JP5256464B2 - Image top feature measurement method, display method, and measurement apparatus - Google Patents

Description

  The present invention relates to a technique for using an image obtained by photographing with a digital camera, and in particular, measures the width, height, and depth of a feature such as a building or a road captured in the image on the image. The present invention relates to a technique for obtaining feature data.

  In recent years, the development of data processing technology has activated the use of images by cameras in various fields. For example, Patent Documents 1 to 7 disclose such examples.

  The “map creation support system” of Patent Document 1 is a case where an image is used for map creation support, and an object is to create road data and a three-dimensional map by distinguishing a vehicle road from a promenade or the like. For this purpose, images are acquired at fixed distances while traveling on the road with a measurement vehicle equipped with position information acquisition means such as a camera and a GPS system, and these images are obtained in association with position information obtained by the position information acquisition means. Data is used to create map data.

  Patent Document 2 “Method for Creating 3D Map Database” uses an image for creating a 3D map database having height information of features. Therefore, in the “method for creating a three-dimensional map database” in Patent Document 2, a mobile measuring vehicle similar to the measuring vehicle in Patent Document 1 is used. That is, a set of corners for obtaining height information of each feature existing in a specific area on a two-dimensional map is determined in advance as a target end point group, and each of two or more measurement points is measured with a mobile measuring vehicle. The target endpoint is imaged, and the target endpoint that appears in the captured image is calculated based on the 3D position coordinates of each measurement point at the time of imaging, and the obtained 3D position coordinates are displayed on the 2D map database of the target endpoint A three-dimensional map database is created by adding to the two-dimensional map database in association with the two-dimensional position coordinates in FIG.

  The “three-dimensional measurement method by photogrammetry” of Patent Document 3 is a case where an image is used for surveying, and measurement points in photogrammetry can be associated with higher accuracy. Therefore, in the “three-dimensional measurement method by photogrammetry” of Patent Document 3, both a wide-angle image and a telephoto image are used so that the subject image of the telephoto image matches the subject image in the wide-angle image taken from the same shooting position. Create a composite photo image that is displayed by combining the images for each shooting position, specify a common reference point in the shooting area of the telephoto image of the composite photo image, and then create multiple composite photos of the reference points An imaging position relationship is calculated based on the coordinate value in the image, and three-dimensional position information in the subject space of the measurement point is measured from the coordinate value of an arbitrary measurement point in the photographic image based on the imaging position relationship. I have to.

  The “vanishing point determination method” of Patent Document 4 is necessary to create a three-dimensional map model by estimating the height of a building on an image, or to specify a path through which a robot can pass through an image. A method for determining the vanishing point on an image is proposed. In the vanishing point determination method, a line segment is extracted from the image data taken from the sidewalk, and then, a useful one for the vanishing point is selected from the line segments, and the area of the triangle formed by the vanishing point and both ends of the line segment is selected. The position of the vanishing point is specified using an evaluation value determined based on the above.

The “camera image-real space coordinate correspondence map generation method” of Patent Document 5 is a case where an image is used for monitoring, and a camera image-real space coordinate correspondence used for obtaining a position of a person or the like in the monitoring image in real time. It proposes a method for generating a map. In the camera image-real space coordinate association map generation method, a square pattern is arranged on the floor surface of the real space to be associated, and the real space coordinates of the position where the square pattern exists are determined. Then, the image coordinates of each vertex of the square pattern in the image obtained by photographing the real space including the square pattern are detected, and the positions of the vanishing points of the two straight lines passing through the vertex image coordinates are calculated. Based on the number of vanishing points, a corresponding map image in which real space coordinates are described for all pixels of the captured image is generated.
JP 2007-206099 A JP 2000-74669 A JP 2006-113001 A JP 2007-57386 A JP 2005-275500 A

  In addition to the above-described example, the camera image can still be used in many forms. For example, one example is the use of images taken of roads and features along the road as data for tax assessment of real estate along the roads. In other words, in the tax assessment of real estate, the situation of the surrounding features of the property to be assessed has a significant effect, so it is important to obtain highly persuasive materials about the situation of such features. In other words, a photographed image, in particular, an image along the road obtained by photographing in a state where the road extends in the optical axis direction of the camera is used.

  However, in order to use roadside images as materials for tax assessment of real estate, it is required to arrange a series of roadside images taken at regular intervals along the road, and the buildings included in the roadside images It is also required as an important factor that feature data including the width, height, depth, etc. of features such as roads and roads can be displayed superimposed on the image along the road.

  Among these requests, a request for a series of roadside images includes roads at regular intervals while running a measurement vehicle equipped with a camera and position information acquisition means as described in Patent Document 1 and Patent Document 2 above. You can respond by acquiring images along the way. However, the above-described conventional technology cannot sufficiently meet the demand for feature data.

  In order to display feature data on an image, it is necessary to measure the feature on the image. The measurement of the feature on the image can be performed by a method using a plurality of images having parallax, as in the examples disclosed in Patent Document 2 and Patent Document 3, for example. However, in the method using a plurality of images, it is necessary to process a plurality of images to measure a feature in one subject space, which increases the burden of data processing, and two locations for the same subject space. Since it is necessary to take an image from the above, the burden on the image taking increases, resulting in a problem of cost performance. Patent Document 4 mentions that the vanishing point on the image is effective for the measurement of the feature on the image, but does not disclose the specific means and responds to the request for the feature data. I can't.

  For example, the image above the road is used to measure the ground height of the cable (distribution cable and communication cable, which is also a kind of feature) including the electric wires along the road and across the road. The same applies to the case of image usage such as collecting cable ground height data. That is, the method of collecting cable ground height data using roadside images obtained while driving a vehicle can be expected to significantly reduce the work load compared to the method of actually measuring the cable ground height at the site. However, in order to make it practical, it is desirable to be able to easily perform measurement with high accuracy of the cable ground height on the image.

  The present invention has been made in the background of the circumstances as described above, and the problem thereof is an on-image feature measurement method that enables easy measurement of features on an image with high accuracy and It is in providing a display method and a measuring device.

  In the present invention, in order to solve the above-described problem, measurement is performed by obtaining the length of the measurement part from the number of pixels of the measurement part in the feature to be measured on the image including the feature, thereby obtaining a single image. Basically, it is possible to perform measurement (measurement in which a single object space is measured with one image), thereby easily performing feature measurement on the image. However, in this case, the length of one pixel is different depending on the depth position with respect to the camera in the subject space of the measurement site, that is, the distance from the camera in the subject space of the measurement site. For this reason, it is necessary to obtain the actual length per pixel for each measurement region. In this regard, in the present invention, by utilizing the fact that the positional relationship between the vanishing point on the image and the measurement part correlates with the manifestation length per pixel, the manifestation length per pixel in the measurement part is obtained. Yes. In this case, it is sufficient to obtain the vanishing point on the image by using the contour of the feature captured in the image, more specifically, the contour of the feature in the subject space and orthogonal to the image plane. For example, it can be obtained by performing a simple operation such as tracing the outline of a feature by using a drawing function on an image displayed on a display device. In addition, the processing for obtaining the actual length per pixel from the positional relationship between the vanishing point and the measurement site is relatively simple. Therefore, it is possible to easily measure the feature data with high accuracy of the feature on the image.

  In the present invention based on the above concept, in the on-image feature measurement method for measuring the feature captured in the image on the image by the camera, the contour of the feature captured in the image is used. While calculating | requiring a vanishing point, calculating | requiring the actual length per pixel in the desired measurement site | part of the target object of a measurement object based on the said vanishing point, and measuring feature data based on the actual length per pixel by it It is characterized by having made it.

  In the present invention, for the above-described image top feature measurement method, when the image is an image along the road, that is, an image taken by capturing the road while extending in the optical axis direction of the camera, the vanishing point is , Use the side line of the road. By doing in this way, the vanishing point which enables the highly accurate feature measurement in a roadside image can be calculated | required more easily.

  Therefore, according to the present invention, in the above image top feature measurement method, the image is an image along the road captured by capturing the road in a state of extending in the optical axis direction of the camera, and the disappearance The point is characterized in that it is obtained using a side line of the road.

  For the above-described image top feature measurement method, it is necessary to obtain the absolute value of the manifestation length per pixel. The absolute value of the manifestation length per pixel can also be obtained by optical analysis of the image. However, the image analysis method leaves a problem in accuracy. It is also possible to obtain the absolute value of the manifestation length per pixel by using a previously known length of the feature captured in the image. However, it is not always easy to capture a feature having a known length in an image. Therefore, it is appropriate to capture a linear body having a known length into an image and obtain the absolute value of the actual length per pixel using the linear body image.

  For this reason, in the present invention, in the above-described image top feature measurement method, a linear object having a known length is taken into the road-side image, whereby the road-side image is used as the image of the linear body. A teacher line whose absolute length is known is generated in the roadside image, and a measurement line is designated on the roadside image with respect to a desired measurement portion of the feature to be measured, The feature data is measured by comparing the measurement line with the teacher line with respect to the number of pixels under the relative positional relationship between the measurement line and the teacher line with respect to a vanishing point. .

  As described above, when using the teacher line, the teacher line is obtained for each coordinate axis of the three-dimensional coordinates in the subject space so that the width, height, and depth of the feature can be measured as necessary. Is preferred.

  Therefore, the present invention is characterized in that the teacher line is generated for each coordinate axis of the three-dimensional coordinates in the subject space of the image along the road in the above-described image top feature measuring method.

  When each of the width, height, and depth of the feature is to be measured as the feature data, the correlation of the positional relationship between the vanishing point and the measurement site with respect to the actual length per pixel differs in each measurement.

  Therefore, according to the present invention, in the above-described image top feature measurement method, the three-dimensional coordinates of the subject space have the X axis as the horizontal direction, the Y axis as the depth direction, and the Z axis as the vertical direction, and along the road. Assuming that the horizontal direction is the x-axis and the vertical direction is the y-axis for the two-dimensional coordinates on the image set in the image, the X-axis component of the measurement line is 1 at the position of the end point of the measurement line on the y-axis. The actual length per pixel is measured based on the teacher line in the X-axis direction, and the Z-axis component of the measurement line is measured per pixel at the position of the end point of the measurement line on the y-axis. Is measured based on the teacher line in the Z-axis direction, the Y-axis component of the measurement line is arbitrarily projected by projection on the image along the road from the virtually projected center, Y-axis formation of the teacher line in the Y-axis direction And the measurement line projection image obtained by projecting the y-axis components of the measurement line onto the line segment parallel to the x-axis in the two-dimensional coordinates on the image with respect to the number of pixels. It is characterized by measuring feature data.

  In measurement using the fact that the positional relationship between the vanishing point and the measurement site correlates with the actual length per pixel, the accuracy of the actual length per pixel differs depending on the distance from the camera, and is too close to the camera. If it is too far from the camera, the accuracy may be insufficient.

  For this reason, the present invention is characterized in that in the above-described image top feature measurement method, a measurement region is set in the roadside image, and measurement can be performed only within the measurement region.

  As a field of use of the feature data obtained by the above-described image top feature measurement method, it is one of the most effective to display and use it on the roadside image as the material for the tax assessment of the real estate as described above. . In this case, it is necessary to arrange a series of road-side images taken at regular intervals along the road as described above.

  Therefore, in the present invention, in the image top feature measurement method as described above, the image along the road is an image obtained by shooting while driving a vehicle on which a camera is mounted. The feature data obtained by the above-described feature measurement method on the image is displayed on the image along the road to form image data.

  On the image according to any one of claims 1 to 8, wherein the feature data includes at least one of a length, height, depth, inclination, distance, and area of the feature. Disclosed is a feature measurement method.

Furthermore, the present invention provides a display means for displaying on a display screen including features captured by a camera,
A first processing means for fetching contour lines of features designated on the display image as data, and obtaining vanishing points from the contour line data;
Second processing means for obtaining an actual length per pixel at a desired measurement site of the measurement target feature on the display image based on the vanishing point data;
Third processing means for obtaining feature data including at least one of the length, height, depth, inclination, distance, and area of the measurement site based on the manifestation length;
An on-image feature measuring apparatus comprising:

  According to the present invention as described above, it becomes possible to easily measure the feature on the image with high accuracy. For example, an image along the road that can be used as a material for taxation evaluation of real estate can be obtained at low cost. It is also possible to obtain cable ground height data using images.

  Hereinafter, modes for carrying out the present invention will be described. FIG. 1 shows a configuration of a feature data acquisition system 1 used for carrying out an on-image feature measurement method according to an embodiment. The feature data acquisition system 1 according to the present embodiment acquires a roadside image 4 as an image while driving the photographing vehicle 3 on the road 2, and performs measurement on the feature in the roadside image 4 to obtain feature data. This is a case, and consists of a photographing vehicle 3 and an image top feature measurement system 5.

  The photographing vehicle 3 is configured by mounting a digital camera 6 and a photographing management system 7 on a vehicle that can travel even in an urban area where a narrow road with a width of 4 m or less is complicated, for example.

  The camera 6 is a still camera or a video camera. Usually, a still camera is used, and two cameras are mounted as a front shooting camera 6f and a rear shooting camera 6r. The front photographing camera 6f performs photographing in the state where the optical axis is along the traveling direction in front of the photographing vehicle 3, and the rear photographing camera 6r has the optical axis in the traveling direction behind the photographing vehicle 3. Take a picture along the line. The roadside image 4 obtained by photographing with the front photographing camera 6f and the rear photographing camera 6r is an image in which the road is captured while extending in the optical axis direction. In this way, the front and rear of the traveling direction can be photographed because the front photographing camera 6 f may be backlit depending on the traveling direction of the photographing vehicle 3, so that such a case can be dealt with. It is to make it.

  The imaging management system 7 includes each subsystem of a positioning system 8, an imaging control system 9, and an image data management system 10. The positioning system 8 functions to acquire position and orientation information, which is information about the current position and traveling direction of the photographing vehicle 3, more specifically, the optical axis direction of the camera 6, at intervals of, for example, every traveling distance 1m. . Therefore, the positioning system 8 is configured by combining a GPS (Global Positioning System) and a distance / inertia measurement system. The GPS acquires position / orientation information by receiving signals from a satellite with two GPS antennas 11 provided before and after the photographing vehicle 3. On the other hand, the distance / inertia measurement system includes an inertial measuring device that detects a traveling direction of the photographing vehicle 3 and an inclination of the vehicle with respect to a horizontal plane by a gyro and a distance measuring device that detects a traveling distance of the photographing vehicle 3, and thereby includes position direction information. To get. As described above, by configuring the positioning system 8 by combining the GPS and the distance / inertia measurement system, it is possible to stably acquire the position and orientation information even in an environment where signals from GPS satellites cannot be received.

  The imaging control system 9 controls imaging by the camera 6 as control that can acquire the roadside images 4 at regular distance intervals. Specifically, the traveling distance of the photographing vehicle 3 is counted, and a photographing command is output for every certain traveling distance such as 5 m to cause the camera 6 to perform photographing. By such photographing, a series of roadside images can be acquired at a constant distance along the road on which the photographing vehicle 3 travels.

  The image data management system 10 is configured by using a workstation-level data processing device, and manages data of the roadside image 4 acquired by photographing by the camera 6. The management is performed by associating the roadside image 4 with the position and orientation information at the time of photographing.

  Here, in the roadside image 4, a teacher line LT necessary for image upper feature measurement processing as described later is generated. The teacher line LT is an image of the roadside image 4 with a direction along the side line of the road 2 in the roadside image 4 (this is used as an outline of a feature for obtaining a vanishing point as described later) as one coordinate axis. Each coordinate axis of three-dimensional coordinates (X, Y, Z coordinates) in the subject space is generated. That is, the teacher line LTx in the X-axis direction, the teacher line LTy in the Y-axis direction, and the teacher line LTz in the Z-axis direction are generated. These teacher lines LTx, LTy, and LTz are generated as images of the linear bodies by taking in linear bodies with known lengths (for example, surveying poles). In principle, it is sufficient to generate such a teacher line LT only for a roadside image that is first captured under the shooting conditions for a range in which a roadside image is shot under the same shooting conditions. However, practically, it is preferable to generate a teacher line by imprinting a linear body for a teacher line for each road-side image taken at the beginning of a day's shooting work, for example. This is because the setting state of the camera 6 on the photographing vehicle 3 affects the photographing conditions, and the setting state may be slightly different for each day of photographing work. If the road is inclined, the camera is also inclined with respect to the horizontal plane. However, the inclination of the camera is detected using the position and orientation information, and correction according to the inclination is performed in the image top feature measurement processing. Will be made.

  The image top feature measurement system 5 is configured by mounting the image top feature measurement means 13 configured as a computer program on the data processing device 12. The on-image feature measuring means 13 manages the roadside image with feature data as described later in which the data of the roadside image 4 captured from the image data management system 10 or the feature data is recorded in the roadside image 4. The data management unit 14, the measurement unit 15 that measures the features captured in the roadside image 4 and acquires the feature data, and records the feature data acquired by the measurement unit 15 in the roadside image 4 A feature data recording unit 16 for generating an image along the road with object data is included.

  Hereinafter, an example of the image upper feature measurement process performed by the image upper feature measurement system 5 will be described. The image top feature measurement processing in this example is a case where the vanishing point is obtained by using the side line of the road 2 in the road-side image 4 as the contour of the feature. As shown in FIG. Each process of S8 is included.

  First, in step S1, a processing target image is displayed. The processing target image display process is a process of displaying the roadside image to be measured on the display screen of the data processing device 12, and is performed in response to designation by the user, or a series of roadside images are sequentially displayed. Is done.

  In step S2, a road side line is designated. Specifically, the user designates a road side line for the roadside image displayed in step S1. The designation is performed as shown in FIG. 3, for example. Specifically, the two designated lines LI and LI are drawn by tracing the side line of the road 2 in the roadside image 4 displayed on the display screen with the drawing function for each of the corresponding straight line portions on the left and right. To do.

  In step S3, vanishing point determination processing is performed. When the vanishing point is obtained using the side line of the road 2, the vanishing point is the intersection V of the two designated lines LI and LI. Therefore, the vanishing point is specified by obtaining the position of the intersection V on the road-side image 4. Specifically, the two-dimensional coordinates on the image with the horizontal direction of the road-side image 4 (corresponding to the horizontal direction in the subject space of the road-side image 4) as the x-axis and the vertical direction as the y-axis are set in the road-side image 4. Then, the position of the intersection V on the road-side image 4 is obtained as the coordinate value (xv, yv) in the two-dimensional coordinates on the image, and this is used as the vanishing point. In FIG. 3, the two-dimensional coordinates on the image are set so that the x-axis is parallel to the teacher line LTx in the X-axis direction.

  In step S4, measurement line designation processing for the measurement region is performed. This process is performed by drawing the measurement line S on the measurement site by the drawing function as in the example shown in FIG. 3, or by specifying the end points A and B at both ends of the measurement site.

  In step S5, measurement processing of the measurement line specified in step S4 is performed. This measurement process is performed on the premise of three-dimensional coordinates in the subject space of the roadside image as in the example shown in FIG. Here, regarding the three-dimensional coordinates, the horizontal direction is the X axis, the depth direction is the Y axis, and the vertical direction is the Z axis. If such three-dimensional coordinates are assumed, there are two measurement lines depending on the orientation that the measurement line takes on the image along the road and the case where there is only one component in each of the X, Y, and Z axial directions. There may be three combinations. In the case of a combination, the length of the measurement line is obtained from each component by the three-square theorem.

  FIG. 5 shows the calculation principle of the X-axis component CX of the measurement line S including the X-axis component and the Y-axis component. The coordinate values of the end points of the teacher line LTx in the X-axis direction at the two-dimensional coordinates on the image are (xta, yta) and (xtb, ytb), and the two-dimensional coordinates on the images of the end points A and B of the measurement line S are used. Is set to (xa, ya) (xb, yb). Note that the measurement line S in FIG. 5 does not necessarily correspond to the measurement line S in FIG.

  Assuming that the distances from the vanishing point perpendicular lines of the end points A and B of the measurement line S are Sa and Sb, these Sa and Sb can be obtained based on the actual length TL of the teacher line LTx. The X-axis component CX can be obtained as CX = Sa−Sb. Here, the vanishing point perpendicular line is a perpendicular line from the vanishing point V, and in the case of FIG. 5, the x axis is parallel to the teacher line LTx, and thus is a line segment orthogonal to the x axis.

In order to obtain Sa from the actual length TL of the teacher line LTx, the teacher line LTx-a virtually assumed at the position of the coordinate value ya is used. Specifically, when the number of pixels of the teacher line LTx is N, the number of pixels of the teacher line LTx-a is Na, and the number of pixels of Sa is na, Sa is Sa = [(TL / N) · ( TL / Na)] · na. Similarly, Sb can be obtained. The above is the calculation principle of CX. Actually, CX is obtained by the following equation.
CX = | TL * {(xa-xv) * (yb-ya) / (ya-yv) + xa-xb} * (2 * yv-yta-ytb) / {2 * (xta-xtb) * (yv- yb)} |

FIG. 6 shows the calculation principle of the Y-axis component CY of the measurement line S. CY is obtained by projection projection. That is, the y-axis component LTy-v of the teacher line LTy in the Y-axis direction and the y-axis component sv of the measurement line S are parallel to the x-axis by projection on the image along the road from the arbitrarily projected projection center P. Projecting onto a straight line segment, a teacher line projection image LTy-h and a measurement line projection image vs-h are obtained. In this case, the teacher line projection image LTy-h and the measurement line projection image vs-h have the same actual length per pixel. Therefore, the length of CY can be obtained by simply comparing the teacher line projection image LTy-h and the measurement line projection image vs-h with respect to the number of pixels. Specifically, the actual length of the teacher line LTy can be determined as TL, and the length of CY can be determined as CY = TL · vs−h / LTy−h. The above is the calculation principle of CY. Actually, CY is obtained by the following equation.
CY = TL · | {(yta-yv) · (ytb-yv) · (yb-ya)} / {(ya-yv) · (yb-yv) · (ytb-yta)} |
If CX and CY are obtained as described above, the actual length SL of the measurement line S can be obtained as SL = (CX ² + CY ² ) ^1/2 .

  FIG. 7 shows the calculation principle of the Z-axis component CZ of the measurement line S. In FIG. 7, the measurement line S is parallel to the teacher line LTz in the Z-axis direction. Therefore, the measurement line S consists only of the X-axis component, and CZ = S. The length of CZ can be obtained based on the teacher line LTz on the same principle as the calculation of CX described above. Specifically, the hypothetical teacher line LTz-s is used at the position of the coordinate value yb of the lower end point of the measurement line S, the number of pixels of the teacher line LTz is N, and the number of pixels of the teacher line LTz-s is Na. If the number of pixels of CZ (= S) is na, CZ can be obtained as CZ = [(TL / N) · (TL / Na)] · na.

The above is the calculation principle of CZ. Actually, CZ is obtained by the following equation. However, “ChiHei_Y” in the equation is a coordinate value of the y axis in the two-dimensional coordinates on the image of the ground plane including the end point B of the measurement line S. This “ChiHei_Y” is the same as the y-axis coordinate value of the end point B when the measurement line S is on the ground plane. However, the end point B of the measurement line S is not necessarily on the ground plane.
CZ = | TL · {(yb-ya) · (yv-yta) / {(yta-ytb) · (yv-ChiHe_Y)} |

  In the measurement using the positional relationship between the vanishing point and the measurement part as described above correlates with the actual length per pixel, the accuracy of the actual length per pixel differs depending on the distance from the camera. If it is too close to or too far from the camera, the accuracy may be insufficient. Therefore, as shown in FIG. 8, it is preferable to appropriately set the measurement region 17 in the roadside image 4 so that the measurement can be performed only in the measurement region 17. In this case, it is appropriate that the teacher line LTx and the teacher line LTz are the outer edges of the measurement region 17. Therefore, the image for generating the teacher line LT is preferably a measurement target image captured under the condition that the teacher line LTx and the teacher line LTz are the outer edges of the measurement region 17.

  Returning to FIG. 2, in step S <b> 6, it is determined whether measurement has been completed for all necessary measurement parts. If this determination is negative, the process returns to step S4 to repeat the process. If the determination is affirmative, the process proceeds to step S7.

  In step S7, the feature data is recorded on the roadside image that has been measured, and image data with feature data is generated. FIG. 9 shows an example of image data with feature data. The image data with feature data 18 in the example of FIG. 9A is a case where the width data of the road 2 and the height data of the building 19 along the road are recorded, and the example of FIG. 9B. The feature data-added image data 20 is a case where the ground height data of the road crossing cable 21 such as an electric wire is recorded.

  In step S8, it is determined whether measurement has been completed for all necessary images. If this determination is negative, the process returns to step S1 to repeat the process. If the determination is affirmative, the process ends.

  Moreover, the above embodiment is a case where a vanishing point is calculated | required using the side line of a road. This is because, for example, when using an image with feature data for collecting tax assessment data for real estate or collecting data for cable ground height data, the image is taken by capturing the road while extending in the optical axis direction of the camera. It is normal for the image to be along the road, and in such an image along the road, the road side line is usually perpendicular to the image plane. This is because it is reasonable to ask for points. In other words, the method of calculating the vanishing point with the side line of the road is particularly effective for images along the road, and if necessary, outlines of features other than the road (for example, the boundary line between the structure along the road and the road, etc.) ), The vanishing point can be obtained using the contour of the feature in the subject space and orthogonal to the image plane.

  Moreover, although the example of the width | variety and height was shown as an example of measurement feature data, the distance, inclination, area, etc. from a certain reference point regarding a feature can also be included. These can be positioned as so-called geometric extraction data of image features in a broad sense, and can be added as an application example of the present invention.

  As mentioned above, although the form for implementing this invention was demonstrated, this is only a representative example and this invention can be implemented with various forms in the range which does not deviate from the meaning. For example, the above embodiment is a case of an image along a road acquired while the photographing vehicle is running, but the present invention is not limited to this, and can be applied to an image by a camera in general.

It is a figure which shows the structure of the feature data acquisition system used for implementation of the on-image feature measuring method by one Embodiment. It is a figure which shows the flow of an image top feature measurement process. It is a figure explaining designation | designated of a road side line and a measurement line. It is a figure which shows the example of the three-dimensional coordinate in to-be-photographed space. It is a figure explaining the principle about calculation of the X-axis component of a measurement line. It is a figure explaining the principle about calculation of the Y-axis component of a measurement line. It is a figure explaining the principle about calculation of the Z-axis component of a measurement line. It is a figure which shows the example of a measurement area | region. It is a figure which shows the example of image data with feature data.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Feature data acquisition system 2 Road 3 Photographing vehicle 4 Image along road 5 Feature measurement system on image 6 Camera 7 Imaging management system 8 Positioning system 9 Imaging control system 10 Image data management system 11 GPS antenna 12 Data processing device 13 On image Feature measurement means 14 Image data management unit 15 Measurement unit 16 Feature data recording unit 17 Measurement area 18 Image data with feature data 19 Building 20 Image data with feature data 21 Road crossing cable LT Teacher line S Measurement line V Loss point

Claims (8)

In the image top feature measurement method for measuring a feature included in the image on the image including the feature by the camera,
The image is an image along a road captured by capturing a road in a state extending in the optical axis direction of the camera, and a vanishing point is obtained using a side line of the road included in the image, and the disappearance In measuring the feature data based on the actual length per pixel based on the actual length per pixel in the desired measurement region of the target feature to be measured based on the points,
By incorporating a linear body having a known length into the roadside image, a teacher line whose absolute length on the roadside image is known is included in the roadside image as the linear body image. And generating a measurement line on the image along the road with respect to a desired measurement part of the feature to be measured, and under the relative positional relationship between the measurement line and the teacher line with respect to the vanishing point The on- image feature measuring method, wherein the feature data is measured by comparing the measurement line with the teacher line for each number of pixels. 2. The image top feature measurement method according to claim 1 , wherein the teacher line is generated for each coordinate axis of a three-dimensional coordinate in the subject space of the image along the road. For the three-dimensional coordinates of the subject space, the horizontal direction is the X axis, the depth direction is the Y axis, and the vertical direction is the Z axis. As the y-axis,
For the X-axis component of the measurement line, the actual length per pixel at the position of the end point of the measurement line on the y-axis is obtained based on the teacher line in the X-axis direction, and measurement is performed.
For the Z-axis component of the measurement line, measure the actual length per pixel at the position of the end point of the measurement line on the y-axis based on the teacher line in the Z-axis direction,
With respect to the Y-axis component of the measurement line, the y-axis component of the teacher line in the Y-axis direction and the y-axis component of the measurement line are respectively projected by projection on the image along the road from an arbitrary virtual projection center. The feature data is measured by comparing the teacher line projection image obtained by projecting onto the line segment parallel to the x-axis in the two-dimensional coordinates on the image and the measurement line projection image with respect to the number of pixels. The image top feature measurement method according to claim 2 , wherein: The image upper ground according to any one of claims 1 to 3 , wherein a measurement area is set in the roadside image, and the feature data can be measured only within the measurement area. Object measurement method. The roadside image, the image upper base material measuring method according to any one of claims 1 to 4, wherein the camera is an image obtained by photographing while traveling vehicle equipped . It claims 1 to any display method comprising displaying on the image containing the feature of the feature data obtained by the image upper base material measuring method according to one of claims 5. The feature data, the length of the feature, height, depth, inclination, distance, image upper base material measuring method according to any one of claims 1 to 6 is intended to include at least one of the area . Display means for displaying on a display screen including features along the road captured by capturing the road in a state extending in the optical axis direction of the camera;
A first processing means for fetching a side line of a road designated on the display image as data and obtaining a vanishing point from the side line data;
Second processing means for obtaining an actual length per pixel at a desired measurement site of the measurement target feature on the display image based on the vanishing point data;
A third processing means for obtaining feature data including at least one of the length, height, depth, inclination, distance, and area of the measurement site based on the manifestation length ;
The third processing means includes
By incorporating a linear body having a known length into the roadside image, a teacher line whose absolute length on the roadside image is known is included in the roadside image as the linear body image. And generating a measurement line on the image along the road with respect to a desired measurement part of the feature to be measured, and under the relative positional relationship between the measurement line and the teacher line with respect to the vanishing point An on- image feature measuring apparatus for obtaining feature data by comparing a measurement line with the teacher line for each number of pixels .

Images