JP2962799B2 - Roadside detection device for mobile vehicles - Google Patents

Description

Description: BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a traveling road edge detection device for a moving vehicle that recognizes a traveling road environment of a vehicle by image processing.

(Prior Art) Conventionally, white lines drawn at both ends of a road are extracted from image information input by a video camera or the like by performing a correction process for removing unnecessary components included in the image, and the roadside edge is extracted. Are aware of

(Problems to be Solved by the Invention) However, in the above-described conventional example, when a shadow or reflection exists on a road surface as a road environment, the processing amount required to correct the shadow and reflection increases, and the correction time is increased. There is a problem that high-speed recognition of the road end cannot be performed because the length of the road becomes long.

The present invention has been made in view of such a point, and an object of the present invention is to propose an effective method for correcting a process of recognizing a road environment, thereby enabling high-speed recognition of a roadside. That is.

(Means for Solving the Problems) The present invention has been made for the purpose of solving the above problems, and has the following configuration as one means for solving the above problems. That is, a traveling roadside detection device for a mobile vehicle including an image input unit for external world recognition, comprising: a luminance extraction unit that extracts luminance information of an input image; and a variance extraction unit that extracts variance information of an input image. Detecting means for detecting a white line on a traveling road based on the luminance information and the dispersion information.

Preferably, the luminance extracting means extracts the luminance information based on a luminance distribution obtained by comparing the luminance of the input image with a predetermined threshold.

Preferably, the variance extracting unit extracts the variance information based on a variance coefficient obtained by comparing a variance of the input image with a predetermined threshold.

Preferably, the luminance extracting means calculates a luminance distribution by comparing a luminance of the input image with the predetermined threshold value as the luminance information, and the variance extracting means calculates a variance of the input image as the variance information. A variance coefficient is calculated by comparing the variance coefficient with a predetermined threshold, and the detecting unit compares the luminance distribution with the variance coefficient, and detects an area where the luminance distribution is large and the variance coefficient is small as a white line on a traveling road. I do.

(Operation) In the above configuration, the white line on the traveling road is recognized at high speed from the luminance information and the dispersion information.

Hereinafter, a preferred embodiment according to the present invention will be described in detail with reference to the accompanying drawings.

FIG. 1 is a block diagram showing a configuration of a traveling roadside detection device (hereinafter, referred to as a device) of a moving vehicle according to an embodiment of the present invention.

The traveling road environment ahead of the vehicle is captured by a video camera 1 such as a CCD camera shown in FIG. 1, and is input to the image input unit 1 as image information. Also, the image input unit 1
Outputs the input image information to the variance calculator 3 and the brightness calculator 4 at the same time.

The variance calculation unit 3 calculates a variance coefficient of the traveling road surface based on the input image information. Further, the brightness calculation unit 4 calculates the brightness of the traveling road surface from the image information. The calculation processes in the variance calculator 3 and the brightness calculator 4 are executed simultaneously,
The calculation results are sent to the white line detection unit 5 as dispersion information and luminance information, respectively. In addition, the white line detection unit 5 extracts a white line region on the traveling road surface based on the input dispersion information and luminance information by a method described later.

The main control unit 6 controls the entire apparatus and recognizes a traveling road structure necessary for the movement of the vehicle based on the white line area detected by the white line detection unit 5.

Hereinafter, recognition of the white line in the present embodiment will be described in detail.

FIG. 2 is an image of the front of the vehicle captured by the video camera 1. In the figure, the broken line F is a horizontal line passing through the point V at infinity, and if the inclination, the focal length, and the size of the imaging surface of the video camera 1 are known, the inclination of the traveling road surface in the area near the own vehicle is Since it can be ignored, the position of the broken line F (coordinates on the screen) can be obtained by numerical calculation.

Further, since the travel path area as the target of white line extraction is located below the broken line F (the coordinates of which are y = F) on the screen, the area is set as a search target for luminance calculation and dispersion calculation. However, since the change in brightness and color near the point V at infinity is complicated and ambiguous, the search area for the above calculation is excluded. Therefore, assuming that the width of the region to be excluded in the y coordinate direction is α, the search region for calculation is y = F−
It is a region below α.

Next, a method of calculating luminance information and shared information in the present embodiment will be described.

The luminance calculator 4 scans a predetermined area with respect to the input image information, and compares the luminance of the image in the area with a preset threshold value using a built-in comparator (not shown). Then, the obtained luminance distribution (histogram) is calculated,
Output it.

For example, scanning is performed in the x-axis direction at the position indicated by y = a in the image shown in FIG. 2, and the above-described comparison is executed to detect an image on the scanning line having a high luminance.
As a result, assuming that the luminance distribution shown in FIG. 3 (a) (the intensity of the distribution is I) is obtained, an area where the luminance is more remarkable than other image areas on the same scanning line is shown in FIG. , P ₁ , P ₂
A peak having a distribution as shown in FIG. FIG. 3C shows the luminance distribution obtained as a result of executing the same processing at y = b in FIG.

 Next, a method of calculating shared information will be described.

In general, the texture of the road surface is uniform,
It is considered that the dispersion of the image on the asphalt surface is about two to three times as large as the dispersion of the white line, the vehicle, the guardrail, and the like. In other words, a uniform reflection is obtained from a white line or the like on the traveling road, which is not random noise, and an image for the white line or the like has little variance.

Therefore, the variance calculation unit 3 scans a predetermined area with the input image information and calculates the variance coefficient of the image in the area, similarly to the luminance calculation unit 4. For example, the variance coefficient of the area is calculated in the x-axis direction at the position indicated by y = a and y = b in the image shown in FIG. FIG. 3 (b) shows the result calculated for y = a, and FIG. 3 (d) shows the result for y = b.

As described above, the dispersion coefficient of the traveling road surface is often significantly different from that of the white line or the like.
As shown in (d), (in the figure, V is shows the dispersion coefficient) dispersion coefficients corresponding to the situation of the road surface valleys P ₅ to P ₈ in the appearance.

Regarding the dispersion coefficient of the road surface, the difference between the white line and the road is remarkable in the image near the vehicle, but it is difficult to distinguish the white line and the road due to the dispersion in the image far away. Therefore, the calculation of the luminance distribution and the dispersion coefficient described above is performed in the vicinity of the vehicle, that is, in the second case.
The processing is performed sequentially from the lower side of the image shown in the figure, and in the distant part of the image, the vicinity of the point V at infinity is excluded as described above.
The process ends when y = F−α. Then, the white line detection unit 5 compares the luminance distribution for each scan obtained by the luminance calculation unit 4 and the variance calculation unit 3 with the dispersion coefficient, and satisfies the condition that the luminance distribution is large and the dispersion coefficient is small. The area is defined as a white line at the end of the traveling road.

This will be specifically described with reference to FIGS. 3A and 3B. X of peaks P ₁ and P ₂ of the luminance distribution shown in FIG.
Positions on the axis and valleys P ₅ and P ₆ of the dispersion coefficient in FIG. 3 (b)
Since the position on the x-axis matches with the position, a white line exists in that region. The same can be said for FIGS. 3 (c) and 3 (d).

As described above, according to the present embodiment, the luminance information is extracted from the image of the traveling road surface, and at the same time, the variance of the traveling road surface is obtained from the image and the feature points of the two are compared to extract the white line. Therefore, the amount of correction required for the process can be reduced, so that the position of the white line on the traveling road can be recognized efficiently and at high speed.

The present invention is not limited to the above embodiment,
For example, in order to increase the recognition accuracy of the white line, the correlation between the images may be taken or the linearity of the white line may be verified.

(Effects of the Invention) As described above, according to the present invention, there is an effect that a road end can be recognized at high speed without being affected by the shadow or reflection of the road surface.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a configuration of a traveling road edge detecting device for a moving vehicle according to an embodiment of the present invention, FIG. 2 is a diagram showing an image in front of the vehicle captured by a video camera 1, and FIG. (A) and (c) are diagrams showing the luminance distribution extracted from the image, and FIGS. 3 (b) and (d) are diagrams showing the dispersion coefficient of the image. In the figure, 1 ... video camera, 2 ... image input unit, 3 ...
Dispersion calculation unit, 4 ... Brightness calculation unit, 5 ... White line detection unit, 6
... This is the main control unit.

────────────────────────────────────────────────── ─── Continuation of the front page (56) References JP-A-1-232401 (JP, A) JP-A-2-162405 (JP, A) (58) Fields investigated (Int.Cl. ⁶ , DB name) G06T 1/00 G05D 1/02

Claims (4)

(57) [Claims] An apparatus for detecting a roadside end of a mobile vehicle, comprising image input means for external world recognition, comprising: a luminance extracting means for extracting luminance information of an input image; and a distributed means for extracting distributed information of the input image. A traveling road edge detection device for a mobile vehicle, comprising: an extraction unit; and a detection unit that detects a white line on a traveling road based on the luminance information and the dispersion information. 2. The vehicle according to claim 1, wherein said luminance extracting means extracts said luminance information based on a luminance distribution obtained by comparing a luminance of said input image with a predetermined threshold. Road end detection device. 3. The mobile vehicle according to claim 1, wherein the variance extracting unit extracts the variance information based on a variance coefficient obtained by comparing a variance of the input image with a predetermined threshold. Road end detection device. 4. The luminance extracting means calculates a luminance distribution by comparing the luminance of the input image with the predetermined threshold value as the luminance information, and the variance extracting means calculates a variance of the input image as the variance information. A variance coefficient is calculated by comparing the variance coefficient with a predetermined threshold value, and the detecting unit compares the luminance distribution with the variance coefficient, and detects a region where the luminance distribution is large and the variance coefficient is small as a white line on a traveling road. The traveling road edge detection device for a mobile vehicle according to claim 1, wherein: