linear connected image obtained by arranging the orthographic image on the distance axis. Since this coordinate system is different from the absolute coordinate system corresponding to the position coordinate data, the position of the sign is only a relative position in the connected image. Therefore, in order to utilize for the above-mentioned alignment and other processes, it is preferable to convert the position of the sign into a position in the absolute coordinate system.
An example of this conversion method is as follows, for example. First, a perpendicular foot from the representative point of the marking to the distance axis is used as the reference point of the marking. Then, the position of the reference point is obtained on the path shown in the absolute coordinate system based on the position coordinate data, and a perpendicular line is drawn from the reference point to the path. In this way, the position coordinates in the absolute coordinate system of the marking can be obtained by arranging the representative point of the marking at the end point of the perpendicular line. In addition, relative coordinates and absolute coordinates corresponding to the position coordinate data are given to the four corners of each sign, and conversion such as bilinear conversion may be performed, or the relative coordinates of each sign and the corresponding frame The absolute coordinates may be recalculated again from the time trajectory data.

The sign may be extracted and stored as image data and used for displaying a road image.
Alternatively, a method may be adopted in which a geometric shape is defined for each type of sign and the position and size are determined so as to include the sign. For example, the geometric shape of the pedestrian crossing is determined in advance so as to be expressed as a rectangle or a parallelogram. Then, based on the result of extracting the pedestrian crossing from the connected image, the position and size of the representative point of this rectangle or parallelogram are determined. The center of gravity can be used as the representative point, and the diagonal length can be used as the size. You may use the position of the vertex in a diagonal as a representative point and a size. The same applies to other signs.
When the sign extraction result is stored in the position and size of the geometric shape, there is an advantage that the data amount can be reduced as compared with the case where the result is extracted as image data. When reproducing a road image with a sign, a typical pattern prepared for each kind of sign may be displayed in an enlarged or reduced manner according to the position and size of the geometric shape.

  In the present invention, when recognizing a sign in a connected image, both end regions separated by a predetermined distance or more from the distance axis may be deleted in advance. When a photographed image is converted to obtain an orthographic image, distortion at the time of conversion tends to be relatively large at both ends of the image. If the two end regions of the connected image obtained by arranging the orthographic image are deleted in advance, the area with such a large distortion is removed in this way, so the sign is recognized using a good image with relatively high position accuracy. It becomes possible to do.

In the present invention, the recognition of the sign can be performed by various methods.
As a first method, there is a method using pattern matching. In this method, first, a predetermined color portion used for marking is extracted from the connected image. Then, the sign is recognized by matching the sign pattern prepared in advance for each kind of sign and the extracted portion. White and yellow can be used as the predetermined colors to be extracted. In order to extract such regions with high accuracy, it is possible to adopt a method in which connected images are grayscaled and regions having a lightness of a predetermined level or more are extracted.
This first method is suitable for extracting a sign having a relatively complicated shape. Examples of indications to which the first method can be applied include arrows, U-turns, traffic indication time indications, numbers such as turn prohibition, speed restrictions, pedestrian crossing notices, bus lane characters, end symbols, and the like.

As a second method for recognizing the sign, there is a method of identifying the sign based on the geometric condition of the portion corresponding to the sign. In this method, as in the first method, first, a portion of a predetermined color used for marking is extracted from the connected image. Then, based on the direction and size of the extracted portion with respect to the distance axis, the type of the sign is recognized. The conditions for recognizing the sign are exemplified below.
Pedestrian crossing ... A region where multiple parallel lines are recognized in a direction parallel to the distance axis;
Bicycle crossing zone: A line segment perpendicular to the distance axis that represents two lines drawn at regular intervals (may be taken into account near the pedestrian crossing);
Stop line: one that represents one line among the line segments orthogonal to the distance axis;
Lane boundary line: A line segment parallel to the distance axis and longer than a predetermined length;
Deceleration zone: A predetermined length line segment orthogonal to the distance axis is arranged parallel to the distance axis direction;
Zebra: Areas with parallel lines drawn at predetermined intervals that do not fall under pedestrian crossings and deceleration zones;

As described above, the present invention can extract a sign from a connected image. However, during actual shooting, various objects other than road markings may appear. For example, when taking a picture while traveling on a road with a vehicle, another vehicle that stops or travels next to the lane in which the vehicle travels may appear in the screen.
A connected image generated using an image having such a reflection includes various unrecognizable images (hereinafter referred to as “garbage images”) other than road markings. If an attempt is made to recognize a sign in this state, there is a risk that the dust image portion may be misrecognized as some sign or may be unrecognizable and result in an error.
In order to suppress such adverse effects, in the present invention, a dust image may be recognized, and a road marking may be recognized reflecting the result.

A first method for recognizing a dust image and reflecting it in recognition of a road marking is as follows.
First, the computer obtains an optical flow between successive frame images. The optical flow is composed of a plurality of vectors representing movement from a plurality of feature points in the previous frame image to a corresponding feature point in the next frame image among two consecutive (adjacent) frame images. Is. The two consecutive frame images do not necessarily need to be consecutively captured frame images, and are defined as a group of frame images obtained by thinning the captured frame images at a predetermined distance interval or time interval. May be.
Based on the optical flow thus obtained, the CPU of the computer extracts an image other than the sign, that is, a dust image from the frame image. This is because when a foreign object different from the road marking appears in the frame image, the optical flow also has a different size and direction from the optical flow of the road marking portion. Reflection extraction may be evaluated based on both the magnitude and direction of the optical flow, or may be evaluated based on either one.

  Next, the CPU generates an orthographic image reflecting the extraction result of the reflection. In addition to the orthographic image generated without considering the dust image, an orthographic image of the dust image may be generated. Further, instead of the orthographic image generated without considering the dust image, the orthographic image may be generated based on an image obtained by removing the dust image from the frame image.

The CPU recognizes the sign reflecting the dust image based on the orthographic image thus generated. When a dust image is reflected as pre-processing for recognizing a sign, a connected image reflecting the reflection extraction result is generated, and the sign is recognized based on the connected image. Such a connected image can be generated by, for example, a method of removing a dust image from a connected image obtained by using an orthographic image generated without considering the dust image. Further, a connected image may be generated using an orthographic image obtained based on an image obtained by removing a dust image from a frame image.
When a dust image is reflected as post-processing for recognizing a sign, first, the CPU recognizes a road sign using a connected image obtained by using an orthographic image generated without considering the dust image. . If it is determined that the dust image is mistakenly recognized as a road marking by comparing the recognition result with the connected image obtained from the orthoimage of the dust image, the recognition result is removed.

  By using the optical flow in this way, it is possible to recognize the portion of the dust image due to the reflection of foreign matter other than the road marking. Moreover, the recognition accuracy of the road marking can be improved by reflecting the recognition result on the recognition of the road marking as pre-processing or post-processing.

A second method for recognizing the dust image and reflecting it in the recognition of the road marking is as follows.
The CPU cuts out a portion where the orthogonal images arranged on the distance axis overlap. Then, the gradation difference of each point between the orthogonal images of the overlapping portion is obtained. The gradation difference is a difference between color components recorded at each point. In the case of a color image, a gradation difference may be obtained for each RGB component, but it is preferable to obtain a lightness (V) difference after decomposing into three elements of HSV. Since the road markings should overlap in the overlapping part between the orthogonal images, the gradation difference with respect to the road markings is a predetermined value or less. On the other hand, foreign objects other than road markings usually have large gradation differences because the shape and size of the reflection on the orthographic image are different when the shooting time and shooting location are different. Therefore, it is possible to extract a dust image by extracting a point where a gradation difference exceeding a predetermined threshold is obtained.
When the dust image is extracted in this manner, the road marking can be recognized by reflecting the dust image extraction result in the pre-processing or post-processing, as in the case of the optical flow.

  Even in the second method described above, it is possible to recognize the portion of the dust image due to the appearance of a foreign object other than the road marking, and to improve the recognition accuracy of the road marking. In addition, the second method has an advantage that the processing load is lighter than that using the optical flow.

  In the case of using the second method, it is preferable to obtain the gradation difference after applying a smoothing filter to the orthogonal image. This is because if the gradation difference is strictly determined for each pixel constituting the image, a large gradation difference may occur in the road marking portion due to a position error or the like when generating an orthographic image. By applying the smoothing filter, the gradation change in the image can be dulled, so that the influence of the position error and the like can be alleviated and the dust image extraction accuracy can be improved.

The dust image extraction may be applied to the entire frame image, but may be applied to only a part of the frame image in order to improve the processing speed.
For example, it is possible to target a portion that is an orthographic image generation target.
As another aspect, a predetermined area from which a predetermined range including a path is removed may be a target for extracting a dust image. Since shooting is performed while moving along the path, it is unlikely that foreign objects other than road markings are reflected in the area around the path. This is because in the travel lane in which the host vehicle travels, it is possible to travel at a greater distance so that a vehicle traveling ahead does not enter the image creation range. On the other hand, in lanes other than the traveling lane, there is currently no means for avoiding the appearance of foreign objects.
Furthermore, a predetermined area from which a predetermined range including a path is excluded may be used as a dust image extraction target in a portion that is an orthographic image generation target when both are combined.

The present invention does not necessarily have all the above-described features, and some of them may be omitted as appropriate, or some features may be appropriately combined.
The present invention is not limited to the road marking map generation method described above, and may be configured as a generation device that generates a road marking map including the markings applied to the road surface by the road marking map generation method. The road marking map generation method described above may be configured as a computer program for causing a computer to realize the method, or may be configured as a computer-readable recording medium on which the computer program is recorded. Examples of the recording medium include a flexible disk, a CD-ROM, a magneto-optical disk, an IC card, a ROM cartridge, a punch card, a printed matter on which a code such as a bar code is printed, an internal storage device of a computer (a memory such as a RAM or a ROM). ) And external storage devices can be used.

  By arranging the orthographic image obtained by orthorectifying the frame image acquired while traveling on the road, it is possible to obtain a connected image with high positional accuracy along the traveling locus (path).

It is explanatory drawing which shows the structure of the road surface imaging | photography system as an Example. It is explanatory drawing which shows the structure of the road marking map production | generation apparatus as an Example. It is explanatory drawing which shows the intermediate data in the production | generation process of a road marking map. It is explanatory drawing which shows the production example of the road image in an Example. It is explanatory drawing which shows the outline | summary of the alignment process. It is explanatory drawing which shows the procedure of the alignment in case an intersection exists. It is a flowchart of a connection image generation process. It is a flowchart of an alignment process. It is a flowchart of a standard path setting process. It is a flowchart of a connection image movement process. It is explanatory drawing which shows the process example (1) of position alignment process. It is explanatory drawing which shows the process example (2) of an alignment process. It is explanatory drawing which shows the process result of the process (2) of position alignment process. It is explanatory drawing which shows the acquisition method of the absolute position coordinate of a road marking. It is explanatory drawing which shows the outline | summary of a transparent polygon setting process. It is a flowchart of a transparent polygon setting process. It is explanatory drawing which shows the example of a road image before setting a transparent polygon. It is explanatory drawing which shows the example of a road image after the setting of the transparent polygon. It is a flowchart of a paint recognition process. It is explanatory drawing which shows the content of the vertical arrangement | positioning process. It is a flowchart of a crossing related paint extraction process. It is explanatory drawing which shows the side cut process example (1). It is explanatory drawing which shows the side cut process example (2). It is explanatory drawing which shows the extraction process etc. of a pedestrian crossing candidate area | region. It is explanatory drawing which shows the extraction process etc. of a bicycle crossing zone / stop line candidate area | region. It is explanatory drawing which shows the content of a boundary line (solid line) extraction process. It is explanatory drawing which shows the content of a boundary line (broken line) extraction process. It is explanatory drawing which shows the content of the arrow extraction process. It is explanatory drawing which shows the content of a zebra extraction process. It is explanatory drawing which shows the content of the U-turn extraction process. It is explanatory drawing which shows the content of the rotation prohibition extraction process. It is explanatory drawing which shows the content of the regulation extraction process. It is explanatory drawing which shows the content of the number extraction process. It is explanatory drawing which shows the content of the pedestrian crossing notice extraction process. It is explanatory drawing which shows the content of the deceleration zone extraction process. It is explanatory drawing which shows the content of the road surface painting extraction process. It is explanatory drawing which shows the content of a bus lane character extraction process. It is explanatory drawing which shows the content of the end symbol extraction process. It is explanatory drawing which shows the content of a relative coordinate conversion process. It is explanatory drawing which illustrates the reflection of a foreign material. It is a flowchart which shows the reflection method of the recognition result of a dust image. It is a flowchart of a dust image recognition process (1). It is an example of an optical flow. It is an example of the optical flow when there is a traveling vehicle. It is an example of the optical flow when there is a stopped vehicle. It is a flowchart of a dust image recognition process (2). It is explanatory drawing which shows the example of recognition of a dust image.

Embodiments of the present invention will be described in the following order.
A. System configuration:
A1. Road surface photography system:
A2. Road marking map generator:
B. Outline of processing:
B1. Intermediate data structure:
B2. Processing example:
B3. Outline of alignment processing:
C. Road marking map generation method:
C1. Connected image generation processing:
C2. Alignment processing:
C3. Standard path setting process:
C4. Connected image movement processing:
C5. Transparent polygon setting process:
C6. Paint recognition process:
D. effect:
E. Variations:

A. System configuration:
In this embodiment, a method for generating a map including road markings (hereinafter referred to as a “road marking map”) using road surface images taken by a video camera mounted on a vehicle will be described.
The system of the present embodiment includes a road surface photographing system and a road marking map generating device. The road surface photographing system is a system for photographing an image of a road surface with a video camera while traveling on the road. In the present embodiment, the target road is traveled a plurality of times with different travel trajectories, and images are respectively captured.
The road marking map generation device is a device that generates a road marking map based on an image of a road surface photographed by a road surface photographing system. First, a connected image of a part of the lane on the road surface is generated by arranging the photographed image after converting it into an orthographic image on each of the above-described traveling trajectories. And the image of the whole road is produced | generated by arrange | positioning the image of a some driving | running | working locus so that a position coordinate may match. In addition, a road surface sign is extracted from the connection image thus generated.
Hereinafter, the system configuration of the road surface photographing system and the road marking map generating device will be described.

A1. Road surface photography system:
FIG. 1 is an explanatory diagram showing a configuration of a road surface photographing system as an embodiment.
The road surface photographing system 100 is a system mounted on a vehicle. The system configuration will be described based on the block diagram below.
The video camera 120 captures an image of a running road surface.
The position measurement unit 110 is a device that measures position coordinates during shooting. The position measuring unit 110 includes a GPS (Global Positioning System) 114, an IMU (Inertial Measurement Unit) 116, a DMI (Distance Measuring Instrument) 118, and a controller 112. The GPS 114 is a global positioning system. The IMU 116 is an inertial measurement device including a triaxial gyro and an acceleration sensor inside. The DMI 118 is a device that measures the distance traveled by detecting the rotation of a wheel.

The controller 112 receives signals from the GPS 114, the IMU 116, and the DMI 118, and sequentially outputs position coordinates at the time of shooting. The position coordinates can take an arbitrary coordinate system, but in this embodiment, latitude and longitude and altitude are used.
In addition, after obtaining these signals, a self-estimated position accuracy σ, which is an evaluation value of the position coordinate measurement accuracy, is also output. In general, it is known that the detection accuracy of the GPS 114 varies depending on the arrangement of artificial satellites used for detecting position coordinates, the reception status of radio waves, the presence of multipath by receiving radio waves reflected on buildings, and the like. Yes. In differential positioning, the detection accuracy is also affected by the operating status of the reference station.
The self-estimated position accuracy σ can be arbitrarily defined. For example, the self-estimated position accuracy σ may be calculated using a precision reduction rate (DOP (Dilution of Precision)) determined by the arrangement of the satellites of the GPS 114.
The self-estimated position accuracy σ may be analyzed when the acquired data is processed by a road marking map generation device to be described later.

The recording device 130 records the output signals of the video camera 120 and the position measuring unit 110 in synchronization. In this embodiment, the recording device 130 is configured by a device in which a recording hard disk 140 is added to a general-purpose personal computer. In the hard disk 140, as shown in the figure, image data 142, synchronization data 144, and measurement data 146 are recorded. The image data 142 is a moving image file of an image taken with a video camera. The measurement data 146 is position coordinates obtained by the position measurement unit 110. The synchronization data 144 is data that associates the acquisition times of the image data 142 and the measurement data 146 with each other. By referring to the synchronization data 144 and the measurement data 146, the position coordinates of the shooting point can be obtained for each frame of the image data 142.
The data structure for recording at the time of shooting is not limited to the structure described above. For example, the measurement data 146 may be data that sequentially stores the position coordinates of each frame of the image data 142. In this way, the synchronization data 144 can be omitted. In order to acquire such data, for example, the recording device 130 can output a synchronization signal to the position measuring unit 110 for each frame of the video camera 120 and acquire the position coordinates at that time.

The state mounted on the vehicle is schematically shown in the upper part of the figure.
The video camera 120 is installed in front of the vehicle so that a front image can be taken. A wide-angle lens may be attached to widen the angle of view.
The antenna 114A of the GPS 114 is installed on the upper roof of the vehicle. In the present embodiment, two main and sub antennas 114A are installed in front of and behind the vehicle so that radio waves from GPS artificial satellites can be reliably received and sufficient positional accuracy can be ensured. Only one of them may be used.
The IMU 116, DMI 118, and controller 112 were installed at the rear of the vehicle. The DMI 118 is mounted so that the rotation of the rear wheel can be detected.
Since the recording device 130 and the hard disk 140 can be installed at any location in the vehicle interior, illustration is omitted.

A2. Road marking map generator:
FIG. 2 is an explanatory diagram showing a configuration of a road marking map generating apparatus as an embodiment. It is an apparatus for generating a road marking map based on an image of a road surface photographed by a road surface photographing system. In this embodiment, a method of advancing the process interactively or semi-automatically while receiving an instruction by a command from an operator as appropriate, instead of generating a road marking map completely automatically.
In the figure, functional blocks of the road marking map generating apparatus 200 are shown. In the present embodiment, the road marking map generating apparatus 200 is constructed in software by installing a computer program for realizing each function shown in a general-purpose personal computer. Some of these functional blocks may be provided by an OS (Operating System). Each of these functional blocks can also be configured in hardware. In addition, here, for convenience of explanation, it is described as a stand-alone operating device, but each functional block may be distributed and prepared in a plurality of computers connected via a network.

  The main control unit 201 performs integrated control of each functional block. The data input unit 204 inputs image data 142, synchronization data 144, and measurement data 146 from the hard disk 140 that records various data acquired by the road surface photographing system 100. In this embodiment, the hard disk 140 is reconnected to the road marking map generating apparatus 200 from the road surface photographing system 100 to transfer these data. However, a method of transmitting data via a network, a DVD, A method of transferring data using a recording medium such as the above may be adopted. Since the measurement data includes the position coordinates obtained by the position measurement unit 110, the data input unit 204 corresponds to the input unit in the present invention.

The command input unit 202 inputs a command from an operator through operations such as a keyboard and a mouse provided in the computer.
The display control unit 203 displays the processing result of the road marking map generation device 200 on a computer display, or displays a screen for an operator to instruct various commands. The functions of the command input unit 202 and the display control unit 203 may be provided by a computer OS (Operating System).

The trajectory data calculation unit 205 generates data representing a travel trajectory (hereinafter also referred to as “pass”) when the image data 142 is captured based on the measurement data 146. In this embodiment, the trajectory data calculation unit 205 corrects the measurement data 146 in which the position coordinates obtained by the road surface photographing system 100 are recorded based on detection information provided from a reference station whose position coordinates are known. Thus, trajectory data is generated. Since the technique for correcting the position coordinates using the information of the reference station is well known, the description thereof is omitted. This processing can improve the accuracy of position coordinates.
However, it is not essential to use data from the reference station. The position coordinates obtained from the measurement data 146 may be used as they are. In such a case, the trajectory data calculation unit 205 can be omitted.

The image conversion unit 206 generates an orthographic image by converting each frame image of the image data 142 into an orthographic projection, that is, a state as viewed from directly above.
A one-pass composition unit 207 serving as a connected image generation unit determines an orthographic image of each frame image obtained by the image conversion unit 206, and a position at which a representative point in the orthographic image is determined based on position coordinates at the time of shooting. By arranging so as to come to the coordinates, an image of the road surface along the traveling locus (path) at the time of photographing is synthesized. The image synthesized in this way is called a connected image. The combined connection image is stored in the processing data storage unit 210.
In this embodiment, each road is photographed by traveling a plurality of times with different traveling trajectories. The 1-pass image composition unit 207 generates a composite image for each pass. As a result, a plurality of connected images are generated according to the number of paths.

The alignment processing unit 220 arranges the plurality of connected images generated by the one-pass image combining unit 207 so that the road image is aligned by correcting the positional coordinate error between the connected images, that is, the connected images. By performing the process, an orthogonal image of the entire road (hereinafter also referred to as “road image”) is generated. The alignment process is performed according to an instruction from the operator. The processing contents will be described later.
The road image obtained by the alignment is stored in the processing data storage unit 210.

  The transparent polygon setting unit 221 sets a transparent polygon on the obtained road image according to an operator's instruction. When performing the above-described alignment, a part of the orthographic image corresponding to the adjacent path may overlap. In the overlapped portion, the road marking may be clearly copied toward the orthographic image arranged on the lower side. In such a case, the transparent polygon is a polygon for designating an area to be subjected to processing for making a part of the upper orthographic image transparent so that the lower image is displayed. By setting the transparent polygon, it is possible to provide a map that can accurately grasp the road marking.

  A paint recognizing unit 223 as a sign recognizing unit recognizes a road surface sign (hereinafter also referred to as “paint”). In the present embodiment, the sign is recognized based on the connected image generated by the one-pass image combining unit 207. The paint recognition result is stored in the processing data storage unit 210.

  The road marking map generating device can output a road marking map based on the road image generated as described above. For example, a road image may be output as a printable file. Moreover, you may output a road image as electronic data so that a road marking map may be produced | generated as an electronic map. In addition, prior to these outputs, a process for obtaining position coordinates and shape data of road markings based on road images may be performed.

B. Outline of processing:
B1. Intermediate data structure:
FIG. 3 is an explanatory diagram showing intermediate data in the process of generating a road marking map. These data are sequentially stored in the processing data storage unit 210 (see FIG. 2).
In this embodiment, data acquired by the video camera 120 and the position measuring unit 110 while traveling on a road is stored in the hard disk 140 by a personal computer as the recording device 130. The stored data includes image data 142, measurement data 146, and synchronization data 144 for synchronizing them.

  The measurement data 146 is a record of position coordinate data at the time of shooting. In this embodiment, the trajectory data 210a is calculated by correcting the measurement data 146 with reference to the reference station data 150. This is a process performed by the trajectory data calculation unit 205 described above with reference to FIG. The reference station data 150 is data representing a detection result by GPS at a reference point whose position coordinates are known. For example, reference point data provided by the Geospatial Information Authority of Japan can be used. The trajectory data 210a obtained here represents the trajectory (hereinafter also referred to as “path”) when an image of the road surface is captured in each processing, in absolute coordinates including latitude, longitude, and altitude. Used as data.

On the other hand, a road surface texture 210c is generated from the image data 142, the synchronization data 144, and the trajectory data 210a. In addition, road surface trajectory data is generated from the synchronization data 144 and the trajectory data 210a.
In this embodiment, each road is traveled a plurality of times, and an image of the road surface is taken. Therefore, the road surface track data 210b and the route track data 210b are generated for a plurality of paths for each road.

A coupled image 210d is generated using the road surface texture 210c and the route trajectory data 210b. The connected image 210d is an image generated by the one-pass image composition unit 207 in FIG. That is, the connected image 210d is a road surface image of each path generated by arranging each road surface texture 210c based on the position coordinates represented by the route locus data 210b. The connected image 210d is also generated for a plurality of paths for each road.
The connected image 210d can also be generated as one image file obtained by combining the road surface texture 210c. In this embodiment, for convenience of subsequent processing, information (hereinafter also referred to as “registered data”) for generating the connected image 210d by arranging the road surface texture 210c is not generated as a composite image. The image is stored in association with each image of the road surface texture 210c. Such information can include position coordinates for placing the road surface texture 210c, posture (angle) at the time of placement, information for specifying the adjacent road surface texture 210c, vertical relationship with the adjacent road surface texture 210c, and the like.

Using the connected image 210d thus obtained, processing such as alignment and transparent polygon setting is performed. These processes are processes performed by the alignment processing unit 220 and the transparent polygon setting unit 221 in FIG. By this process, the linked images 210d for a plurality of paths can be synthesized to obtain a road image 210e for each road.
The road image 210e may be generated as a composite image, or information for generating the road image 210e by arranging the road surface texture 210c may be stored in association with each image of the road surface texture 210c. Good. In this embodiment, the latter method is adopted. Information such as the position coordinates at which each road surface texture 210c is arranged and the posture (angle) at the time of arrangement are stored as road image registration data 210f. In addition, since a process for correcting the position error is performed on the route trajectory data 210b in the alignment process, information indicating the correction process for the original data is stored as the trajectory registration data 210g.

  In addition, the data of the connected image 210d (including the road surface texture 210c and the route trajectory data 210b) are also stored. It is preferable to store the image data 142 and the trajectory data 210a as the original data. If the connected image 210d is stored in the form of a composite image, the road image 210e is combined with the connected image 210d, so that the image quality is deteriorated compared to the original data due to repeated combining. There is a risk. On the other hand, as in the present embodiment, by leaving data close to the original data including the road surface texture 210c, the road image 210e can be generated using these data. Therefore, it is possible to suppress the image processing being superimposed on the image data, such as repeated synthesis, and to improve the image quality of the road image 210e.

B2. Processing example:
Next, in order to facilitate understanding of the outline of processing in the present embodiment, a processing example will be shown.
FIG. 4 is an explanatory diagram illustrating a generation example of a road image in the embodiment. FIG. 4A shows an example of generation of a connected image obtained along one path, and FIG. 4B shows an example of a road image obtained by arranging connected images of a plurality of paths. Is shown.
Straight lines L <b> 41 to L <b> 44 in FIG. 4A represent travel trajectories (paths) when a road image is captured while traveling on the road surface photographing system 100. The PIC 41 in FIG. 4A is a linked image generated based on image data obtained by traveling on the path L43. In this embodiment, since the image is taken using the wide-angle lens, it is possible to obtain a connected image that only covers a plurality of lanes even in one pass. The fact that both ends of the connected image are serrated in a saw-tooth shape is due to the influence of shape distortion that occurs when each frame of image data is orthographically projected. The connected image PIC41 is generated by arranging orthogonal images (road surface texture) having the number of frames corresponding to the number of jagged peaks.
Such a connected image is obtained for each of the paths L41 to L44 in the drawing.

  FIG. 4B shows a road image PIC42 obtained by synthesizing connected images for the paths L41 to L44. It can be seen that road images are generated that are wider than those in FIG. 4A and include the opposite lane. When combining a multi-pass connected image, if there is an error in the position coordinates of each pass, a shift occurs between the connected images. If these shifts exist, the signs such as the pedestrian crossing and the lane boundary in FIG. 4B are also displayed in a shifted state. In this embodiment, the composition is performed while correcting the positional coordinate error between the connected images of each path. This process is called alignment. By performing alignment in this way and synthesizing the connected images, as shown in FIG. 4B, a road image in which signs such as a pedestrian crossing and a lane boundary line are matched can be obtained.

B3. Outline of alignment processing:
FIG. 5 is an explanatory diagram showing an outline of the alignment process. In the present embodiment, alignment is performed by moving the connected images in parallel so that the positions of corresponding points designated by the operator are aligned based on the signs photographed in common with the plurality of connected images.
FIG. 5A shows a processing method when only one corresponding point is designated. In the figure, two linked images PIC51 and PIC52 are drawn. Each of these includes a diamond sign, that is, a warning sign for a pedestrian crossing. However, in the state on the left side of FIG. 5 (a), the linked images PIC51 and PIC52 have a relative position error, so that the positions of the markings are shifted.
The operator designates corresponding points using a pointing device such as a mouse while looking at the display screen. In the example of the figure, a state is shown in which P51 and P52 corresponding to the apex of the notice of pedestrian crossing are designated. These corresponding points P51 and P52 are originally points that should overlap at the same position if the connected images PIC51 and PIC52 have no position error. Therefore, in this embodiment, the connected images PIC51 and PIC52 are translated as indicated by arrows in the drawing so that the corresponding points P51 and P52 coincide.
At this time, a method was adopted in which one of the connected images PIC51 and PIC52 is translated and the other is translated. In the example of the figure, an example in which the connected image PIC52 is moved with reference to the connected image PIC51 is shown. By moving in this way, it is possible to obtain the road image PIC53 in a state in which the shift of the warning sign is eliminated.

FIG. 5B shows a processing method when a plurality of corresponding points are designated. In the figure, two linked images PIC54 and PIC55 are drawn. Each of these includes a warning sign for a pedestrian crossing. However, in the state on the left side of FIG. 5B, since the linked images PIC54 and PIC55 have a relative position error, the positions of the markings are shifted.
In this state, it is assumed that the operator has specified two sets of corresponding points. A pair of corresponding points P54 and P53 and a pair of corresponding points P56 and P55. In the linked image PIC54, the entire warning sign M52 included in the linked image PIC55 is drawn, and a part of the warning sign M51 included in the linked image PIC54 has disappeared. Even in such a state, since it is clear that the corresponding points P55 and P56 correspond, it is possible to designate them as corresponding points.
When a plurality of sets of corresponding points are designated in this way, the linked image PIC 55 is moved so that the corresponding points match with the linked image PIC 54 as a reference. However, the first movement amount for matching the corresponding point P53 with P54 and the second movement amount for matching the corresponding point P55 with P56 are not necessarily the same. Therefore, in the region between the corresponding points P53 and P55, the first movement amount and the second movement amount are linearly interpolated to set the movement amount of each point. By doing so, it is possible to obtain the road image PIC 56 in a state in which the shift of the notice sign is eliminated.

FIG. 5B also shows an example of setting a transparent polygon.
In this example, half of the notice sign M51 in the linked image PIC54 is missing. If alignment is performed in this state, in this example, the connected image PIC54 is displayed so as to be superimposed on the upper side of the PIC55, so the notice sign M52 of the connected image PIC55 is covered by the connected image PIC54. As a result, the sign M52 drawn in the complete state in the connected image PIC55 cannot be utilized in the road image PIC56.
Therefore, in such a case, the transparent polygon TP50 is set so as to surround the notice sign M52 according to an instruction from the operator. At the place where the transparent polygon TP50 is set, the upper connected image is made transparent and displayed as if it was cut out. As a result, in the part of the transparent polygon TP50, the notice sign M52 drawn on the connection image PIC55 arranged below the connection image PIC54 is displayed.
In this embodiment, by making the transparent polygons settable in this way, the signs drawn in the respective connected images can be used effectively in the road image.

  FIG. 6 is an explanatory diagram showing a procedure for alignment when an intersection exists. In order to avoid complication of the figure, only the positional relationship of the paths of the connected images is shown here. In the figure, roads around two intersections are drawn. It is assumed that connected images are obtained along the paths BP61 and BP62 on the vertical road, respectively. For the horizontal road, it is assumed that a connected image is obtained along the paths BP63b, BP64b, and NP61b indicated by broken lines.

In this embodiment, when positioning between a plurality of paths, any one path is set as a reference path, and the other paths are translated to match the reference path. A path other than the reference path is hereinafter referred to as a standard path. Although the reference path and the standard path can be set by an arbitrary method, in this embodiment, as described later, a path with high position accuracy is set as the reference path.
In the example of FIG. 6, there is only a single path for each vertical road, so the paths BP61 and BP62 are the reference paths.
For the side road, in the section D61, the side having higher position accuracy among the paths BP63b and NP61b is set as a reference path, and for the section D62, the side in BP64b and NP61b having higher position accuracy is set as the reference path. Here, it is assumed that the paths BP63b and BP64b are set as reference paths, respectively. Further, the superiority or inferiority is determined by comparing the positional accuracy between the paths BP63b and BP64b. This is because the paths BP63b and BP64b are reference paths of the sections D61 and D62, respectively, but are continuous paths arranged on one road, and therefore it is necessary to perform alignment between these paths. In the example of FIG. 6, it is assumed that the path accuracy of the path BP63b is higher than that of the path BP64b.
As a result, for the horizontal path, the alignment priority is determined in the order of the reference path BP63b> reference path BP64b> standard path NP61b.

Next, each path is aligned according to the above-described priority. Assume that the vertical paths BP61 and BP62 have already been aligned.
First, alignment of the reference path BP63b is performed. It is assumed that the corresponding point P63b on the reference path BP63b is designated by the operator's instruction, and the point P63a is designated as its original position. As a result, the reference path BP63b is moved so that the corresponding point P63b coincides with the point P63a, and the reference path BP63a indicated by the solid line is obtained.
Although not shown, the connected image corresponding to the reference path BP63b also moves in accordance with the reference path BP63a. In the present embodiment, the connected images are displayed by arranging road surface textures along the reference path BP63b, and these road surface textures are not synthesized. Therefore, when the movement to the reference path BP63a is performed, a connected image of the reference path BP63a can be obtained by translating the position of each road surface texture along the reference path BP63a.

Next, the reference path BP64b is aligned. It is assumed that corresponding points P65b and P64b on the reference path BP64b are designated by the operator's instruction, and points P65a and P64a are designated as their original positions. This corresponding point is designated based on the connected image of the reference path BP63a. That is, the alignment of the reference path BP64b is affected according to the result of the process of aligning the reference path BP63b with the reference path BP63a.
When the corresponding point is designated, the reference path BP64b is moved so that the corresponding points P65b and P64b coincide with the points P65a and P64a, and the reference path BP64a indicated by the solid line is obtained. In accordance with this, the road surface texture constituting the connected image of the reference path BP64b is also translated on the reference path BP64a.

Finally, the standard path NP61b is aligned. It is assumed that corresponding points P68b, P67b, and P66b on the standard path NP61b are specified by an operator's instruction, and points P68a, P67a, and P66a are specified as their original positions. This corresponding point is designated based on the connected image of the reference paths BP63a and BP64a. That is, the alignment of the standard path NP61b is affected by the result of the process of aligning the reference path BP63b with the reference path BP63a and the process of aligning the reference path BP64b with the reference path BP64a.
When the corresponding point is designated, the standard path NP61b is moved so that the corresponding points P68b and P67b coincide with the points P68a and P67a, and the corresponding points P67b and P66b coincide with the points P67a and P66a. Moved. Since these three points are not on a straight line, as a result, the standard path NP61b is moved to the broken line-shaped standard path N61a. In accordance with this, the road surface texture constituting the connected image of the standard path NP61b is also translated on the standard path NP61a.

In the present embodiment, when there are a plurality of paths as shown in FIG. 6, alignment is performed preferentially from a path with high position accuracy by the procedure described above. By doing so, it is possible to perform alignment while ensuring sufficient overall position accuracy.
For example, in the processing of FIG. 6, it is assumed that alignment is performed in the order of low position accuracy, that is, in order of the standard path NP61b, the reference path BP64b, and the reference path BP63b. In this case, the alignment of the reference path BP64b is affected by the alignment of the standard path NP61b, and the position accuracy decreases. The alignment of the reference path BP63b is affected by the alignment of the standard path NP61b and the reference path BP64b, and the position accuracy is lowered. Therefore, if the alignment is performed in the order from the lowest position accuracy, the overall position accuracy decreases due to the interaction between the paths.
In the present embodiment, on the contrary, the alignment is performed in the descending order of positional accuracy. Therefore, it is possible to perform the entire alignment without deteriorating the position accuracy of the path having the highest position accuracy.

C. Road marking map generation method:
Hereinafter, the method for generating the road marking map described with reference to FIGS. 1 to 6 will be described in detail by taking as an example a case where an operator gives an instruction as necessary.
First, a connected image generation process, that is, a process of obtaining a connected image 210d of each path based on the road surface texture 210c and the road surface trajectory data 210b in FIG. 2 will be described.
Next, registration processing, that is, processing for positioning the connected image 210d with respect to a plurality of passes, reference path setting processing, and connected image movement processing performed during the positioning processing will be described.
Also, the transparent polygon setting process will be described.
In the series of processes described above, if the road marking map generating apparatus displays the position of the marking in the connected image, the operator can easily specify the corresponding point and the position of the transparent polygon. It is also possible to automatically identify corresponding points and set transparent polygons. Finally, a paint recognition process will be described as a process for enabling such a process.

C1. Connected image generation processing:
FIG. 7 is a flowchart of linked image generation processing. In terms of hardware, this is a process executed by the CPU of the road marking map generating apparatus 200. This corresponds to the processing of the image conversion unit 206 and the one-pass synthesis unit 207 shown in FIG.
When the process is started, the CPU first reads frame data (step S10). The frame data is an image of each frame constituting the image data 142 photographed by the video camera 120 of the road surface photographing system 100 (FIG. 1).

An example of frame data is shown in the figure. Since the video camera 120 is installed toward the front of the road surface photographing system 100, the frame data includes a road ahead of the vehicle, a vehicle ahead, and the like. In this embodiment, since it is desired to generate an image of the road surface, a partial region of this frame data is cut out and used. A region A71 in the figure represents a cut-out region set to include only the road surface. In this embodiment, the area A71 is set so as to acquire an image of an area 5 to 7 m ahead of the vehicle. The relative position of the area A71 in each frame is constant.
The area A71 is not limited to the above example, and can be set arbitrarily. Since the video camera 120 captures images at a constant frame rate, the frame data is an image group obtained by intermittently capturing the road surface. Therefore, it is preferable to determine the range of the area A71 so that the road can be reproduced as a continuous image when the group of images photographed intermittently are arranged. For example, if the vertical width of the area A71 is narrowed, a gap is likely to be generated between an area cut out from certain frame data and an area cut out from the next frame data when the vehicle speed is high. On the other hand, if the vertical width of the area A71 is increased, miscellaneous images different from the road image such as the preceding vehicle, the sky, and the building are easily included. The area A71 may be set in consideration of these influences.

Next, the CPU converts the acquired frame data into an orthogonal image (road surface texture) (step S12). The outline of the processing is shown in the figure. On the upper side is an example of frame data. Here, an example is shown in which only the road surface state is photographed, and the left and right lane regulation lines L71 and L72 and the sign M7 are photographed. Since this is an image of the front, the lane restriction lines L71 and L72 that are essentially parallel are shown in a square shape under the influence of perspective (perspective).
As described above, a partial area A71 of this frame data is cut out and used.
The lower part illustrates a state in which the image of the area A71 is orthographically converted. In order to convert the road into an image viewed from directly above, the left and right lane regulation lines L71 and L72 are converted into parallel line segments as shown in the figure. The sign M7 is similarly converted into a shape as seen from directly above.

An orthographic projection conversion method will be described.
First, it is assumed that a vehicle on which the road surface photographing system 100 is mounted is traveling on a horizontal plane, and a road as a subject is also on the same horizontal plane.
At this time, the road image, that is, the two-dimensional coordinates on the screen of the frame data is m = [u, v] ^T. Further, the three-dimensional coordinate of the world coordinate system fixed to the ground is M = [X, Y, Z] ^T. A vector obtained by adding one element to each of these coordinates as a direct product is defined as the following equation (1).

  The relationship between the three-dimensional coordinate M and the two-dimensional coordinate m of the projected image is modeled by the following relational expressions (2) and (3).

Where s is the scale factor;
[Rt] is an external parameter matrix;
R is a rotation matrix;
t is a translation matrix;
A is an internal parameter matrix.

The internal parameter matrix A is an internal parameter considering the focal length of the video camera 120 and the like, and represents a mapping parameter from the real image coordinate system (xy coordinate system) to the frame coordinate system (uv coordinate system).
α and β are scale factors in the u-axis and v-axis directions, respectively, and γ is a parameter represented by the skew of the two image axes;
[U ₀ , v ₀ ] ^T is the coordinate (principal point coordinate) of the principal point of the image.
Assuming that the pixel size of the image is (k _u , k _v ), the angle between the u axis and the v axis is θ, and the focal length is f, α, β, and γ are expressed by the following equation (4).

The external parameter matrix [Rt] is an external parameter depending on the installation position, installation posture, and the like of the video camera 120, and represents a mapping parameter from the world coordinate system (XYZ coordinate system) to the real image coordinate system (xy coordinate system). . In the world coordinate system, the road surface directly below the video camera 120 is the origin, the horizontal axis perpendicular to the traveling direction of the vehicle is the X axis, the vertical axis is the Y axis, and the horizontal axis in the traveling direction is the Z axis.
The parallel movement vector t is a movement vector of the image principal point of the real image with respect to the origin in the world coordinate system.
When the height of the video camera 120 (the height of the main image point of the actual image) is h, the translation vector t is expressed by the following equation (5).

In the world coordinate system, if the rotation angle (yaw angle) in the heading direction of the real image is φ, the pitch angle is ω, and the roll angle is κ, the rotation matrix R is expressed by the following equation (6).

The internal parameter matrix A is obtained by a prior measurement.
The yaw angle φ, pitch angle ω, roll angle κ, and height h of the image principal point are obtained by the following procedure. First, in an initial state, that is, a state where the vehicle is installed on a horizontal ground, reference values of the yaw angle φ ₀ , the pitch angle ω ₀ , the roll angle κ ₀ , and the height h ₀ are measured. Next, while driving, the change in the vehicle attitude angle and the change in vehicle height are recorded with a gyroscope, an acceleration sensor, etc., and this change is reflected in the above-mentioned reference value. The angle φ, the pitch angle ω, the roll angle κ, and the height can be obtained.

The orthographic projection conversion is performed by using the equation (2) based on these parameters, and a road image in the frame coordinate system (uv coordinate system) is converted into a projected road image in the world coordinate system (XYZ coordinate system). Can be converted. The procedure is as follows.
First, it is assumed that the road surface as a subject is an image of a horizontal plane (Y = 0). At this time, the relationship of the following equation (7) is established from the equation (2).

  As a result, the world coordinates (X, Z) and the scale parameter s for the pixel (u, v) can be obtained by the following equation (8).

Next, the CPU of the road marking map generating apparatus 200 performs correction in consideration of the inclination of the road surface that is the subject.
First, from the position coordinate data (X ₀ , Y ₀ , Z ₀ ) of each point where the frame data is acquired, and the position coordinates (X _i , Y _i , Z _i ) of a plurality of points near the road surface that is the subject, The slope of the road surface that is the subject is calculated. In this embodiment, it is assumed that the gradient is uniform.
Specifically, the height change Δh is obtained from position coordinate data in the vicinity of the world coordinate point (X, Y, Z) of the shooting point. That is Δh = _{Y-Y 0.} At this time, assuming a uniform gradient, the depth Z ′ of the points in the world coordinate system (X ′, Y ′, Z ′) on the road surface can be obtained by the following equation (9).

  When the corrected depth Z ′ on the road surface is determined, the relationship between the frame coordinate point (u, v) and the world coordinate point (X ′, Y ′, Z ′) is expressed by the following equation (10) from the equation (2). It becomes as follows.

Thus, X ′ and Y ′ of the world coordinate point can be calculated by the following equation (11).

As described above, an orthographic image (road surface texture) can be obtained by mapping the point (u, v) on the frame data to (X ′, Z ′), respectively. As shown in FIG. 7, when the frame data is cut out in a rectangular area A71 and orthographically projected, a trapezoidal orthographic image (road surface texture) A72 spreading upward is obtained.
In the present embodiment, for the convenience of the subsequent processing, the orthographic image (road surface texture) is generated in two ways of low resolution / high resolution. A high-resolution orthographic image (road texture) (hereinafter referred to as a “high-resolution image”) is generated with the same resolution as the original image, ie, an image generated by using the cut-out area A71 of the original frame data as it is. It is an image that was made. A low-resolution orthographic image (road surface texture) (hereinafter referred to as a “low-resolution image”) is an image whose resolution is lower than that of the original data. The resolution of the low-resolution image is preferably set to such a value that the road marking map generating apparatus 200 can display the image with a light load, and can be arbitrarily set such as half the resolution of the original image.

Next, the CPU of the road marking map generating apparatus 200 arranges the obtained orthographic image (road surface texture) and synthesizes a one-pass image (step S14). An example of 1-pass image synthesis is shown in the figure. In this example, orthographic images (road surface texture) A72 [0] to A72 [5] are synthesized.
In each orthographic image (road surface texture) A72, a point corresponding to the origin of the frame coordinate system (uv coordinate system) may be arranged based on the position coordinates at the time of shooting each frame data. Since the frame data is a forward image of the position of the vehicle, the orthographic image (road surface texture) needs to be arranged at a point moved forward by a certain distance from the vehicle position. In this embodiment, the positional relationship between the vehicle position and the frame coordinate system is calculated and arranged for each frame data. In addition, the orthographic image (road surface texture) is sequentially arranged from the old image to the new image in time series.
By arranging the orthographic image (road surface texture) in this way, the lane boundary lines L71 and L72 and the marking M7 on the road surface are reproduced.
In the present embodiment, at the stage of the connected image generation processing, the orthographic image (road surface texture) is kept arranged and displayed without being combined with one image. Therefore, what is generated by the one-pass image synthesis process (step S14) is not a synthesized image but information that determines the arrangement of each orthogonal image (road surface texture). However, in this process, a method of combining an orthographic image (road surface texture) into one image can be adopted.

C2. Alignment processing:
FIG. 8 is a flowchart of alignment processing. In terms of hardware, this is a process executed by the CPU of the road marking map generating apparatus 200. This corresponds to the processing of the alignment processing unit 220 shown in FIG.
When the process is started, the CPU first inputs a designation from the operator regarding a road to be processed (hereinafter referred to as “target road”) (step S20). Then, a connected image corresponding to the target road is input (step S22). In this embodiment, each road is traveled a plurality of times while changing the travel position, and a road surface image is taken. Therefore, a connected image is generated based on the path corresponding to each run. In step S22, the plurality of connected images are read.

  Next, the CPU sets a reference path (step S30). The reference path is a path that serves as a reference when positioning a plurality of paths. In the present embodiment, among the paths corresponding to the target road, the evaluation value of the position accuracy, that is, the one having the highest self-estimated position accuracy is selected. The reference path setting method will be described later.

When the reference path is set, the CPU performs processing for setting corresponding points for each path in accordance with the operation of the operator (step S40).
In this embodiment, as shown in the figure, a method of displaying a linked image of a reference path and a standard path on a display, and an operator operating a pointing device such as a mouse to set corresponding points in this screen. Was taken. In the example of the figure, the vertex of the pedestrian crossing warning sign having a rhombus shape in the standard path image is designated as the corresponding point, and then the corresponding vertex is designated in the reference path image. The corresponding points are not limited to one point, and a plurality of points can be designated.
When a paint recognition process, which will be described later, is performed and a sign in each connected image has been extracted, the CPU displays the extracted sign on the display, and the sign that the operator should use as a corresponding point from this is displayed. You may make it select. Further, it is possible to identify a corresponding label based on the positional relationship between the labels extracted between connected images, and to automatically set corresponding points and transparent polygons.

  In this embodiment, a low-resolution image is used for displaying the connected image. By doing so, there is an advantage that when the corresponding point is designated, the display can be smoothly moved, enlarged and reduced, and the work efficiency can be improved.

When the corresponding point is designated, the CPU performs a process of moving the standard path connected image to match the reference path connected image so that the corresponding points match each other, and ends the alignment process (step S50). ).
As described above, in the present embodiment, the connected image is not generated as a single composite image, but an orthographic image (road surface texture) is arranged and displayed. Therefore, in the process of step S50, the movement process of a connected image is performed by moving each orthogonal image (road surface texture). Along with the movement process, a process of replacing each orthogonal image from a low resolution image to a high resolution image is performed. Processing for rearranging the orthographic image may be performed using the high-resolution image.
The contents of the connected image moving process will be described in detail later.

C3. Standard path setting process:
FIG. 9 is a flowchart of the reference path setting process. This process corresponds to step S30 of the alignment process (FIG. 8), and is a process for setting a reference path having the highest self-estimated position accuracy when aligning a plurality of paths.

  When starting the processing, the CPU inputs the position accuracy at each point where the frame image is acquired for each path of the target road (step S31). At the time of shooting, as shown in the figure, a frame image is shot along the path at points P91, P92, P93, etc., and position accuracy AC1 in the east-west direction and position accuracy AC2 in the north-south direction are recorded for each point. Yes.

  In general, it is known that the detection accuracy of the GPS 114 varies depending on the arrangement of artificial satellites used for detecting position coordinates, the reception status of radio waves, the presence of multipath by receiving radio waves reflected on buildings, and the like. Yes. In differential positioning, the detection accuracy is also affected by the operating status of the reference station. The position accuracy is a quantitative evaluation of these effects. The position accuracy can be arbitrarily defined. For example, a precision reduction rate (DOP (Dilution of Precision)) or the like may be used.

  The CPU calculates the self-estimated position accuracy σ for each path based on the position accuracy of each point (step S32).

The self-position estimation accuracy may be a value determined based on a difference between GPS and IMU, DMI, or the like. In this case, for example, a standard deviation of the deviation amount may be used. Alternatively, the sum of the square of the standard deviation in the east-west direction and the square of the standard deviation in the north-south direction may be obtained, and this square root may be used as the self-position estimation accuracy. As described above, when the GPS, IMU, DMI, and other values are used, the self-position estimation accuracy increases as the deviation increases. That is, the smaller the value of the self-estimated position accuracy is, the higher the accuracy is.
When the self-estimated position accuracy σ of each path is obtained, the CPU sets the path having the minimum value as the reference path (step S33). If there is only a single path for the target road, that path is unconditionally set as the reference path. Let the self-estimated position accuracy of this reference path be σ _B.

If the self-estimated position accuracy σ _{B of} the reference path set in step S33 is lower than the predetermined threshold σ _TH (step S34), the reference path setting process is terminated.
On the other hand, if the self-estimated position accuracy σ _B is equal to or greater than the predetermined threshold σ _TH , an error display is performed (step S35), and the process ends. In this case, it means that the position accuracy of the reference path is not sufficiently ensured, and therefore the position accuracy is not sufficiently ensured even if the alignment process is performed.
As described above, the predetermined threshold σ _TH can be arbitrarily set based on the position accuracy to be secured as the road marking map.

It is only the self-estimated position accuracy σ _B of the reference path that is the object of determination as to whether or not to perform error display (step S35). This is because, with respect to other standard paths, even if the self-estimated position accuracy is low, it is possible to improve the position accuracy by performing alignment with reference to the reference path.
However, it can be said that the correction in the alignment process is preferably as small as possible for any path. Therefore, in step S34, the self-estimated position accuracy of all the paths is compared with the threshold value _σTH, and if any one of the paths has an accuracy lower than the threshold value, an error display may be performed. .
However, if the standard path is required to have the same position accuracy as the reference path, there is a possibility that error display is frequently performed. In order to avoid such an adverse effect, a higher threshold σ _TH may be used in the standard path than in the reference path. In other words, for the standard path, the positional accuracy requirement is relaxed compared to the reference path. By doing so, it is possible to avoid frequent error display while guaranteeing the minimum position accuracy for the standard path.

C4. Connected image movement processing:
(1) Flow chart:
FIG. 10 is a flowchart of the connected image movement process. This corresponds to the processing in step S50 of the alignment processing (FIG. 8).
When the process is started, the CPU inputs standard path data and corresponding point data to be moved (step S51). The standard path data is trajectory data including a point sequence in which position coordinates when a frame image is taken are sequentially recorded. The corresponding point data is the coordinate value of the corresponding point designated by the operator in the screen on which the reference path and the standard path are displayed in step S20 of FIG.

Next, the CPU calculates a movement vector for each point where the orthographic image (road surface texture) is arranged on the standard path (step S52).
An example of movement vector calculation is shown in the figure. In this example, it is assumed that corresponding points P101 and P103 are designated for the standard path NP10. On the standard path, an orthographic image (road surface texture) is arranged as shown by a trapezoid in the drawing.

  As points corresponding to the corresponding points P101 and P103, it is assumed that corresponding points P102 and P104 are designated on the reference path. The CPU obtains a movement vector for the corresponding points based on these designation results. In the illustrated example, a movement vector V10 from the corresponding point P101 to P102 of the standard path and a movement vector V11 from the corresponding point P103 to P104 are obtained.

  The corresponding point is not necessarily specified on the standard path NP10 because the operator specifies a point such as the apex of the sign that the operator can easily correspond with the standard path and the standard path. When the corresponding point is specified at a location deviated from the standard path NP10, the movement vector V10a is obtained at a location deviated from the standard path NP10 as indicated by a broken line in the drawing. Therefore, the movement vector V10 may be obtained by moving in the vertical direction to the standard path NP10 so that the starting point of the movement vector V10a is on the standard path NP10.

  When the movement vectors V10 and V11 at the corresponding points are obtained, the CPU obtains the movement vectors at the respective points located between the corresponding points P101 and P103 by interpolating these. For example, as shown in the figure, when the movement vector is obtained at the shooting point PP10 of the frame image, the movement vectors V10 and V11 are translated so that this point is the starting point, and a line segment connecting the end points of both vectors is obtained. Is internally divided by the ratio of the distance between corresponding points P101 to PP10 and the distance between P103 to PP10. By doing so, it is possible to obtain a movement vector VP10 starting from the point PP10 and ending at this internal dividing point.

For points that do not exist in the section between the two movement vectors V10 and V11, the movement vector at the closest position is used as it is. In the example in the figure, the movement vector V10 is used as it is in the section on the right side of the point P101, and the movement vector V11 is used as it is in the section on the left side of the point P103.
Further, when only one corresponding point is specified and only one movement vector is given, this movement vector is used.

The CPU translates the orthographic image (road surface texture) according to the movement vector obtained by the above processing (step S53), and ends the connected image movement processing. In the example of the figure, an example is shown in which the road surface texture TX11 arranged at the point PP10 of the standard path NP10 is translated to the position of the road surface texture TX12 according to the movement vector VP10.
Along with this processing, the position of the point PP10 on the standard path NP10 is also corrected by the movement vector VP10. Therefore, in the process of step S53, the locus of the standard path NP10 is also corrected along with the movement of the road surface texture.

(2) Positioning processing example (1):
FIG. 11 is an explanatory view showing a processing example (1) of alignment processing. Each of FIGS. 11A to 11C shows a state in which the linked images for the standard path NP11 and the reference path BP11 are displayed in an overlapping manner. FIG. 11A shows a state in which the connected image of the standard path NP11 is arranged above the connected image of the reference path BP11. As described above, the connected image is configured by arranging a large number of road surface textures, but for the sake of convenience of explanation, one road surface texture TX11 is illustrated with an outline.
The operator designates the corresponding point P111 in the standard path NP11 in this screen. The corresponding point P111 can be arbitrarily set. In this embodiment, one of the end points of the white stripe diagonal stripe pattern of the separation band sign M11 is selected as the corresponding point P111.

  FIG. 11B shows a state where the connected images of the reference path BP11 are arranged on the upper side. In this state, the positions of the standard path NP11 and the reference path BP11 are shifted. Therefore, when the connected image of the reference path BP11 is displayed on the upper side, the position of the corresponding point P111 is deviated from the diagonal stripe pattern of the white line of the separation band sign M12.

FIG. 11C shows a state in which the corresponding point P112 is specified with the connected image of the reference path BP11 facing upward. That is, in the image with the reference path BP11 on the upper side, the end point of the white stripe of the separation band sign M11 may be selected as the corresponding point P112.
When the corresponding point P112 is designated, the movement vector V11 is obtained so as to go from the corresponding point P111 of the standard path NP11 to the corresponding point P112 of the reference path BP11. If the road texture TX11 is moved according to the movement vector V11, the corresponding point P111 coincides with the corresponding point P112, and the positions of the separation band signs M11 and M12 can also coincide.

  In the alignment process, not only the road surface texture TX11 but also other road surface textures constituting the standard path NP11 are similarly moved according to the movement vector V11. Although an example of processing in which only one corresponding point is specified is shown here, a plurality of corresponding points may be specified. For example, in the example of the figure, it is conceivable to use the crossing stripe pattern, the stop line, the end point of the lane boundary line, or the like as the corresponding point.

(3) Positioning processing example (2):
FIG. 12 is an explanatory view showing a processing example (2) of alignment processing. The state where the connected images of the standard path NP12 and the reference path BP12 are overlapped is shown. For convenience of explanation, both road markings are shown in a visible state. Before the alignment, the positions of the standard path NP12 and the reference path BP12 are deviated, and thus the positions of the markings such as the lane boundary line are deviated.
Here, the operator selects one of the end points of the lane boundary line indicated by a broken line as a corresponding point. For the standard path NP12, the end point of the lane boundary line L122 is selected as the corresponding point P122, and for the reference path BP12, the end point of the lane boundary line L121 is selected as the corresponding point P121. As a result, a movement vector V12 from the corresponding point P122 of the standard path NP12 toward the corresponding point P121 of the reference path BP12 is determined.

FIG. 13 is an explanatory diagram showing a processing result of the positioning processing (2).
As described above, the position of the lane boundary line can be adjusted by moving the connected image of the standard path NP12 according to the movement vector V12. The result of the alignment is the lane boundary line L13.
In addition, by this alignment process, the standard path is also aligned with the position of the reference path. In the present embodiment, an image of a road surface is generated by aligning a plurality of paths originally traveling at different positions. At this time, as can be seen from the comparison between FIG. 12 and FIG. 13, by moving the standard path in parallel according to the movement vector set based on the corresponding point, the positional relationship of the road marking and the position of the path between the plurality of paths. Relationships can be matched very well.

(4) Acquisition of absolute coordinates:
FIG. 14 is an explanatory diagram showing a method for obtaining the absolute position coordinates of the road marking. In the illustrated example, the road surface texture TX142 on the standard path NP14 and the road surface texture TX141 on the reference path BP14 are illustrated. In the road surface textures TX141 and TX142, signs M141 and M142 are included, respectively.
The road surface textures TX141 and TX142 are arranged so that their representative points coincide with a point P141 on the reference path BP14 and a point P143 on the standard path NP14.

  In the road surface texture TX141, the position of the vertex P142 of the sign M141 can be specified by relative coordinates (x142, y142) with the representative point as the origin. Therefore, if the absolute coordinates of the representative point, that is, the position coordinates (X141, Y141) where the road surface texture TX141 is arranged are known, the absolute position of the vertex P142 of the sign M141 is added by adding the relative coordinates described above. Coordinates can be acquired.

  Similarly, in the road surface texture TX142, the position of the vertex P145 of the sign M142 can be specified by relative coordinates (x145, Y145) with the representative point as the origin. Accordingly, if the absolute coordinates of the representative point, that is, the position coordinates (X143, Y143) where the road surface texture TX142 is arranged are known, the relative position described above is added to the absolute position of the vertex P145 of the sign M142. Coordinates can be acquired.

  For the road surface texture TX142, it is assumed that the position P143 of the representative point is moved to the point P144 according to the movement vector V14 by the alignment process. At this time, the absolute position coordinates of the point P144 after alignment can be obtained by adding the components (VX14, VY14) of the movement vector V14 to the position coordinates (X143, Y143) of the point P143 before movement. Furthermore, if the relative coordinates (x145, Y145) of the point P145 are added to the absolute position coordinates of the point P144 obtained in this way, the absolute position coordinates of the vertex P145 of the sign M142 after the alignment processing are obtained. Can do.

  Here, the method of obtaining the absolute position coordinates for the vertices of the signs M141 and M142 in the road texture has been shown. However, each arbitrary point in the road texture is a relative coordinate based on the representative point of the road texture. Since it can be specified, the absolute position coordinates of an arbitrary point can be obtained by a similar method.

C5. Transparent polygon setting process:
(1) Process overview:
FIG. 15 is an explanatory diagram showing an outline of the transparent polygon setting process. In the transparent polygon setting process, the transparent polygon is set on the superimposed road image according to the operator's instruction so that the orthographic image corresponding to the adjacent path overlaps at the upper orthographic image. This is a process in which a part of the image is made transparent so that the lower orthographic image can be seen through.
In the center of the figure, a state in which the orthographic image P151 is superimposed on the orthographic image P152 is shown in a perspective view. The lower orthographic image P152 includes the pedestrian crossing A154 in a divided state, and includes the stop line A153 in a complete state. In the upper orthographic image P151, the pedestrian crossing A152 is included in a complete form, and the stop line A151 is included in a divided state. Each divided part is shown surrounded by a broken line.
When the orthographic images P151 and P152 are overlapped in this state, they are displayed as shown on the left side. That is, in the part where both overlap, only the image of the upper orthographic image P151 is displayed, so the pedestrian crossing A152 is displayed in a complete state, but the stop line A151 is shown in a divided state. It is.
If the vertical relationship between the orthogonal images P151 and P152 is changed, the stop line A153 can be displayed in a complete state, but the pedestrian crossing A154 is displayed in a divided state. . Thus, it is not possible to display both the pedestrian crossing and the stop line in a complete state only by the vertical relationship between the orthogonal images P151 and P152.

  Therefore, in this embodiment, the transparent polygon POL15 is set. In this example, the upper orthographic image P151 is set so as to cover the divided stop line A151. In the transparent polygon POL15, the upper orthographic image P151 is displayed in a transparent state. Therefore, as shown on the right side of the figure, when the orthographic images P151 and P152 are overlapped, the lower orthographic image P152 is displayed inside the transparent polygon POL15, and the upper orthographic image is displayed in other portions. An image P151 is displayed. As a result, the stop line A153 included in the lower orthographic image P151 and the pedestrian crossing A152 included in the upper orthographic image P152 are displayed, and both the stop line and the pedestrian crossing can be displayed in a complete form. it can.

(2) Flow chart:
FIG. 16 is a flowchart of the transparent polygon setting process. In terms of hardware, this is a process executed by the CPU of the road marking map generating apparatus 200. This corresponds to the processing of the transparent polygon setting unit 221 shown in FIG.
When the process is started, the CPU inputs designation of the target road from the operator (step S100), and inputs a connected image corresponding to the target road (step S102). When a plurality of paths correspond to the target road, a plurality of connected images corresponding to these paths are input.

The CPU displays these connected images and inputs designation of a priority path based on the operation of the operator (step S104). The priority path refers to a path having the best road image among a plurality of paths, and refers to a path positioned at the top when overlapping connected images of a plurality of paths. The priority path is different from the reference path used in the alignment process. The reference path means the one with the best position accuracy, but just because the position accuracy is good does not mean that the road surface image is good. Regardless of the top-to-bottom relationship of the overlay of connected images between multiple paths, alignment can be performed without any problem, so the alignment reference path and priority path are independent of each other. And can be set.
In this embodiment, the priority path can be arbitrarily set while the operator compares the linked images of the paths. The path with the roughest road surface image may be designated as the priority path. In such a case, only the number of transparent polygons to be described later increases.

When the priority path is set, the CPU sets a transparent polygon according to the operation of the operator (step S106).
An example of transparent polygon setting is shown in the figure. In this example, the road surface texture TX161 along the priority path and the road surface texture TX162 along the other paths are shown.
At the time of shooting, a rectangular image becomes a trapezoid by orthographic image conversion. Therefore, when the road surface textures TX161 and TX162 are arranged, it becomes a saw blade shape as shown in the figure. From the saw blade portion, the appearance of the road surface image is deteriorated and only the divided road surface image can be obtained. Therefore, the portion is unnecessary for the purpose of obtaining a complete road surface image. Therefore, in the example shown in the figure, in the portion where the road surface textures TX161 and TX162 overlap, the transparent polygon POL161 is set at the left end portion of the road surface texture TX161 having a saw blade shape so that the saw blade portion is not displayed. ing.

On the other hand, transparent polygons are not set in the portions where the road surface textures TX161 and TX162 do not overlap, in the example shown in the figure, the regions A161 and A162 at both ends. This is because in this portion, the images obtained by the road surface textures TX161 and TX162 are the only image information. If transparent polygons are set in the regions at both ends, the information on the road surface image included in this portion cannot be used. In the present embodiment, the information on the road surface image included in the road surface texture can be effectively used by not setting the transparent polygon in the portion that does not overlap with the other road surface texture.
Such a setting may be realized by an operation in which the operator sets a transparent polygon by simply avoiding a portion where the road surface texture does not overlap, but in the transparent polygon setting process (step S106), the transparent polygon You may make it restrict | limit a setting position. In other words, the transparent polygon setting operation by the operator may be accepted only for the portion where the road surface texture overlaps.

  When there is a sign hidden by the road surface texture TX161, the operator sets the transparent polygon so that the sign can be visually recognized. In the example shown in the figure, the transparent polygon POL 162 is set so as to cover the arrow mark. The arrow mark is an image included in the texture arranged on the lower side of the texture TX161.

  In order to set the transparent polygon POL162 that covers the sign in this way, the vertical relationship is temporarily changed so that the road texture TX161 is positioned below the other road texture, or the road texture TX161 is not displayed. Just do it. With these operations, the sign hidden in the road texture TX161 is made visible, the transparent polygon POL162 is set so as to cover the sign, and the display of the road texture TX161 is restored.

  When the setting of the transparent polygon is completed by the above processing, the CPU outputs the setting result and ends the transparent polygon setting processing.

(3) Processing example:
FIG. 17 is an explanatory diagram showing an example of a road image before setting a transparent polygon. In this example, a road image generated by performing alignment of connected images obtained along two paths P171 and P172 is shown. The shape of both ends of the saw blade in the connected image of the path P172 and the connected image of the path P171 are opposite to each other in the direction in which the vehicle of the road surface photographing system 100 travels on these paths P171 and P172. This is because the opposite is true.

In a portion where the connected image of the path P172 overlaps with the connected image of the path P171, a saw-toothed boundary at the end B17 of the connected image of the path P172 appears, degrading the image quality of the road image. However, in FIG. 17, for the convenience of illustration, the shape of the end portion B <b> 17 is emphasized with a saw-blade contour.
Further, since the road surface image of the path P172 is unclear at the end, for example, in the region A171, the stripe pattern of the pedestrian crossing is distorted. In the area A172, the stop line is divided. In the area A173, the character “bus only” indicating that the route is a preferential traffic zone such as a route bus (so-called bus lane) is broken so that it cannot be read. In the area A174, the broken-line lane boundary line is divided in the middle.

In order to avoid these influences, in FIG. 17, the transparent texture POL <b> 17 including the regions A <b> 171 to A <b> 174 and the end B <b> 17 is set as indicated by a one-dot chain line in the drawing.
When the transparent polygon POL17 is set in this way, the road surface texture on the path P172 side is in a transparent state inside the transparent polygon POL17, and the road surface texture on the path P171 side arranged below is visually recognized.

FIG. 18 is an explanatory diagram showing an example of a road image after setting a transparent polygon. Due to the action of the transparent polygon described above, the lower image is displayed in the area A181, so that the divided state of the pedestrian crossing shown in FIG. 17 is eliminated. Similarly, in the area A182, the stop line is displayed in a complete state. Further, as exemplified in the region B18, the saw-tooth profile at the end of the road surface texture is not visually recognized, and the image quality of the entire road image is improved.
In the area A183, the bus-specific characters are clearly readable. In the area A184, the lane boundary line is displayed in a complete state.
As described above, in this embodiment, by setting the transparent polygon, the image quality of the road image can be improved and the image quality of the marking on the road surface can be improved.

C6. Paint recognition process:
(1) Overall processing:
FIG. 19 is a flowchart of the paint recognition process. In terms of hardware, this is a process executed by the CPU of the road marking map generating apparatus 200. This corresponds to the processing of the paint recognition unit 223 shown in FIG.
When the processing is started, the CPU reads from the processing data storage unit 210 an image of a road to be processed, that is, road texture and path data (step S300).
Next, the CPU performs vertical arrangement processing (step S302), generates a vertical arrangement image, and stores it in the processing data storage unit 210.

  FIG. 20 is an explanatory diagram showing the contents of the vertical arrangement processing. FIG. 20A shows a connected image in normal processing. A one-dot chain line arrow PA20 in the figure represents a path when an image is taken. In normal processing, the road surface texture Tx20 is arranged along the path PA20. The position coordinates of the path PA20 are obtained in an absolute coordinate system XY such as latitude and longitude. Therefore, when the path PA20 and the road surface texture Tx20 are arranged in the absolute coordinate system, the connected image may be displayed obliquely as shown in FIG. In this example, a straight road is illustrated, but when the road is curved, the connected image is also curved.

FIG. 20B shows an example of vertically arranged images. Since the road is straight, the image in FIG. 20A is rotated in the direction of arrow A20.
The vertically arranged image can be generated by the following procedure. First, a distance axis is set in the longitudinal (y) direction of the two-dimensional coordinate xy. The distance axis is an axis representing a moving distance from the start point when an image is taken along the path PA20. When the path PA20 is linear, the traveling direction of the path PA20 is displayed upward. When the path is curved, the path is straightened.

For each road texture Tx20, since the position coordinate data at the time of shooting and the movement distance from the start of shooting are obtained (see the position measuring unit 110 in FIG. 1), the road texture on the distance axis is based on these data. Tx20 is arranged. The road surface texture Tx20 is arranged such that the representative point in the image is placed on the distance axis and the left-right symmetry axis is parallel to the distance axis. By doing so, it is possible to display a connected image in which the traveling direction is linearly extended in the vertical direction regardless of whether the path is curved. According to this connected image, as shown in the figure, the pedestrian crossing and the stop line are drawn in a direction perpendicular to the distance axis (left and right direction in the figure), and the lane boundary line is in a direction along the distance axis (up and down direction in the figure). Drawn in. In the vertically arranged image, road markings are drawn in a fixed positional relationship in this way, so that there is an advantage that these can be easily recognized.
Here, an example is shown in which the distance axis is arranged vertically, but the distance axis can be arranged in any direction such as laterally or diagonally.

Returning to FIG. 19, the paint recognition process will be described.
When the vertical arrangement process is completed, the CPU performs a crossing-related paint extraction process (step S310). This is a process of extracting signs in the vicinity of intersections such as pedestrian crossings, bicycle crossings, and stop lines. Details of the processing will be described later. In the present embodiment, a vertically arranged image stored in the processing data storage unit 210 is used, and a sign drawn there is extracted by image processing, and a condition based on the positional relationship, length, etc. of the line segment included in the sign Judgment type is judged by judgment.

When the crossing-related paint process ends, the CPU performs various paint extraction processes (step S350). Indices to be extracted in this process are shown in the figure.
The “boundary line” is a lane boundary line drawn with a solid line or a broken line.
“Arrow” is an arrow shown in each lane in the vicinity of the intersection in order to indicate regulation of the traveling direction within the intersection.
Unlike a pedestrian crossing, the “zebra” is a striped pattern that is marked on the median strip or where there are more lanes for turning left and right.
“U-turn” is a U-shaped arrow drawn on a U-turn prohibited road.
“Turning prohibited” is a cross marked with a U-turn arrow.
“Regulation” means time marking of traffic regulation. For example, traffic mode restrictions such as a sign “17-19” (meaning from 17:00 to 19:00) drawn together with a sign such as a bus lane.

The “number” is a number such as speed regulation.
“Pedestrian crossing notice” is a diamond-shaped symbol drawn in front of the pedestrian crossing.
The “deceleration zone” is a striped pattern drawn by arranging a plurality of line segments in a direction perpendicular to the path on the road surface in parallel along the traveling direction of the path in order to promote the deceleration of the vehicle speed. .
“Road surface painting” is a pavement area such as red, which is provided for safety of traffic at sharp curves and other places where the driver should be alerted.
The “bus lane character” is a character “bus exclusive” or “bus priority” attached to a lane used as a bus lane. In the present embodiment, the bus lane is illustrated, but the present invention is not limited to the bus lane, and may be a general character displayed on the road surface.
The “end symbol” is a “0” -shaped symbol indicating an end point such as a bus lane.
In the present embodiment, these signs are targeted, but these are merely examples, and more signs may be targeted, and some of them may be excluded from the extraction process.

In various paint extraction processes (step S350), pattern matching is performed using a model prepared in advance. In this embodiment, two types of artificial model images and OCR model images are used. These models are stored in the processing data storage unit 210 in advance.
An artificial model image is a model generated by computer graphics. It is suitable as a model for relatively simple marking matching, such as U-turn and turn prohibition.
An OCR model image is a model generated based on an image manually cut out by an operator from a captured image. For example, it is suitable as a model for matching signs having complicated shapes such as numbers and bus lanes. Although it is not impossible to generate these models by computer graphics, in order to generate a model by combining the shape of characters with actual road markings, it is necessary to trace the captured image after all. As a result, it is not much different from using the OCR model.

The CPU executes the above processing in units of connected images, stores the paint recognition result in the processing data storage unit 210, and ends the paint recognition processing.
Hereinafter, the crossing-related paint extraction process (step S310), various paint extraction processes (step S350), and the relative coordinate conversion process (step S370) will be described in detail.

(2) Cross-related paint extraction processing:
FIG. 21 is a flowchart of the cross-related paint extraction process. When this process is started, the CPU reads a vertically arranged image to be processed (step S311).
Next, the CPU performs side cut processing (step S312). The side cut process is a process of deleting a predetermined range at both ends of the vertically arranged image. Any of the two processing examples shown below may be used for the side cut processing.

(3) Side cut processing:
FIG. 22 is an explanatory view showing a side cut processing example (1). FIG. 22A shows a connected image in which the road surface texture Tx22 is arranged. Since the orthographic image becomes an inverted trapezoid at the time of image conversion, the connected image becomes jagged in a sawtooth shape at both ends. The jagged portion is a portion in which a large distortion is caused by the conversion to the orthographic image, and even if a sign is included, the positional accuracy is not necessarily high. In addition, the mark included in the jagged portion may be partially missing, and it is highly possible that it cannot be obtained in a complete form. For this reason, in this embodiment, this jagged portion is deleted as the side cut processing. However, if the part excluding the jagged portion is carelessly deleted, part or all of the sign drawn in the complete form may be deleted, so the area to be deleted should be barely deleted. Set to the extent that you can.

  In order to determine a range to be deleted at both ends, the CPU obtains a rectangular region RCT22 that includes a connected image. The rectangular center line CL22 is a straight line parallel to the distance axis, but is not necessarily coincident with the distance axis. When the size or position of the road surface texture TX22 varies, a portion that does not come into contact with the road surface texture Tx22 also appears in the rectangular RCT22 as shown in the region A22. The sides at both ends of the rectangle RCT22 are lines that define the outside of the area to be deleted.

FIG. 22B shows an example of processing for defining a line inside the area to be deleted. In the present embodiment, a circle C22 that is inscribed inside the jagged surface of the entire connected image is obtained. When the circle C22 set in this way is moved in the vertical direction in the drawing, a part of the jaggedness does not enter the circle C22. The tangent lines SC22L and SC22R that are in contact with the both ends of the circle C22 set in this way in the vertical direction are the boundary lines of the area to be deleted.
FIG. 22C shows an area to be deleted in the side cut process. The area to be deleted (the hatched area in the drawing) sandwiched between the sides of the rectangle RCT22 set in FIG. 22A and the tangent lines SC22L and SC22R set in FIG. It becomes.

FIG. 23 is an explanatory view showing a side cut processing example (2). FIG. 23A shows a connected image. For convenience of illustration, the connected images are shown here in a horizontal arrangement.
The CPU binarizes the connected image and converts the road portion into a solid image (FIG. 23B). In this state, the edge L231 of the road portion is extracted (FIG. 22C).
Next, the CPU thickens the extracted edge L231. An enlarged view of the edge L231 is shown in the lower right of FIG. The solid line indicates the extracted edge L231, and the broken line indicates the result of thickening. The amount of thickening can be arbitrarily set within a range where the jagged edges disappear.

By the thickening process, as shown in FIG. 23D, the jagged area becomes two bands L232 slightly thicker than the amplitude of the unevenness. A region W23 between the two bands L232 is a road image from which a jagged region is removed.
Therefore, the CPU generates a solid image of a portion corresponding to the region W23 as shown in FIG. Then, as shown in FIG. 23 (f), by extracting only the overlapping portion of the connected image (FIG. 23 (a)) and the solid image (FIG. 23 (e)), the connected state in which the jagged area is deleted. Get an image.

(4) Pedestrian crossing extraction:
Returning to FIG. 21, the crossing-related paint extraction process will be described.
When the side cut processing ends, the CPU performs edge extraction (step S313).
Extraction can be performed in various color systems such as RGB. In this embodiment, the connected image is converted into a color space of H (hue), S (saturation), and V (brightness), and then a marking area is extracted using the V (brightness) image. The V image is an achromatic image having only lightness as a component, and is a so-called gray scale image. As a result of the extraction in various color spaces, it was found that the sign drawn in white or yellow can be extracted with the highest accuracy when the V image is used.
When edge extraction is performed, contour lines of various signs in the connected image are extracted. This includes many striped patterns representing pedestrian crossings.

The CPU extracts a pedestrian crossing candidate area based on the edge extracted image (step S314). This process is a process for extracting candidate areas for a pedestrian crossing by deleting an area that is clearly too small or too large as a pedestrian crossing for the extracted edge image.
Next, the CPU performs a pedestrian crossing extraction process (step S316), a pedestrian crossing area process (step S317), and a pedestrian crossing information output process (step S318) for each extracted candidate area (step S315).
The pedestrian crossing extraction process (step S316) is a process of extracting a portion corresponding to the pedestrian crossing from the candidate areas for the pedestrian crossing.
The pedestrian crossing area process (step S317) is a process for obtaining an outer shape including a portion extracted as a pedestrian crossing.
The pedestrian crossing information output process (step S318) is a process for obtaining and outputting a rectangular shape that approximates the outer shape including the pedestrian crossing.

FIG. 24 is an explanatory diagram showing extraction processing of a pedestrian crossing candidate area. Each processing example of steps S314 to S318 in FIG. 23 is shown.
FIG. 24A shows an image obtained by extracting an edge from a connected image (hereinafter referred to as “edge image”). In this image, contour portions of various signs drawn on the road are extracted. A region A24 in the figure is a portion corresponding to a pedestrian crossing.
As the pedestrian crossing candidate area extraction process, the CPU deletes, from the edge image, an area that is too small as a pedestrian crossing (for example, area C24 in the figure) or an area that is too large (area B24 in the figure). In the edge image, the line of the pedestrian crossing included in the area A24 is not necessarily continuous, so the pedestrian crossing may be deleted in the same manner as the area C24. Therefore, in this embodiment, the pedestrian crossing candidate area extraction process is performed according to the following procedure.

  First, each line segment of the edge image is divided into a vertical line and a horizontal line. A line segment having an inclination within ± 45 ° with respect to the vertical axis may be defined as a vertical line segment, and a line segment having an inclination within ± 45 ° with respect to the horizontal axis may be defined as a horizontal line segment. . And these line segments are connected. As shown in a region C24 in the figure, the connection may be performed by thickening each line segment by a certain size in the vertical and horizontal directions. By this processing, adjacent line segments are connected to each other. The fattening amount may be set to a value such that each line segment constituting the pedestrian crossing is connected. After connecting the line segments by the thickening process, it is preferable to restore the original thickness by the thinning process.

  When the connection is completed in this way, a line segment that is too long and a line that is too short for a crosswalk is deleted. The threshold for deletion can be arbitrarily set based on the standard size as a pedestrian crossing. In this way, for example, the region B24 can be deleted as a too large portion, and the region C24 can be deleted as a too short portion.

FIG. 24B shows the result of extracting the pedestrian crossing candidate area. The CPU performs a pedestrian crossing extraction process, that is, a process of extracting a portion corresponding to the pedestrian crossing from the pedestrian crossing candidate area. In the example in the figure, the processing is to delete the areas D24 and E24 and extract the area A24. In this embodiment, statistically long lines, short lines, etc. that are extremely longer than other line segments are deleted.
In order to perform this process, the CPU first extracts vertical line segments in the area. Then, the position information of each line segment, that is, the center position, length, and width is obtained, and the average value and standard deviation thereof are obtained.

Next, the CPU deletes a line segment having a length exceeding “average height of line segment + standard deviation” as a line segment that is too long.
Next, the CPU obtains the horizontal interval of each line segment, its average value, and standard deviation. Then, in the case where the distance between the adjacent line segments exceeds the “average value + standard deviation”, in order to form a stripe pattern of the pedestrian crossing, it is determined that it is in a distant position and is deleted.
Further, the CPU linearly approximates the point sequence at the center position of the remaining line segment. Then, a line segment whose center position is at a predetermined distance or more from the approximate straight line is determined to be a line segment at a position shifted from the pedestrian crossing, and is deleted.

FIG. 24C shows the result of the pedestrian crossing extraction process. The vertical lines that make up the stripe pattern of the pedestrian crossing are extracted. The reason why there is a protruding portion in the region F24 is because the lane boundary line drawn close to the pedestrian crossing is erroneously recognized. Similarly, horizontal line segments can be extracted.
As the pedestrian crossing area process, the CPU obtains an outer shape that includes a portion extracted as a pedestrian crossing (FIG. 24C). In order to perform this processing, the CPU combines the vertical line segments in FIG. 24C and the horizontal line segments extracted in the same manner so that the portion corresponding to the stripe pattern of the pedestrian crossing becomes a closed figure.
Then, the area of each closed figure is fattened. Since the adjacent striped patterns are combined by this process, a convex region G24 shown in FIG. 24D is obtained. After such a shape is obtained, the region G24 may be contracted. By doing so, the convex portion is reduced, and an area close to a rectangle can be specified.

When the above processing is completed, the CPU finally obtains and outputs a rectangular shape that approximates the outer shape that includes the pedestrian crossing as pedestrian crossing information output processing. As shown in FIG. 24 (e), the size of the rectangle H24 may be determined so that the deviation from the broken line region obtained by the pedestrian crossing region process is minimized.
In the example of FIG. 24E, the case where the convex region G24 is used is shown. However, a reduction process is performed on the convex region G24 to reduce the convex portion, and then the rectangle H24 is obtained. May be.

(5) Extraction of bicycle crossing zone:
Returning to FIG. 21, the crossing-related paint extraction process will be described.
When extraction of the pedestrian crossing is completed for all candidate regions (step S315 in FIG. 21), the CPU performs bicycle crossing zone / stop line candidate region extraction (step S320). This process is a process of extracting these candidate areas by deleting an area that is clearly too small or too large as a bicycle crossing band / stop line with respect to the extracted edge image.
Next, the CPU performs a bicycle crossing zone / stop line extraction process (step S322) for each extracted candidate area (step S321). This is a process of extracting a portion corresponding to the bicycle crossing band / stop line from the candidate region for the bicycle crossing band / stop line.
If the extracted area is larger than the threshold value Wth (step S323), a bicycle crossing zone region process (step S324) and a bicycle crossing zone information output process (step S325) are performed.

If it is equal to or less than the threshold value Wth, stop line area processing (step S326) and stop line information output processing (step S327) are performed.
The bicycle crossing band region process (step S324) and the stop line region process (step S326) are processes for obtaining an outer shape including a portion extracted as a bicycle crossing band or a stop line.
The bicycle crossing band information output process (step S325) and the stop line information output process (step S327) are processes for obtaining and outputting a rectangular shape that approximates the outer shape including the bicycle crossing band or the stop line.

FIG. 25 is an explanatory diagram showing a process of extracting a bicycle crossing zone / stop line candidate region. Each processing example of steps S320 to S327 in FIG. 23 is shown.
FIG. 25A shows a connected image to be processed. The bicycle crossing zone is an area A25 composed of two straight lines drawn beside the pedestrian crossing and parallel to the pedestrian crossing.
FIG. 25B shows an edge image. Here, an example in which only a line segment in a direction orthogonal to the road path is extracted is shown. A region D25 represents a bicycle crossing zone, and a region C25 represents a stop line. Region B25 is noise.
The CPU extracts the next procedure, bicycle crossing zone / stop line area. First, a vertical line segment is combined with a horizontal line segment to form a region. Then, a fattening process is performed on each area. The fattening amount is set to a value such that two parallel lines constituting the bicycle crossing band are combined. As a result of the fattening process, the bicycle crossing band becomes one wide area where two lines are joined. After performing the fattening processing in this way, the CPU is too small (for example, a region B25 in the drawing) or a region that is too small as a line segment constituting the boundary line and the stop line of the bicycle crossing band for each region. Remove too much space.
Next, the CPU extracts a region in which the longitudinal direction extends in the direction orthogonal to the distance axis, that is, in the left-right direction, from among the remaining regions, and sets a thick one as a bicycle crossing band and a thin one as a stop line candidate region.

  FIG. 25C shows a road image E25 corresponding to a candidate region for a bicycle crossing zone. In this embodiment, in this way, after extracting the edge from the road image and determining the candidate area of the bicycle crossing zone, the road image corresponding to the determined area is cut out again, and this image is used for more details. The method of extracting the bicycle crossing zone is adopted.

FIG. 25D shows the result of edge extraction performed on the extracted image (FIG. 25C). The reason why the rectangular wavy irregularities appear on the upper side is that a part of the crosswalk drawn so as to be adjacent to the bicycle crossing zone is extracted. The CPU performs a bicycle crossing band extraction process, that is, a process of extracting a portion corresponding to the bicycle crossing band from the candidate areas of the bicycle crossing band.
In order to perform this process, the CPU performs a thickening process on the contour of the extracted area as shown in the enlarged view in FIG. The solid line indicates the outline, and the broken line indicates the result of the fattening process. When the thickening process is performed, the concavo-convex portions in FIG. 25D are combined with the adjacent portions and disappear. As a result, as shown in FIG. 25E, it is possible to obtain an outer shape that approximates a rectangle without unevenness. After combining with adjacent portions by thickening processing, the outer shape may be reduced to the extent that the original boundary line is obtained.

When the above processing is completed, the CPU finally obtains and outputs a rectangular shape that approximates the outer shape including the bicycle crossing band as a bicycle crossing band information output process. As shown in FIG. 25 (f), the size of the rectangle F25 may be determined so that the deviation from the broken line region obtained by the bicycle crossing band region process is minimized.
FIG. 25 shows an example of a bicycle crossing band, but the same applies to a stop line. As a result of performing the fattening process in FIG. 25B, a region having a width smaller than the bicycle crossing band may be extracted as a stop line.
With the process described above, the CPU completes the crossing-related paint extraction process.

(6) Boundary line (solid line) extraction processing:
FIG. 26 is an explanatory diagram showing the contents of the boundary line (solid line) extraction process. FIG. 26A illustrates an image including a lane boundary line. For convenience of illustration, the image is shown in a horizontally arranged state. As shown in the figure, this image includes solid lane boundary lines that divide the three lanes, stop lines, and arrows indicating the restriction of the traveling direction.

  FIG. 26B shows a state in which the sign portion is extracted by binarizing the image. In this embodiment, the image shown in FIG. 26A is converted into a color space of H (hue) S (saturation) V (lightness), and a portion having a lightness equal to or higher than a predetermined value is obtained using the V image. Extracted. The predetermined value for binarization may be set to a value with which a white line and a yellow line can be accurately extracted.

  Next, as shown in FIG. 26C, a line segment having a predetermined length or more is extracted from this image. The predetermined length can be arbitrarily set as long as an arrow or the like can be excluded. Depending on the image quality of the connected image, there is a possibility that the lane boundary line may be divided in the binarized image (FIG. 26B). Therefore, the line segment is connected before the line segment is extracted. May be. In other words, each line segment is thickened within a certain range to fill a minute gap, and then returned to the original thickness by a reduction process. By doing so, the lane boundary line can be extracted with high accuracy.

  In the state of FIG. 26 (c), a part of the stop line is extracted by mistake. Therefore, next, only the line segment parallel to the distance axis is extracted from the extracted line segments. In this figure, since the connected image is shown in a horizontal arrangement, a line segment in the left-right direction is extracted. By doing so, as shown in FIG. 26 (d), the stop line is eliminated and the lane boundary line can be extracted.

  The recognition result may be output as image data of the lane boundary line thus obtained. Although not shown, linear approximation may be applied to these lane boundary lines, and the obtained straight line may be output as a lane boundary line recognition result. Further, from the H (hue) image of the original image data (FIG. 26 (a)), a portion overlapping the region extracted in FIG. 26 (d) is extracted, and it is determined whether the lane boundary line is white or yellow. Also good.

(7) Boundary line (broken line) extraction processing:
FIG. 27 is an explanatory diagram showing the contents of the boundary line (broken line) extraction process. FIG. 27A illustrates an image including a lane boundary line. As shown in the drawing, this image includes a broken lane boundary line L27 that divides the three lanes, and a character C27 indicating restriction. Although not included in this figure, depending on the location, an arrow indicating the restriction of the traveling direction may be included.

FIG. 27B shows a state in which the sign portion is extracted by binarizing the image. The binarization method is the same as in the case of extracting a solid lane boundary line (FIG. 26B).
Next, a line segment having a predetermined length or more is extracted from this image. The predetermined length can be arbitrarily set within a range in which a dashed boundary line can be extracted while excluding an arrow or the like. Depending on the image quality of the connected image, there is a possibility that the lane boundary line may be divided in the binarized image (FIG. 27 (b)). May be. In other words, each line segment is thickened within a certain range to fill a minute gap, and then returned to the original thickness by a reduction process. However, as shown in FIG. 27A, the character C27 is drawn to such an extent that the lane boundary line L27 is likely to come into contact with the lane boundary line L27. preferable. The thickening process may be performed only in a direction parallel to the distance axis (vertical direction in the figure).
Moreover, when extracting a line segment, it is preferable to extract only a line segment in a direction parallel to the distance axis.

FIG. 27C shows a result of extracting a broken lane boundary line L27.
The recognition result may be output in the form of image data of the lane boundary line thus obtained. Although not shown, linear approximation may be applied to these lane boundary lines, and the obtained straight line may be output as a lane boundary line recognition result. Further, a portion overlapping with the region extracted in FIG. 27C is extracted from the H (hue) image of the original image data (FIG. 27A) to determine whether the lane boundary line is white or yellow. Also good.

(8) Arrow extraction process:
FIG. 28 is an explanatory diagram showing the contents of the arrow extraction process. FIG. 28A illustrates an image including arrows A281 and A282 and lane boundaries, pedestrian crossings, stop lines, and the like. For convenience of illustration, the distance axis is shown in a horizontal arrangement in the left-right direction.
FIG. 28B shows a state in which this image is binarized to extract a marking portion. The binarization method is the same as in the case of extracting a solid lane boundary line (FIG. 26B). Marked portions such as arrows A281 and A282 and a pedestrian crossing A283 are extracted from the binarized image.

Next, a region having a length and a width corresponding to an arrow is extracted from the image.
The length of the arrow is similar to the length of each line segment on the crosswalk. Although not shown in the figure, the size is similar to the speed regulation numbers on the road. Therefore, it is preferable to apply a thickening process to each line segment in advance in order to distinguish these marks from the arrows. The thickening process combines the stripe pattern of the pedestrian crossing and the line segments constituting the character, so that a large area that can be clearly distinguished from the arrow can be formed. Since this thickening process is performed with the aim of such an effect, it is preferable that the thickening amount be large in the direction orthogonal to the distance axis, that is, in the vertical direction in the figure.
By performing the fattening process in this way, pedestrian crossings, characters, and the like can be eliminated.

FIG. 28 (c) shows the result of eliminating pedestrian crossings and the like. In this embodiment, an arrow is extracted from this image by pattern matching using an OCR model. The OCR model is image data of an arrow for pattern matching generated by cutting out an image obtained by photographing a road surface in advance. The OCR model is prepared according to the type of arrow drawn on the road surface. For example, a straight arrow, a right turn arrow, a left turn arrow, a straight right turn arrow, a straight left turn arrow, and the like can be given.
In pattern matching, these OCR models are superimposed on the image of FIG. 28C, and the position and size of the OCR model are changed so that the deviation between the two is minimized. Then, it is determined whether or not the sign drawn on the image matches the OCR model according to the amount of deviation when it becomes the minimum. In this embodiment, the sign is recognized as an arrow when the reliability is 0.99 or more.

  The recognition result may be output in the form of the image data of the arrow thus obtained. Although not shown, the recognition result may be output in the form of an existing area for these arrows. That is, as shown in the extraction of the pedestrian crossing (FIG. 24 (e)), the position and size of the geometric shape showing the area where the pedestrian crossing exists are output as the recognition result. In the case of an arrow, the type of arrow and the position and size of a geometric shape such as a rectangle including the arrow can be output as a recognition result. By doing so, there is an advantage that the recognition result can be saved with low-capacity data.

(9) Zebra extraction processing:
FIG. 29 is an explanatory diagram showing the contents of the zebra extraction process. Zebra refers to a striped pattern on a road surface that is different from a pedestrian crossing or a deceleration zone described later. It is often provided at the part where the median strip and right / left turn lanes increase, and is characterized by a diagonal stripe pattern with respect to the distance axis.

FIG. 29 (a) shows an example of an image including a zebra. For convenience of illustration, the distance axis is shown in a horizontal arrangement in the left-right direction. In this example, a zebra A29 is drawn in the central separation zone.
FIG. 29B shows a state in which this image is binarized and a marking portion is extracted. The binarization method is the same as in the case of extracting a solid lane boundary line (FIG. 26B).
FIG. 29 (c) shows the result of extracting a region of a predetermined value or more from the binary image. The zebra A29 has a relatively large overall area because the boundary line in the direction along the distance axis (the left-right direction in the figure) and the hatched portion are coupled. The predetermined value for extracting the zebra A29 can be arbitrarily set in a range larger than the lane boundary line and smaller than the zebra A29.

Next, a region having a narrow width in the direction orthogonal to the distance axis is removed. For this purpose, each region is contracted in a direction (vertical direction in the figure) perpendicular to the distance axis. By doing so, the boundary line of the zebra A29 is removed as shown in FIG. After the boundary line is removed, a process for expanding each region in the vertical direction is performed so as to restore the contraction.
This time, the striped pattern obtained in FIG. 29D is thickened in the direction along the distance axis. By doing so, striped patterns are combined and a solid region including the entire zebra can be obtained.
Then, a zebra is extracted from the original image in this solid area. FIG. 29 (e) shows a state in which only the zebra A29 is newly cut out from the original image (FIG. 29 (a)) in the solid region thus obtained. That is, the processes from FIG. 29B to FIG. 29D are processes for narrowing down the candidate area of the zebra A 29 from the connected image. The part corresponding to the candidate region thus obtained is extracted again from the original image (FIG. 29A), and the zebra is extracted with higher accuracy.

The processing applied to the extraction result (FIG. 29 (e)) is the same as in FIGS. 29 (b) to 29 (d). In other words, the zebra B29 is extracted as shown in FIG. 29 (f) by binarizing the image of FIG. 29 (e), thickening the obtained line segments, and combining them. However, in this process, signs other than the zebra such as the lane boundary line are almost eliminated, so that the fattening amount in the fattening process can be set to an optimum value for extracting the zebra B29. is there.
In this manner, first, the candidate region where the zebra A29 exists is narrowed down, and the zebra B29 is extracted with high accuracy by performing the two-step process of processing the image cut out from the original image again based on the result. It is possible. This two-stage process can be applied not only to zebra but also to other signs.

(10) U-turn extraction process:
FIG. 30 is an explanatory diagram showing the contents of the U-turn extraction process. FIG. 30A shows a road image including a U-turn. For convenience of illustration, the distance axis is shown in a horizontal arrangement in which the distance axis is horizontal (in the horizontal direction in the figure). The U-turn arrow is an arrow A30 in the figure.

  In this embodiment, the U-turn arrow is extracted by pattern matching using an artificial model. FIG. 30B shows an artificial model of a U-turn arrow. The artificial model is created by computer graphics.

FIG. 30C shows a binarized image of the original image (FIG. 30A). The binarization method is the same as in the case of extracting a solid lane boundary line (FIG. 26B).
With respect to this image, for each region, the length in the direction along the distance axis (left-right direction in the figure) and the width in the direction orthogonal to the distance axis (up-down direction in the figure) are obtained. And the area | region which this length and width does not exist in the predetermined value range corresponded to a U-turn arrow is excluded. FIG. 30D shows an exclusion result. It can be seen that, in addition to the U-turn arrow A30, a turn prohibition symbol B30 (x mark in the figure) having an equivalent size is also extracted.
The artificial model shown in FIG. 30B is applied to the extraction result thus obtained, and a U-turn arrow is extracted by pattern matching. In the present embodiment, an indication having a reliability of 0.85 or higher is determined as a U-turn arrow.

  The recognition result may be output in the form of image data of the U-turn arrow thus obtained. Although not shown, the recognition result may be output in the form of an existing area for these U-turn arrows. That is, the position and size of a geometric shape such as a rectangle including the U-turn arrow can be output as a recognition result. By doing so, there is an advantage that the recognition result can be saved with low-capacity data.

(11) Inversion prohibition extraction processing:
FIG. 31 is an explanatory diagram showing the contents of the turn prohibition extraction process. FIG. 31A shows a road image including turning prohibition. For convenience of illustration, the distance axis is shown in a horizontal arrangement in which the distance axis is horizontal (in the horizontal direction in the figure). The rotation prohibition is a cross A31 in the figure.
In this embodiment, turning prohibition is extracted by pattern matching using an artificial model. FIG. 31 (b) shows an artificial model that prohibits turning. The artificial model is created by computer graphics.

FIG. 31C shows a binarized image of the original image (FIG. 31A). The binarization method is the same as in the case of extracting a solid lane boundary line (FIG. 26B).
With respect to this image, for each region, the length in the direction along the distance axis (left-right direction in the figure) and the width in the direction orthogonal to the distance axis (up-down direction in the figure) are obtained. And the area | region which this length and width does not exist in the predetermined value range corresponded to rotation prohibition is excluded. FIG. 31D shows the exclusion result. It can be seen that in addition to the rotation prohibition A31, a U-turn arrow B31 having an equivalent size is also extracted.
The artificial model shown in FIG. 31B is applied to the extraction result thus obtained, and the turn prohibition is extracted by pattern matching. In this embodiment, a sign having a reliability of 0.8 or more is determined to be prohibited from turning.

  The recognition result may be output in the form of image data for which rotation is prohibited thus obtained. Although not shown in the figure, the recognition result may be output in the form of an existing area for turning prohibition. That is, the position and size of a geometric shape such as a rectangle including the rotation prohibition can be output as a recognition result. By doing so, there is an advantage that the recognition result can be saved with low-capacity data.

(12) Regulation extraction process:
FIG. 32 is an explanatory diagram showing the contents of the restriction extraction process. FIG. 32A illustrates an image including the regulation A32. The distance axis is indicated by a vertical arrangement in the vertical direction.
Regulations refer to letters and numbers indicating traffic restrictions, such as letters representing time zones dedicated to bus lanes (numbers written as “17-19” in the figure). In the figure, the word “dedicated” is also visible, but this is not included in the regulation process because it is processed by the “bus lane” marking described later.
FIG. 32 (b) shows a state in which this image is binarized and a marking portion is extracted. The binarization method is the same as in the case of extracting a solid lane boundary line (FIG. 26B).

Next, a white area is extracted from the image area. The reason why white color is extracted is that in this example, the regulation is for white numbers. In the case of targeting a restriction drawn in yellow, it is preferable to extract a yellow region.
When white areas are extracted, as shown in FIG. 32C, fattening processing is performed on these areas. This is because, in the regulation extraction process, since signs such as numbers and letters are processed, it is easier to extract them by combining them once as one area. In the example of the figure, it can be seen that the line segments constituting the “dedicated” character B32 and the regulation numeral (17-19) A32 are combined to form a unified area. As described above, the fattening amount can be arbitrarily set within a range in which letters or numbers can be combined into one area.

After thickening the numbers and letters into one area in this way, the length of each area in the direction along the distance axis (vertical direction in the figure) and the direction orthogonal to the distance axis (horizontal direction in the figure) Based on the width, the region corresponding to the restriction is extracted. For example, in the example of FIG. 32C, the area where the “dedicated” character is crushed has a large length in the direction along the distance axis (the vertical direction in the figure), and can be distinguished from the restriction. The lane boundary line has a large length in the direction along the distance axis (up and down direction in the figure) and is thin in the direction orthogonal to the distance axis (left and right direction in the figure), so that it can be distinguished from regulation.
In this way, the restricted area can be extracted as shown in FIG.

  After extracting the restriction region, the portion overlapping this region is extracted again from the original image (FIG. 32 (a)), binarized, and the white region is extracted. Then, pattern matching using the OCR model is performed. Determine whether it is a regulation. The OCR model is regulated image data for pattern matching generated by clipping from an image obtained by photographing a road surface in advance. The OCR model is prepared according to the type of regulation drawn on the road surface. In this embodiment, when the reliability is 0.9 or more, the sign is recognized as the restriction.

  The recognition result may be output in the form of regulated image data obtained in this way. Although not shown in the figure, the recognition result may be output in the form of an existing area for these restrictions. That is, the position and size of the geometric shape showing the area where the restriction exists are output as a recognition result. For example, the type of restriction and the position and size of a geometric shape such as a rectangle including the restriction can be output as a recognition result. By doing so, there is an advantage that the recognition result can be saved with low-capacity data.

(13) Number extraction processing:
FIG. 33 is an explanatory diagram showing the contents of the number extraction process. FIG. 33A illustrates an image including the number A33. The distance axis is indicated by a vertical arrangement in the vertical direction.
Here, the number represents a speed regulation number. In the figure, the number “50” is drawn.
FIG. 33 (b) shows a state in which this image is binarized and a marking portion is extracted. The binarization method is the same as in the case of extracting a solid lane boundary line (FIG. 26B). At this time, only the yellow region may be extracted.

  Next, a fattening process is performed on the image as shown in FIG. This is because, in the number extraction process, two-digit numbers are processed, and it is easier to extract them by combining them once as one area. In the example of the figure, it can be seen that the numbers A33 “5” and “0” are combined to form a unified region. The fattening amount can be arbitrarily set in such a range that numbers can be combined into one area.

After thickening the numbers into one area in this way, the length of each area in the direction along the distance axis (vertical direction in the figure) and the width in the direction perpendicular to the distance axis (horizontal direction in the figure) Based on the above, a region corresponding to a number is extracted. For example, in the example of FIG. 33 (c), the lane boundary line has a large length in the direction along the distance axis (vertical direction in the figure) and a direction perpendicular to the distance axis (horizontal direction in the figure) is thin. Can be distinguished from numbers.
In this way, a numerical area A33 can be extracted as shown in FIG.

  After extracting the number area, the part that overlaps with this area is extracted again from the original image (Fig. 33 (a)), binarized, and then determined whether it is a number by pattern matching using the OCR model To do. The OCR model is prepared according to the type of numbers drawn on the road surface. In this embodiment, the sign is recognized as a number when the reliability is 0.9 or more.

  The recognition result may be output in the form of numeric image data obtained in this way. Although not shown, the recognition result may be output in the form of an existing area for these numbers. That is, the position and size of the geometric shape representing the area where the number exists are output as a recognition result. For example, the type of number and the position and size of a geometric shape such as a rectangle including the number can be output as a recognition result. By doing so, there is an advantage that the recognition result can be saved with low-capacity data.

(14) Crosswalk notice extraction process:
FIG. 34 is an explanatory diagram showing the contents of the pedestrian crossing trail extraction process. A pedestrian crossing notice is a rhombus sign in front of a pedestrian crossing. In this embodiment, an artificial model was used for extracting the pedestrian crossing notice. FIG. 34A shows an artificial model for notice of a pedestrian crossing. The artificial model is created by computer graphics.

FIG. 34B shows a road image including a pedestrian crossing notice. For convenience of illustration, the distance axis is shown in a horizontal arrangement in which the distance axis is horizontal (in the horizontal direction in the figure). The pedestrian crossing notice is a diamond-shaped sign A34 in the figure.
FIG. 34C shows a binarized image of the original image (FIG. 34B). The binarization method is the same as in the case of extracting a solid lane boundary line (FIG. 26B).
With respect to this image, for each region, the length in the direction along the distance axis (left-right direction in the figure) and the width in the direction orthogonal to the distance axis (up-down direction in the figure) are obtained. Then, an area in which the length and width are within a predetermined value range corresponding to the pedestrian crossing notice is extracted. FIG. 34 (d) shows the extraction result. It can be seen that two rhombus marks A34 are extracted.

  The artificial model shown in FIG. 34A is applied to the extraction result thus obtained, and a pedestrian crossing notice is extracted by pattern matching. In this embodiment, a sign having a reliability of 0.85 or more is determined to be a pedestrian crossing notice.

  The recognition result may be output in the form of image data for the pedestrian crossing notice obtained in this way. Although not shown, the recognition result may be output in the form of an existing area for the pedestrian crossing notice. That is, the position and size of a geometric shape such as a rectangle including a pedestrian crossing notice can be output as a recognition result. By doing so, there is an advantage that the recognition result can be saved with low-capacity data.

(15) Deceleration zone extraction processing:
FIG. 35 is an explanatory diagram showing the contents of the deceleration band extraction process. The deceleration zone refers to a sign in which a predetermined length line segment orthogonal to the distance axis is arranged in parallel to the distance axis direction.
FIG. 35A shows a road image including a deceleration zone. For convenience of illustration, the distance axis is shown in a horizontal arrangement in which the distance axis is horizontal (in the horizontal direction in the figure). The deceleration band is a striped pattern A35 in the figure.
FIG. 35B shows a binarized image of the original image (FIG. 35A). The binarization method is the same as in the case of extracting a solid lane boundary line (FIG. 26B). Since the deceleration zone is usually drawn in yellow / red, a yellow or red region may be extracted at this stage.

Next, a region having a predetermined area or more is extracted for each region of the image. Instead of the area, an area in which the length in the direction along the distance axis (left and right direction in the figure) and the width in the direction orthogonal to the distance axis (up and down direction in the figure) is a value corresponding to the deceleration band You may make it extract. FIG. 35C shows a result extracted based on the area. Regardless of the area, length, or width, a value that serves as a criterion for determining whether or not to extract may be adjusted according to the size of the deceleration zone. In FIG. 35 (c), it can be seen that the line segment constituting the striped pattern is partially broken.
The extraction result thus obtained is subjected to a fattening process on each area as shown in FIG. By doing so, it can be seen that the fractured portion seen in FIG. 35 (c) is filled.

Next, a region overlapping the region A35 subjected to the thickening process is extracted from the original image (FIG. 35A). FIG. 35 (e) shows the extraction result.
By extracting a yellow or red area from the image thus extracted, a deceleration zone can be extracted as shown in FIG. The image may be further subjected to contraction processing to delete the horizontal line generated in the figure. Alternatively, a very small area having a predetermined area or less may be deleted, and only the deceleration band may be clearly extracted.
Furthermore, you may arrange the length of each line segment which comprises a striped pattern.

  The deceleration band obtained in this way may output a recognition result in the form of image data. Although not shown, the recognition result may be output in the form of an existing area for the deceleration zone. That is, the position and size of a geometric shape such as a rectangle including the deceleration zone can be output as a recognition result. By doing so, there is an advantage that the recognition result can be saved with low-capacity data.

(16) Road surface paint extraction process:
FIG. 36 is an explanatory diagram showing the contents of the road surface paint extraction process. Road surface painting is a pavement area such as red that is applied for safety of traffic to sharp corners and other places where attention should be given to the driver.
FIG. 36A illustrates an image including road surface coating. The distance axis is shown as vertical (vertical direction in the figure). For convenience of illustration, the coating A36 is shown with hatching, but actually, the entire region is solidly painted in red.

  While the other markings are drawn in white or yellow, the painted portion to be treated here is drawn in red. Therefore, as the extraction of the painted portion, the red portion is extracted using the H (hue) image in the H (hue) S (saturation) V (lightness) color space. FIG. 36B shows the extraction result. Although the painted portion A36 has been extracted, since there is a change in color tone due to uneven coating, light reflection, shadows, etc., a part of the area leaks from the extraction, resulting in a hole H36.

In order to fill the hole H36 in the extracted area, the extracted area is thickened. Alternatively, the portion H36 surrounded by the extracted region may be unconditionally filled.
FIG. 36C shows a result of filling the hole H36 in this way. Although the hole H36 is filled, minute regions B36 are scattered around the hole H36.

Therefore, next, a region whose area is equal to or greater than a predetermined value is extracted. The predetermined value can be arbitrarily set to a value that can sufficiently extract the painted portion while excluding a minute region.
By doing so, the painted portion A36 can be extracted.
Since the painted portion A36 does not have a fixed shape, the recognition result is output as image data.

(17) Bus lane character extraction processing:
FIG. 37 is an explanatory diagram showing the contents of the bus lane character extraction process. The bus lane characters are characters such as “bus exclusive” and “bus priority”.
FIG. 37A illustrates an image including the bus lane character A37. The distance axis is indicated by a vertical arrangement in the vertical direction. In the figure, the numbers for "Bus only" are drawn in white.
FIG. 37 (b) shows a state in which this image is binarized and a marking portion is extracted. The binarization method is the same as in the case of extracting a solid lane boundary line (FIG. 26B). It can be seen that the lane boundary line is extracted in addition to the characters “bus only”.

  Next, a fattening process is performed on the image as shown in FIG. In this case, the processing is performed with a thickening amount that can combine the muddy point of “bus exclusive” with “c”. For this result, a process of deleting a minute area having a predetermined area or less is performed. By doing this, it is possible to eliminate a minute region while leaving the cloud point of “B”.

  After deleting the minute area, an area is extracted in which the length in the direction along the distance axis of each area (vertical direction in the figure) is a value corresponding to each character of “bus only”. Since the lane boundary line is long in the vertical direction, this makes it possible to eliminate the lane boundary line and extract only the characters “bus only”.

  A fattening process is performed on the result thus extracted as shown in FIG. This process is a process for recognizing a character composed of separated line segments such as “B” as one area. The fattening amount can be arbitrarily set within a range in which such characters are combined.

  When the fattening process is completed, the portion overlapping the region shown in FIG. 37 (d) is extracted again from the original image (FIG. 37 (a)), binarized, and then the bus lane characters by pattern matching using the OCR model. It is determined whether or not. The OCR model is prepared according to the type of bus lane character drawn on the road surface. In this embodiment, the sign is recognized as a bus lane character when the reliability is 0.9 or more.

  The bus lane character thus obtained may be output as a recognition result in the form of image data. Although not shown in the figure, the position and size of an existing area that includes all of these characters may be output as a recognition result. For example, the character type and the position and size of a geometric shape such as a rectangle including the character can be output as a recognition result. By doing so, there is an advantage that the recognition result can be saved with low-capacity data.

(18) End symbol extraction processing:
FIG. 38 is an explanatory diagram showing the contents of the end symbol extraction process. The end symbol is a sign having a shape in which an oblique line is drawn to “0” indicating an end point such as a bus lane.
FIG. 38A illustrates an image including the end symbol A38. The distance axis is shown in a vertical arrangement with the vertical direction.
FIG. 38 (b) shows a state in which this image is binarized and a marking portion is extracted. The binarization method is the same as in the case of extracting a solid lane boundary line (FIG. 26B).

  Next, as shown in FIG. 38C, this image is subjected to processing for filling holes in the area. In other words, a portion surrounded by the extracted area is filled in like an end symbol. By doing so, the end symbol becomes an area having a relatively large area, so that there is an advantage that it can be easily distinguished from other signs.

  When the hole filling process is completed, the area, length in the direction along the distance axis (up and down direction in the figure), and width in the direction perpendicular to the distance axis (left and right direction in the figure) are extracted as the end symbol. To do. FIG. 38D shows the extraction result. The character “bus” included in FIG. 38C is the end symbol, the length in the vertical direction, and the width in the left-right direction are the same as the end symbol, but it is composed of line segments and has a larger area than the end symbol. Can be distinguished from the end symbol. The lane boundary line can be distinguished from the end symbol because the vertical length is larger than the end symbol and the width in the left-right direction is narrower than the end symbol. In this way, only the end symbol can be extracted as shown in FIG.

  Next, a portion overlapping with the region shown in FIG. 38 (d) is extracted again from the original image (FIG. 38 (a)), binarized, and then whether or not it is an end symbol by pattern matching using an OCR model. judge. In this embodiment, the sign is recognized as an end symbol when the reliability is 0.9 or more.

  The end symbol thus obtained may be output as a recognition result in the form of image data. Although not shown, the position and size of the existing area including the entire end symbol may be output as the recognition result. By doing so, there is an advantage that the recognition result can be saved with low-capacity data.

(19) Relative coordinate conversion processing:
FIG. 39 is an explanatory diagram showing the contents of relative coordinate conversion processing. In the present embodiment, the recognition results of the various signs described above are all obtained using linear images in which road textures are arranged along the distance axis. Therefore, as a recognition result, only a relative position in the coordinate system displaying this linear image is acquired. The relative coordinate conversion process is a process of converting this relative position into an absolute coordinate system position corresponding to the position coordinate at the time of photographing.

  The left side of the figure shows a state in which the sign is recognized by a linear connected image. Here, rectangular existence areas M391 and M392 are shown. The positions of these existence areas M391 and M392 are represented by the position coordinates of the centroids G391 and G392 as representative points. These coordinates are given by a coordinate system xn, yn for displaying a linear connected image. In the coordinate systems xn and yn, for example, the distance axis NP39 is preferably defined as the yn axis.

  In the connected image, the representative points of the road surface texture Tx39 are arranged on the distance axis according to the position at the time of shooting. The relative coordinates from the representative point of each point in each road surface texture Tx39 are known. Accordingly, since the relative coordinates within the road surface texture Tx39 of the centroids G391 and G392 of the respective existence areas are obtained, the coordinates of the perpendicular legs R391 and R392 taken from the centroids G391 and G392 on the distance axis can also be obtained.

The right side of the figure shows a state converted to the absolute coordinate system XY. Since the road is not always a straight line, in the absolute coordinate system, the path RP39 may be drawn in a curved line according to the position coordinates acquired at the time of shooting.
The positions and orientations of the respective existence areas M391 and M392 are obtained by the following method.
First, in accordance with the distance from the reference point of the curved path RP39, the positions of the perpendicular legs R391 and R392 of the existing area on the path RP39 are obtained. Then, normal vectors V391 and V392 are drawn from the points R391 and R392, respectively, and the end points are set as the center of gravity G391 and G392. The magnitudes of the normal vectors V391 and V392 are equal to the distance between the points R391 and G391 and the distance between the points R392 and G392 in the linear connected image, respectively. The directions of the existence regions M391 and M392 are arranged so that the longitudinal direction is orthogonal to the normal vectors V391 and V392. In other words, the relative positional relationship between the existence areas M391, M392, the centroids G391, G392, and the perpendicular legs R391, R392 is expressed by a linear connected image (left side in the figure) and an absolute coordinate system (right side in the figure). The existence areas M391 and M392 are arranged so as to be the same.
By this conversion process, the position of each extracted sign can be expressed by the position coordinates of the absolute coordinate system.

You may perform the conversion process of the position coordinate to an absolute coordinate system mentioned above with the following method.
First, the constituent points of each existence area, that is, the positions of the vertices of the existence area are obtained with relative coordinates based on the representative point of the road texture to which each vertex belongs. Then, when each road surface texture is converted into an arrangement in the absolute coordinate system XY, the same conversion is applied to each constituent point. By doing so, the position coordinates of the constituent points in the absolute coordinate system can be obtained relatively easily.
However, in this method, since the coordinate conversion is performed for each constituent point, the shape of the existing area may be destroyed during the conversion. Therefore, a correction process may be performed to adapt the arrangement of the converted component points to the shape of the existing area.

D. effect:
According to the road surface photographing system 100 and the road marking map generating device 200 of the embodiment described above, the road surface texture obtained by orthogonal transformation of the frame image acquired while traveling on the road is arranged, thereby driving. A connected image with high positional accuracy can be obtained along a trajectory (path). Furthermore, a road surface image of the entire road can be obtained by aligning and synthesizing connected images obtained along a plurality of paths. At this time, a road surface image is generated while ensuring the overall position accuracy by adopting a method that uses the path with the highest position accuracy of each path when taking an image as a reference path and aligning other paths with this reference path. can do.
In the present embodiment, the connected images of the respective paths are limited until the road surface texture is arranged, and these are not combined as a single image. Therefore, it is possible to easily synthesize a multi-pass connected image by translating the arrangement in units of road surface texture.

  In this embodiment, both the generation of the connected image of each path and the synthesis of the combined images of a plurality of paths do not need to perform affine transformation on the road surface texture, and are performed by simple parallel movement. Therefore, it is possible to avoid deterioration in image quality due to complicated image processing, and to obtain a road surface image displayed with a clear road marking. Further, since the translation is performed by parallel movement, a relative coordinate system based on the representative point in the road texture is maintained before and after the generation and synthesis of the connected images. As a result, if the absolute position coordinates of the representative point are obtained, the absolute position coordinates of each point in the road texture can be easily obtained, and the absolute position coordinates of the road marking can also be obtained.

Further, in this embodiment, by arranging the road surface texture along the distance axis, it is possible to generate a linear connected image and extract a road surface sign. As a result, the pedestrian crossing has a striped pattern in which the line segments parallel to the distance axis are arranged in a direction perpendicular to the distance axis. It becomes possible.
Since the recognition result of the sign recognized in this way can be expressed by the position and size of the existing area, there is an advantage that the recognition result can be saved with relatively low volume of data.
Furthermore, the position of the recognized sign can be obtained in the absolute coordinate system by converting the coordinate system for drawing a linear connected image into the coordinate system of absolute coordinates. Therefore, according to the present embodiment, it is possible to accurately acquire the position of the road surface marking from the image of the road.

E. Variations:
(1) Impact of dust images:
In the embodiment, the process of recognizing the paint when the paint is copied relatively well in the connected image is exemplified. However, during actual shooting, various objects other than paint may appear. For example, when taking a picture while traveling on a road with a vehicle, another vehicle that stops or travels next to the lane in which the vehicle travels may appear in the screen.

FIG. 40 is an explanatory diagram illustrating the reflection of foreign matter.
FIG. 40A illustrates a frame image. The inside of the frame A40 around the center in the figure is a portion used for generating road surface texture (see FIG. 7). In this example, it is assumed that the vehicle is traveling in the lane on which the arrow P40c is drawn. In this frame, paint such as an arrow P40c and a lane boundary line P40b is shown, and a rear part P40a of a truck traveling in the right lane is shown as a foreign object.
FIG. 40B shows a connected image generated using a frame image in which a foreign object is reflected. The part in which the foreign object is reflected as shown in FIG. 40A is represented as a part that cannot be recognized as paint in the connected image, as shown in a region G40 in the figure. Hereinafter, the foreign matter reflected in the frame image or the foreign matter shown in the connected image is referred to as a dust image.
If an attempt is made to recognize a sign in a state where there is such a dust image, there is a risk that the part of the dust image may be misrecognized as some kind of sign or an error may occur due to unrecognition.
In order to suppress such an adverse effect, in the modified example, a dust image is recognized, and the paint is recognized by reflecting the result.

(2) Method of reflecting the recognition result of dust images:
FIG. 41 is a flowchart showing a method for reflecting the result of recognition of dust images. The result of recognizing the dust image can be reflected by a pre-processing performed prior to paint extraction or by a post-processing performed after paint extraction.
In the flowchart in the figure, the reflection method in the case of pre-processing on the left side and post-processing on the right side is shown.
In executing these processes, it is assumed that a dust image in a frame image or a connected image is recognized separately by dust recognition processing (step S410), and the result is stored in dust image data. The contents of the dust recognition process (step S410) will be described later.
Here, an example is shown in which the dust recognition process is performed independently, but it may be incorporated in the process of pre-processing or post-processing.

In the case of preprocessing, the CPU of the computer reads the frame image (step S300A), and deletes the dust image from the frame image using the dust image data (step S301). By this process, for example, in the image shown in FIG. 40A, the reflection P40a at the rear of the track is deleted.
The CPU extracts the paint according to the procedure described in the embodiment (see FIG. 21) using the frame image from which the dust image has been removed. That is, vertical arrangement processing is performed (step S302), cross-related paint is extracted (step S310), and various paints are extracted (step S350). Then, a relative coordinate conversion process (see FIG. 39) is performed (step S370).
According to the preprocessing method, the paint is extracted using the frame image from which the dust image is removed, so that erroneous recognition due to the dust image can be suppressed.

In the post-processing, the CPU reads the frame image (step S300A), and in accordance with the procedure described in the embodiment (FIG. 21), the vertical arrangement process (step S302), the crossing related paint extraction (step S310), and various paint extractions (step S350) and relative coordinate conversion processing (step S370) are performed. Since the processing so far is performed using a connected image including a dust image, misrecognition due to the dust image is also included.
Next, the CPU reads the dust image location using the dust image data (step S411). The presence position of the dust image is obtained by orthographically projecting the dust image recognized in the frame image (see FIG. 7) and performing alignment processing (see FIG. 8). . Since dust images are not used for paint extraction, vertical arrangement processing is unnecessary, and relative coordinate conversion processing for converting the result of vertical arrangement into a position in the absolute coordinate system is unnecessary.
When the dust image existence area is thus obtained, the CPU checks the positional relationship between the paint and the dust image existence area and deletes the paint overlapping the existence area (step S412). This is because it is determined that such a paint has erroneously recognized a dust image.
According to the post-processing method, since the part of the recognized paint that overlaps with the dust image is removed, it is possible to prevent the paint that erroneously recognized the dust image from remaining.

In FIG. 41, the pre-processing and the post-processing have been described as separate independent processes. The reflection of the dust recognition result may be performed by either the pre-processing or the post-processing, or the post-processing may be further applied after performing the pre-processing.
Hereinafter, the details of the dust image recognition process in step S410 will be described in detail. As a dust image recognition method, there are an optical flow method and a gradation difference method.

(3) Dust image recognition processing (1) to optical flow:
FIG. 42 is a flowchart of the dust image recognition process (1). An example of processing using optical flow is shown.
First, the CPU reads two pieces of frame data arranged at an interval of about 1 m from the photographing position. Frame images taken while traveling are obtained at distance intervals shorter than 1 m, but in order to reduce the processing load, frame images are thinned out at intervals of about 1 m and two consecutive images are read. did. The reason why it is set to about 1 m is that the frame images are not strictly taken at intervals of 1 m.
The CPU calculates the distance between the frames based on the position coordinates together with the reading of the frame image (step S422). The calculated distance is used for normalization described later.

Next, the CPU extracts left and right inspection areas from each frame image (step S423). The outline of the inspection area is shown in the figure. The inspection areas are the left and right areas A42L and A42R excluding the central part A42C in the rectangular area used for generating the road surface texture in the frame image FL42. The reason for limiting to the area used for generating the road surface texture is that the recognition of the paint is not affected even if a dust image exists in the other area. The reason why the central portion A42C is excluded is that there are no other stopped vehicles or other foreign objects in the lane in which the vehicle is traveling. This is because in a travel lane in which the host vehicle travels, it is possible to travel with a large inter-vehicle distance so that a vehicle traveling ahead does not enter the image creation range.
The inspection area is not obtained by analyzing the frame image, but may be a fixed area in the frame image. Since shooting is performed with the camera fixed to the vehicle, the portion of the frame image FL42 that can be used for texture is fixed, and the area A42C in front of the vehicle that does not require recognition of the dust image is also fixed. This is because that. The area A42C in front of the vehicle can be set arbitrarily, but can be set to be an area corresponding to the lane width in which the host vehicle travels, for example.

Next, the CPU generates an optical flow based on the images of the left and right inspection areas (step S424). The magnitude of the optical flow also depends on the interval between two frame images used for generation. The CPU normalizes the optical flow to a value that does not depend on the frame image interval. In this example, as a normalization, a method of converting to a value corresponding to a value obtained from a frame image having an interval of 1 m is adopted. Normalization can be performed by the following method, for example.
(Vx, Vy) = (OFx / DIST, OFy / DIST);
Vx, Vy ... x, y components of the normalized movement vector;
OFx, OFy ... optical flow x and y components;
DIST: interval between two frame images;
The CPU obtains the following three types of evaluation values for each normalized movement amount vector obtained by normalizing each optical flow in the inspection region (step S425).
| V |: the size of the normalized movement amount vector;
Vx: x component of the normalized movement amount vector;
Vy: y component of the normalized movement amount vector;

Then, a point where any evaluation value becomes abnormal, that is, a point exceeding a preset threshold range is extracted (step S426). For example, when the magnitude | V | of the normalized movement amount vector is smaller than a preset minimum value | V | min or larger than a maximum value | V | max, it is determined as abnormal. The Similarly, regarding the x component and the y component, if they become smaller than the minimum values Vxmin and Vymin or become larger than the maximum values Vxmax and Vymax, they are determined to be abnormal.
The threshold range may be set as an allowable range of the absolute value of the evaluation value. In this case, it is determined that there is an abnormality when the magnitude | V | of the normalized movement amount vector exceeds a preset threshold value | V | th. Similarly, regarding the x component and the y component, if the absolute values | Vx | and | Vy | exceed the respective threshold values | Vx | th and | Vy | th, they are determined to be abnormal.
A threshold value for determining whether or not the evaluation value is abnormal can be arbitrarily set. For example, it is possible to obtain a normalized movement amount vector of a screen on which no foreign object is shown and set based on a range of evaluation values obtained therefrom.
In addition, here, a uniform threshold value is used for all points in the inspection area, but different threshold values may be used for the right and left inspection areas. The threshold value may be changed for each portion obtained by further subdividing the inspection area.
In step S426, the point at which any one of the magnitude of the normalized movement amount vector, the x component, and the y component is determined to be abnormal is extracted, but the points where all the points are abnormal are extracted. Other evaluation methods can also be taken.

Next, the CPU generates a convex polygon including the point determined to be abnormal, and stores this as a dust image (step S427).
The CPU repeatedly executes the above process for all the frame data (step S428), and ends the dust image recognition process.

FIG. 43 shows an example of an optical flow. The portions surrounded by solid lines on both sides of the area A43C in front of the traveling lane of the host vehicle are inspection areas A43L and A43R. The optical flow V43 is shown in the form of dots in the drawing. In this example, the optical flow V43 is obtained over the entire image regardless of the inspection areas A43L and A43R. Since no foreign matter is present in the inspection areas A43L and A43R, the optical flow is a vector indicating the amount and direction of the road marking to move within the screen as the vehicle travels. A movement of about 1 m is only a vector having a magnitude represented in a dot shape as illustrated.
After normalizing this optical flow, if the magnitude of the normalized movement amount vector, the x component, and the y component distribution are obtained, the dust recognition process (FIG. 42) step S426 is performed based on the maximum value and the minimum value. It is possible to set a threshold value used in.

FIG. 44 shows an example of an optical flow when there is a traveling vehicle. An example in which the vehicle C44 overtakes the left lane of the own vehicle is shown. An optical flow V44 is shown for the inspection area on the left side. Since the feature point of the vehicle overtaking the own vehicle moves at a speed faster than the movement of the own vehicle, it moves forward with respect to the own vehicle, that is, in the upper right direction in the screen. Therefore, the optical flow V44 is a vector pointing in the upper right direction.
As a result, in the portion corresponding to the vehicle A44, the evaluation value (size, x component, y component) of the optical flow V44 exceeds a preset threshold value and is determined to be abnormal.
The dust image A44 is recognized by generating a convex polygon including a point determined to be abnormal. The upper and left sides of the vehicle C44 are not included in the dust image A44 because these portions are outside the inspection area. A curve L44 in the dust image A44 represents a classification according to which evaluation value has exceeded the threshold value among the magnitude of the normalized movement amount vector, the x component, and the y component. In the modified example, if any evaluation value exceeds the threshold value, all the images are determined as dust images, and therefore the curve L44 has no particular meaning.

FIG. 45 is an example of an optical flow when there is a stopped vehicle. The example in case the vehicle C45 has stopped on the left lane of the own vehicle was shown. An optical flow V45 is shown for the inspection area on the left side. The stopped vehicle C45 is not different from the road marking in that the moving speed is 0, but is different from the road marking, and is a three-dimensional vehicle. Therefore, an optical flow different from the road marking is obtained in the frame image. The overtaking vehicle (FIG. 44) has an optical flow in the upper right direction, whereas the stationary vehicle has an optical flow in the lower left direction on the right side of the vehicle, while Direction and other complex optical flows.
The result of normalizing these optical flows V45 and generating convex polygons including abnormal points exceeding the threshold range is a dust image A45. In this example, the entire vehicle C45 cannot be recognized as a dust image, but the entire vehicle C45 can also be recognized as a dust image by adjusting the threshold range.

(4) Dust image recognition processing (2)-gradation difference:
FIG. 46 is a flowchart of the dust image recognition process (2). An example of processing using gradation difference is shown.
First, the CPU extracts an overlapping region of consecutive texture images in the connected image (step S461). An example of processing is shown in the figure. As shown on the left side, attention is paid to adjacent textures T461 and T462 in the connected image. Both overlap in the hatched part. Therefore, as shown on the right side, the CPU cuts out a region A461 that overlaps the texture T462 from the texture T461. Similarly, a region A461 is cut out from the texture T462.

Next, the CPU applies a smoothing filter to blur the image of the overlapping area (step S462). Any smoothing filter can be applied.
The reason for smoothing is as follows. In this dust recognition process, as will be described later, it is determined whether or not the image is a dust image based on the gradation difference for each pixel in the overlapping region. This is because road markings should overlap each other in the overlapping area, and a gradation difference should occur only in the dust image portion. However, since various errors are actually included at the time of photographing or in the processing process, the position of the road marking does not always coincide completely in the overlapping region. If the gradation difference is calculated for each pixel in this state, the road marking portion may be erroneously recognized as a dust image. Therefore, in step S462, a smoothing filter is applied to the image of the overlapping area so that a large gradation difference does not occur even when the position of the road marking is shifted in the overlapping area.

Using the image thus smoothed, the CPU obtains the gradation difference between the overlapping regions, and extracts pixels whose gradation difference exceeds the threshold value (step S463).
An example of processing is shown in the figure. Assume that the color component of the pixel Pxa of the overlapping image A461 is expressed as (Ha, Sa, Va) in the H (hue) S (saturation) V (brightness) system. In the overlap image A462, the color component of the pixel Pxb corresponding to this is expressed as (Hb, Sb, Vb).
Then, a pixel whose gradation difference exceeds the threshold is extracted.
For the hue, pixels satisfying | Ha−Hb |> threshold THH are extracted.
For saturation, pixels satisfying | Sa−Sb |> threshold THS are extracted.
For brightness, pixels satisfying | Va−Vb |> threshold THV are extracted.
Here, a method of extracting pixels whose brightness difference exceeds a threshold value is adopted. This is because white or yellow is used for the road marking, so that the lightness gradation difference is most prominent between the road marking and other dust images. Of course, pixels whose tone difference regarding hue and saturation exceeds a threshold value may also be extracted. Further, it is possible to take a method of extracting only pixels that exceed a threshold value in all of lightness, saturation, and hue.

The CPU combines the extracted pixels to specify a dust image area (step S464). The bonding can be performed, for example, by the following method. First, adjacent portions of pixels extracted as dust images are combined. Next, each pixel is thickened by predetermined pixels vertically and horizontally. The fattening process is the same process as described in the side cut process example of FIG. As a result of the fattening process, if there is a combined pixel, that portion is also recognized as one polygon. By doing so, the CPU can generate a dust image area. A convex polygon may be formed by filling a concave portion of a polygon obtained by combining pixels.
Finally, the CPU deletes a region having a minute area that is equal to or smaller than the threshold (step S465). This is because a stopped vehicle or other foreign object is normally recognized as a large dust image, and even if a small area dust image is left, it does not have a significant adverse effect on the recognition of paint. However, this process can be omitted.

FIG. 47 is an explanatory diagram showing an example of dust image recognition. 47 (a) and 47 (b) are overlapping regions cut out from two textures. In both cases, dust images G472 and G473 are reflected on the right side. In FIG. 47A, another dust image G471 is shown.
47 (c) and 47 (d) show the results of applying the smoothing filter to FIGS. 47 (a) and 47 (b), respectively. It can be seen that the contours of road markings are blurred compared to the original data.
47 (e) and 47 (f) show the results of obtaining the gradation difference between FIGS. 47 (c) and 47 (d), respectively. The range where the lightness gradation difference exceeds the threshold value THV is dust images G474 to G477 indicated by solid lines. The dust images G474, G476, and G475, G477 are the same. Here, in order to understand the comparison with FIG. 47 (c) and FIG. 47 (d), they are merely superimposed on each other.
For example, when only FIG. 47 (f) is viewed, the portion of the dust image G476 appears to be an appropriate road marking. However, since there is the dust image G474 of FIG. 47E superimposed on this, the gradation difference is an abnormal value exceeding the threshold value.
On the other hand, the dust images G475 and G477 are reflections of foreign matters. In this example, the entire reflection of the foreign object cannot be recognized as a dust image, but the entire image can be recognized depending on the adjustment of the threshold value.

  FIG. 47G shows a recognition result for another texture. As a result of obtaining the gradation difference by the above method, the dust image G478 is recognized. However, since these dust images G478 are very small, the CPU excludes them from the dust images (see step S465 in FIG. 46).

(5) Effect:
According to the dust recognition process described above, a dust image in a frame image or texture can be recognized and extracted.
By reflecting this dust image extraction result as a pre-process of the paint recognition process, the paint can be recognized based on the connected image from which the dust image has been removed, and the recognition accuracy can be improved.
In addition, if the extraction result of the dust image is reflected in the post-processing of the paint recognition process, it is possible to determine whether or not it is a dust image based on the paint recognition result, and erroneous recognition of paint based on the dust image can be determined. Can be eliminated.

  In the modified example, two methods, the method using the optical flow and the method using the gradation difference, have been described as the dust image recognition method. These methods may be used alone or in combination. By using in combination, there is an advantage that the dust image recognition accuracy can be further improved.

As mentioned above, although the various Example of this invention was described, it cannot be overemphasized that this invention is not limited to these Examples, and can take a various structure in the range which does not deviate from the meaning.
For example, the connected image may be generated as a single image obtained by combining road textures. In this case, when synthesizing a plurality of paths, the connected image may be divided into a plurality of regions corresponding to the road texture, and then translated for each region.
In this embodiment, an example of using an image taken with a video camera mounted on a vehicle is shown. However, not only a vehicle but also various other moving bodies can be used, and a method of taking a picture while walking is adopted. Also good.

DESCRIPTION OF SYMBOLS 100 ... Road surface imaging | photography system 110 ... Position measurement part 110 ... Measurement data 112 ... Controller 114 ... GPS
114A ... Antenna 116 ... IMU
118 ... DMI
DESCRIPTION OF SYMBOLS 120 ... Video camera 130 ... Recording apparatus 140 ... Hard disk 142 ... Image data 144 ... Synchronization data 146 ... Measurement data 150 ... Base station data 200 ... Road marking map generator 201 ... Main control part 202 ... Command input part 203 ... Display control part 204 ... Data input unit 205 ... Track data calculation unit 206 ... Image conversion unit 207 ... 1-pass image synthesis unit 210a ... Track data 210b ... Road surface track data 210c ... Road surface texture 210d ... Linked image 210e ... Road image 210f ... Road image registration data 210g ... Registration data for locus 210 ... Processing data storage unit 220 ... Alignment processing unit 221 ... Transparent polygon setting unit

Claims (12)

A road marking map generation method for generating a road marking map including a marking given to a road surface by a computer,
As a process executed by the computer,
(A) a step of inputting image data of a continuous image obtained by photographing the road surface on which the sign is given while moving on a road to be photographed, and position coordinate data representing a photographing position of the image data;
(B) converting each frame image constituting the input image data to obtain an orthographic image in a state in which the road surface is viewed from directly above;
(C) Along a path along a linear distance axis representing a distance from a reference point on the path, which is a movement locus when the photographed road is moved in a predetermined direction on a two-dimensional plane. Generating a linear connected image representing a road surface of the path by arranging the orthographic image based on a distance from the reference point; and
(D) A road marking map generation method comprising: a sign recognition step for recognizing a type and position of a sign given to the road surface based on the connected image. The road marking map generation method according to claim 1, further comprising:
(F1) Based on the indication recognized in the step (d), corresponding points corresponding to each other in a region that is photographed in common in a connected image of two or more passes among a plurality of passes. Identifying process;
(F2) setting one of the plurality of paths as a reference path;
(F3) In the absolute coordinate system representing the position coordinate data, a parallel movement is performed for each region constituting a path other than the reference path based on a movement vector set so that the positions of the corresponding corresponding points coincide with each other. And a step of generating a composite image of the road surface across the plurality of paths. The road marking map generating method according to claim 1 or 2, further comprising:
(G) A perpendicular foot from the representative point of the marking to the distance axis is used as a reference point of the indication, and the reference point on the path shown in the absolute coordinate system based on the position coordinate data is perpendicular to the path. A road marking map generation method comprising a step of obtaining a position coordinate in an absolute coordinate system of the marking by drawing a perpendicular and arranging a representative point of the marking at an end point of the perpendicular. A road marking map generation method according to any one of claims 1 to 3,
The road marking map generation method wherein the step (d) determines the position and size of the geometric shape determined for each type of the marking so as to include the marking. A road marking map generation method according to any one of claims 1 to 4,
The step (d) is a road marking map generation method for recognizing the sign after deleting both end areas that are separated from the distance axis by a predetermined distance or more in advance for the connected image. A road marking map generation method according to any one of claims 1 to 5,
In the step (d), after extracting a portion of a predetermined color used for the indication from the connected image, the indication is performed by matching the indication pattern prepared in advance for each type of the indication and the extracted portion. Road marking map generation method for recognizing. A road marking map generation method according to any one of claims 1 to 5,
In the step (d), after extracting a portion of a predetermined color used for the marking from the connected image, the type of the marking is recognized based on the direction and size of the extracted portion with respect to the distance axis. To create a road marking map. The road marking map generation method according to any one of claims 1 to 7, further comprising:
(E) obtaining an optical flow representing movement of a plurality of feature points from two consecutive frame images;
(F) based on the optical flow, extracting a reflection other than the marking in the frame image, and in the step (b), together with the orthographic image or instead of the orthographic image , Generating an orthographic image reflecting the extraction result of the reflection in the frame image,
In the step (c) or the step (d), based on the orthographic image generated in the step (b), the generation of a connected image reflecting the extraction result of the reflection, or the indication applied to the road surface A road marking map generation method for recognizing the type and position of the road. The road marking map generation method according to any one of claims 1 to 8, further comprising:
(H) in a portion where the orthographic images arranged on the distance axis overlap with each other, comprising a step of extracting a reflection other than the marking based on a gradation difference of each point between the orthographic images,
In the step (c) or the step (d), a road surface that generates a connected image reflecting the extraction result of the reflection in the step (h) or recognizes the type and position of a sign drawn on the road surface. Marking map generation method. A road marking map generation method according to claim 9,
In the step (h), a road marking map generation method for obtaining the gradation difference after applying a smoothing filter to the orthogonal image. A road marking map generating device for generating a road marking map including a marking applied to a road surface,
An input unit for inputting image data of a continuous image obtained by photographing the road surface on which the sign is given while moving on a road to be photographed, and position coordinate data representing a photographing position of the image data;
An image converter that converts each frame image constituting the input image data and obtains an orthographic image in a state in which the road surface is viewed from directly above;
The reference point along the path on a linear distance axis that represents the distance from the reference point on the path, which is the movement trajectory when the photographed road is moved in a predetermined direction on a two-dimensional plane A connected image generation unit that generates a linear connected image representing the road surface of the path by arranging the orthographic image based on a distance from
A road marking map generating apparatus comprising: a sign recognition unit that recognizes a type and position of a sign drawn on the road surface based on the connected image. A computer program for generating a road marking map including a marking applied to a road surface,
An input subprogram for inputting image data of a continuous image obtained by photographing the road surface on which the sign is given while moving on a road to be imaged, and position coordinate data representing a photographing position of the image data;
An image conversion subprogram that converts each frame image constituting the input image data to obtain an orthographic image in a state in which the road surface is viewed from directly above;
The reference point along the path on a linear distance axis that represents the distance from the reference point on the path, which is the movement trajectory when the photographed road is moved in a predetermined direction on a two-dimensional plane A connected image generation subprogram for generating a linear connected image representing the road surface of the path by arranging the orthographic image based on the distance from
A computer program comprising a sign recognition subprogram for recognizing the type and position of a sign drawn on the road surface based on the connected image.