JP2001108434A - Method and apparatus for measuring distance - Google Patents

Abstract

PROBLEM TO BE SOLVED: To realize a highly accurate distance measurement based on the image from a single visible camera even when the camera itself is moving or the road surface is not horizontal. SOLUTION: An object detecting means 12 detects the position based on image data 15 received from an image input section 11. A road structure recognizing means 13 recognizes the road structure based on the image data and a distance calculating means 14 calculates the distance accurately by integrating the position of the object and the road structure.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

[0001] 1. Field of the Invention [0002] The present invention relates to a distance measuring technique for calculating a distance to an object existing in an image.

[0002]

2. Description of the Related Art Conventionally, techniques for measuring a distance include a technique using a laser radar and a technique using a stereo method using two visible cameras.

As a system using only one visible camera, there is a system disclosed in Japanese Patent Application Laid-Open No. 11-44533, entitled "Advance Vehicle Detector".

FIG. 10 is an explanatory diagram of a conventional system. FIG.
In the figure, reference numeral 91 denotes an imaging surface, and 92 denotes a lens. The prime number of the y coordinate from the image center of the preceding vehicle estimation point is dy2, the focal length of the lens 92 of the CCD camera of the image input unit is f,
When the number of pixels of the imaging surface 91 is IY and the screen size is DH, the angle dθ2 between the optical axis and a straight line connecting the estimated vehicle ahead point on the imaging surface 91 and the center point of the lens 92 is represented by the following equation. .

Dθ2 = atan (DH / IY · dy2 / f) At this time, the distance in the real space to the preceding vehicle estimation point is represented by L2.
Assuming that the vertical angle between the optical axis direction of the lens 92 and the horizontal line is θ1 and the mounting height of the CCD camera is H, the distance L2 is expressed by the following equation.

L2 = H / tan (θ1-dθ2) The distance L2 obtained by the above equation is regarded as the inter-vehicle distance.

[0007]

In such prior art, a laser radar and two visible cameras were required to measure the distance. Further, measuring the distance from one visible camera is limited to a case where the camera is fixed, or a case where the road from the camera to the target object is a horizontal plane even when the camera is not fixed.

According to the present invention, when the camera itself is moving,
Even when the road surface is not horizontal, the visible camera 1
An object of the present invention is to accurately measure a distance using only a table.

[0009]

SUMMARY OF THE INVENTION In order to solve this problem, the present invention relates to an object detecting means for detecting an object existing on a road from an image input from a camera, and at the same time, recognizing a structure of the road surface. Road structure recognizing means, and distance calculating means for calculating the distance in real space from the camera to the object using the position of the detected object in the image and the road surface structure, and configured to measure the distance Things.

Thus, a distance measuring device capable of accurately calculating a distance from an image of only one visible camera can be obtained even when the camera itself is moving and the road surface is not horizontal.

[0011]

According to the first aspect of the present invention, a position in an image of an object existing on a road is detected from an image input from a camera, and the structure of the road surface is recognized from the image. A distance measuring method comprising calculating a real space distance from a camera to an object from the detected position of the object and the road surface structure, wherein the camera itself is moving, and the road surface is Even when not horizontal,
This has the effect that the distance can be measured from an image of only one visible camera.

According to a second aspect of the present invention, in the distance measuring method according to the first aspect, the position of the object in the image is detected by:
The characteristics of multiple detection target objects are stored in advance,
The feature amount is obtained from the input image, the object most similar to the feature of the detection target object is extracted, and the position of the object is detected.When the camera itself is moving, and when the road surface is not horizontal, Also has the effect that the distance can be measured from an image of only one visible camera.

According to a third aspect of the present invention, in the distance measuring method according to the second aspect, the feature of the object is obtained by extracting linear components from a differential binary image of the object, and determining the lengths of the linear components and the object. It is expressed by at least one of the gender and the positional relationship, and has an effect that the distance can be measured from the image of only one visible camera even when the camera itself is moving and the road surface is not horizontal.

According to a fourth aspect of the present invention, in the distance measuring method according to the first aspect, the position of the object in the image is detected by:
In order to detect the position of the object by registering multiple types of learning images in advance and comparing the input image with each learning image to determine the most similar area, if the camera itself is moving, In addition, even when the road surface is not horizontal, the distance can be measured from the image of only one visible camera.

According to a fifth aspect of the present invention, in the distance measuring method according to the fourth aspect, a learning image is provided for each object to be detected at each of a plurality of different distances when the object exists at a plurality of different distances. And has the effect that the distance can be measured from the image of only one visible camera even when the camera itself is moving and the road surface is not horizontal.

According to a sixth aspect of the present invention, in the distance measuring method according to the first aspect, the road structure is recognized by assuming a road model in advance and detecting the positions of left and right road edges from an input image. Recognizes the structure of the road surface by obtaining the positions of the left and right road edges in the real space based on the model. Even when the camera itself is moving and the road surface is not horizontal, only one visible camera is used. Has the effect that the distance can be measured from the image.

According to a seventh aspect of the present invention, in the distance measuring method according to the sixth aspect, the distance between the left and right road edges is a fixed length, and a line segment connecting the left and right road edges is always horizontal. It is configured based on the assumption that there is an object, and has an effect that the distance can be measured from an image of only one visible camera even when the camera itself is moving and the road surface is not horizontal.

According to an eighth aspect of the present invention, in the distance measuring method according to the sixth aspect, the road edge is detected from a white line normally provided at the road edge, and a position where the white line is a broken line is detected. Detects by complementing between already detected white lines, and can measure the distance from the image of only one visible camera even when the camera itself is moving and the road surface is not horizontal. Having.

According to a ninth aspect of the present invention, in the distance measuring method according to the first aspect, the distance is calculated by finding left and right road edges corresponding to the position in the image of the detected object, and calculating the road edge from this position. The distance is calculated by detecting the position in the real space based on the recognition of the structure, and even when the camera itself is moving and the road surface is not horizontal,
This has the effect that the distance can be measured from an image of only one visible camera.

According to a tenth aspect of the present invention, in the distance measuring method according to the ninth aspect, the left and right road edges corresponding to the position in the image of the object are line segments connecting the left and right road edges and passing through the position. Is determined as a combination that minimizes the distance, and has an effect that the distance can be measured from an image of only one visible camera even when the camera itself is moving and the road surface is not horizontal.

According to an eleventh aspect of the present invention, there is provided an object detecting means for detecting a position in an image of an object existing on a road from an image input from a camera, and a road structure for recognizing a structure of a road surface from the image. Recognition means, and a distance calculation means for calculating a real space distance from the camera to the object from the detected position of the object and the road surface structure, when the camera itself is moving, and Has an effect that the distance can be measured from an image of only one visible camera even when is not horizontal.

According to a twelfth aspect of the present invention, in the distance measuring device according to the eleventh aspect, the road structure recognizing means is arranged such that a camera is installed behind the traveling vehicle to continuously photograph a direction opposite to the traveling direction. In this case, it is assumed that the road is horizontal at a position sufficiently close to the camera, and based on this assumption, the position of the road end in real space is calculated from the position of the road end in a range sufficiently close to the camera in the input image at the current time. At the same time, the movement amount of the vehicle itself in the real space from the previous time to the current time is obtained from the vehicle, and the road surface structure is recognized by connecting to the position of the road edge obtained by the previous time based on the movement amount. This has the effect that the distance can be measured from an image of only one visible camera even when the camera itself is moving and the road surface is not horizontal.

Hereinafter, an embodiment of the present invention will be described with reference to FIG.
This will be described with reference to FIG.

(Embodiment 1) FIG. 1 is a block diagram of a distance measuring apparatus according to Embodiment 1 of the present invention. In FIG. 1, reference numeral 11 denotes an image input unit for inputting an image of an object, and 12 denotes an area of the object which is extracted from image data 15 input by the image input unit 11 to obtain position information 1.
The object detection means 13 for detecting the image data 6 is used to output road structure information 17 to the image data 15 input by the image input unit 11.
Is a distance calculating means for calculating the distance from the camera to the object using the position information 16 and the road structure information 17 and outputting a measurement result 18.

The operation of the distance measuring device configured as described above will be described below. Here, FIG. 2 shows an example of the image data 15, where the detection target object 21 is on the road 23, and the detection target object 21 and the road 23 are in contact at the position 22. White lines 24 and 25 are drawn on the road 23. FIG. 3 shows an example of a differential binary image, and 31 indicates a detected model of the preceding vehicle.
FIG. 4 shows an example of a road structure recognition result, where 41 is a road shape on a horizontal plane when viewed from above, 42 is a road shape on a vertical cross section when viewed from the side, and both ends of a line segment 43. Have a one-to-one correspondence. Also, 44
Represents a position in the real space. FIG. 5 shows the road edges detected from the image data 15 and their correspondence.
52 indicates a detected road edge, and 53 indicates a line segment representing the correspondence between the left and right road edges.

The image input unit 11 inputs an image obtained by photographing a road to be recognized and an object to be detected on the road with one camera. The present invention can cope with a case where the camera itself is moving by mounting the camera on a car or the like, and a case where the road surface is not horizontal due to a slope as shown in FIG.

The object detecting means 12 is provided with the image input unit 11
2 is read, the position 22 in the image of the detection target object 21 is detected, and this is stored as position information 16.

Here, as an example of a method of detecting the position of an object from an image, an information processing report CV37-4 (1985)
Years) There are those shown in "Identification and tracking of preceding vehicles on expressways". The above method assumes a vehicle traveling ahead in particular as a detection target object, and its content will be briefly described below. First, the input image obtained by photographing the rear surface of the preceding vehicle is subjected to a secondary differentiation process and a binarization process to obtain a differential binary image as shown in FIG. Next, a target horizontal edge component is extracted from the obtained image, and this set is used as a model 31 of the preceding vehicle.

The road structure recognizing means 13 outputs the image data 15 as shown in FIG.
To recognize the structure of the road 23 in the real space,
This is stored as road structure information 17. Here, as an example of a method of recognizing the structure of the road surface in the real space from the image, there is a method shown in Jikken Kenkyu CV62-3 (1989) "Road shape restoration by local plane approximation" and the like. . The above-mentioned method obtains a road structure as shown in FIG. 4 from a road edge in an image based on knowledge about a road shape called a road model, and its content will be briefly described below. The road model is defined as a road edge when a line segment having a constant length on a smooth curve moves horizontally while intersecting vertically at a middle point.

First, as shown in FIG. 5, road edges 51 and 52 are detected with respect to an input image, and a one-to-one correspondence between left and right road edges is set so as to correspond to both ends of the line segment having a constant length. Ask for. Next, for the corresponding points of the left and right road edges, the three-dimensional position is restored based on the road model, and a smooth curve is applied to obtain the three-dimensional road shapes 41 and 4.
2.

The distance calculating means 14 comprises the object detecting means 1
The distance from the camera is calculated by calculating the position of the detection target object 21 in the real space from the position information 16 detected by Step 2 and the road structure information 17 recognized by the road structure recognition means 13, and the distance is measured. Output as result 18. For example, a target object 21 is detected from the image data as shown in FIG. 2 by the object detection means 12, and a position 22 in the image is obtained.

On the other hand, the road structure as shown in FIG. At this time, as shown in FIG. 5, the white lines in the image are detected as road edges 51 and 52, and the correspondence between the left and right road edges is determined, so the line segment 53 in the image corresponding to the detected position 22 is determined. .
The line segment 53 is converted into the line segment 43 by the road structure recognizing means 13.
Therefore, the position 44 in the real space is obtained from the position 22 in the image corresponding to the position. From this position 44, the distance from the camera to the object is calculated.

(Embodiment 2) FIG. 6 shows a block diagram of a distance measuring apparatus according to Embodiment 2 of the present invention. The difference from Embodiment 1 is the object detecting means 12 and the object learning image database 19. Yes, only different parts will be described below.

The operation of the object detecting means 12 according to the second embodiment will be described below. Here, FIG. 7 shows an example of the learning image.

When the position of the object is detected by the object detecting means 12, a plurality of learning images as shown in FIG. 7 are registered in the object learning image database 19 in advance.
The learning image registered in the object learning image database 19 is
For each detection target object, an image having a size corresponding to the appearance at each distance is registered for each distance from the camera set at an appropriate interval.

The image data 15 as shown in FIG. 2 input by the image input unit 11 is read and collated with the learning images registered in the respective object learning image databases 19, and the most similar learning image and the input image are compared. And the area of The position in the input image at this time is held as position information 16.

At the same time, a rough distance from the camera to the target object can be obtained from a learning image of a size corresponding to the appearance at the collated distance.

(Embodiment 3) Next, a distance measuring apparatus according to Embodiment 3 of the present invention will be described.
Is different from the processing of the road structure recognizing means 13, and how the different road structure recognizing means 13 will be described.

The operation of the road structure recognition means according to the third embodiment will be described below. Here, FIG. 8 shows the relationship between the image and the real space, where 71 is the lens center, 7
2 is the coordinate axis of the image, 73 is the coordinate axis of the real space, 74 is the position in the image, and 75 is the position of the real space. However, in both cases, the y-axis is a direction perpendicular to the paper surface. 9A and 9B show the road structure in the real space. FIG. 9A shows the road structure at a position sufficiently close to the camera at the current time, FIG. 9B shows the road structure recognition result at the previous time, and FIG. ) Indicates a road structure recognition result at the current time, and 81 indicates a coordinate axis at the previous time.

In a case where the road structure recognizing means 13 considers a case where a camera is installed behind the traveling vehicle and continuously photographs in a direction opposite to the traveling direction, it is assumed that the road is horizontal at a position sufficiently close to the camera. . The image input unit 11
2 is read, and the position (ix, iy) of the white line in the image is detected at a position sufficiently close to the camera. As shown in FIG. 8, assuming that the focal length is f, the height of the lens center is h, and the angle from the horizontal direction to the camera optical axis is θ, the position (px, py, pz) is as shown in (Equation 1).

[0041]

(Equation 1) The position of the white line in the real space is calculated from all positions of the white line in the range of the detected image that is sufficiently close to the camera by (Equation 1), and the curve is applied to obtain a position sufficiently close to the camera as shown in FIG. The road structure in the real space is required.

Here, at the previous time, it is assumed that the road structure in the real space as shown in FIG. Further, the movement amount of the vehicle itself from the previous time to the current time, that is, the parallel movement amount and the rotation angle from the coordinate axis at the previous time to the coordinate axis at the current time are separately obtained from the vehicle side. Therefore, the road structure shown in FIG. 9B at the previous time is connected to the road structure sufficiently close to the camera shown in FIG. The road structure at the current time as shown in (c) can be obtained, and this is stored as the road structure information 17.

[0043]

As described above, according to the present invention, even when the camera itself is moving and the road surface is not horizontal, the distance can be accurately measured from the image of only one visible camera. The effect is obtained.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a distance measuring device according to a first embodiment of the present invention.

FIG. 2 is a diagram showing an example of image data according to the first embodiment of the present invention.

FIG. 3 is a diagram showing an example of a differential binary image according to the first embodiment of the present invention.

FIG. 4 is a diagram showing an example of a recognition result of a road structure according to the first embodiment of the present invention;

FIG. 5 is a diagram showing an example of road edge detection and its correspondence in the first embodiment of the present invention.

FIG. 6 is a block diagram showing a distance measuring device according to a second embodiment of the present invention.

FIG. 7 is a diagram showing an example of a learning image according to the second embodiment of the present invention.

FIG. 8 is a diagram showing a relationship between an image and a real space according to the third embodiment of the present invention.

FIG. 9 is a diagram showing a road structure in a real space according to the third embodiment of the present invention;

FIG. 10 is an explanatory diagram of a conventional distance measuring method.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 11 Image input part 12 Object detection means 13 Road structure recognition means 14 Distance calculation means 15 Image data 16 Position information 17 Road structure information 18 Measurement result 19 Object learning image database 21 Object to be detected 22 Object position 23 Road 24, 25 White line 31 Preceding vehicle model 41 road shape on horizontal plane 42 road shape on longitudinal section 43 line segment representing correspondence of road edge 44 position in real space 51, 52 road edge 53 line segment representing correspondence of road edge 71 lens center 72 image 73 Coordinate axis of real space 74 Coordinate axis of image 74 Position in image 75 Position of real space 81 Coordinate axis at previous time

 ──────────────────────────────────────────────────続 き Continued on front page F term (reference) 2F065 AA12 CC11 DD02 FF04 FF09 JJ03 JJ19 JJ26 MM07 QQ04 QQ13 QQ21 QQ32 QQ39 RR02 2F112 AD05 AD10 BA09 CA05 FA08 FA27 FA35 FA38 FA39 FA45 5L096 BA04 CA08 FA03 FA04 FA66 FA66 HH21 HH23 KK56

Claims (12)

[Claims] 1. A position in an image of an object existing on a road is detected from an image input from a camera, a structure of the road surface is recognized from the image, and a position of the detected object and the road surface structure are detected. A distance measuring method, wherein a distance in a real space from a camera to an object is calculated from the distance. 2. A method for detecting a position of an object in an image, wherein a plurality of characteristics of a plurality of objects to be detected are held in advance, a characteristic amount thereof is obtained from an input image, and an object most similar to the characteristic of the object to be detected is extracted. The method according to claim 1, wherein the position of the object is detected. 3. The feature of the object is that a linear component is extracted from a differential binary image of the object, and is represented by at least one of the length, symmetry, and positional relationship of the linear component. The distance measuring method according to claim 2, wherein 4. The position of an object in an image of an object is detected by registering a plurality of types of learning images in advance and comparing the input image with each of the learning images to determine a region most similar to the input image. 2. The method according to claim 1, wherein
The distance measurement method described in. 5. The learning image according to claim 4, wherein each of the objects to be detected is held for each size on the image when the object is present at a plurality of different distances.
The distance measurement method described in. 6. The road structure is recognized by assuming a road model in advance, detecting the positions of left and right road edges from an input image, and obtaining the positions of the right and left road edges in a real space based on the road model. The distance measuring method according to claim 1, wherein a structure of a road surface is recognized. 7. The road model according to claim 6, wherein the road model is constructed on the assumption that the distance between the right and left road edges is a fixed length and the line segment connecting the road edges is always horizontal. The distance measurement method described. 8. The road edge is characterized by detecting a road edge by interpolating a white line provided at the road edge or a portion where the white line is a broken line between already detected white lines. The distance measuring method according to claim 6, wherein 9. The distance is calculated by finding left and right road edges corresponding to the position of the detected object in the image, and detecting the position in the real space based on the recognition of the road structure from the position to calculate the distance. The distance measuring method according to claim 1, wherein: 10. The method according to claim 9, wherein the left and right road edges corresponding to the position in the image of the object are obtained as a combination that connects the left and right road edges and passes the shortest line segment. The distance measurement method described. 11. An object detecting means for detecting a position in an image of an object present on a road from an image input from a camera, a road structure recognizing means for recognizing a structure of a road surface from the image, A distance measuring device, comprising: distance calculating means for calculating a distance in a real space from a camera to an object from a position of the object and the road surface structure. 12. In the road structure recognizing means, when a camera is installed behind the traveling vehicle and continuously photographs in a direction opposite to the traveling direction, it is assumed that the road is horizontal at a position sufficiently close to the camera. Based on this assumption, the position of the road edge in real space is obtained from the position of the road edge sufficiently close to the camera of the input image at the current time, and at the same time, the amount of movement of the vehicle itself in the real space from the previous time to the current time is calculated. The distance measuring device according to claim 11, wherein the distance measuring device recognizes a structure of a road surface by connecting to a position of a road edge obtained up to a previous time based on the movement amount obtained from the vehicle.