JPH03166662A - Vehicle detecting device - Google Patents

Abstract

PURPOSE:To surely detect the stop of a vehicle by preparing a detecting mask with respect to a principal edge obtained from an edge image, comparing the previous image and the present image at every detecting mask, and deciding that the vehicle stops, when the prescribed number coincides in plural detecting masks. CONSTITUTION:A histogram of an edge image is derived, and with respect to a principal edge obtained by executing a threshold processing to the direction of each of the horizontal direction and the vertical direction, a labeling processing for giving an intrinsic number to every picture element of continuous edges as shown in a fiture (a) is executed. Subsequently, with regard to each level 1 - 9, a block in which only a prescribed picture element is expanded becomes a detecting mask. For instance, in the case of the label 7, a range shown by a thick line in a figure (b) becomes a detecting mask M to the label 7 (oblique line part). At every detecting mask thereof M, the previous image and the present image are compared, and when the prescribed number coincides in plural detecting mask M, it is decided that a vehicle is stopping. In such a way, the stop of the vehicle can be detected surely.

Description

[Detailed Description of the Invention] [Field of Industrial Application] The present invention relates to a vehicle detection device that automatically detects whether a vehicle is stopped in a monitoring area by applying image processing technology. It is.

[Conventional technology]

Conventionally, the following method has been used to determine whether a vehicle has stopped using a vehicle detection device that applies this type of image processing. That is, for each stored frame of the image, spatial differentiation processing is performed to extract the edges of the vehicle, and the degree of overlap CD of edges between two consecutive frames is calculated using the formula shown below.

D=(Number of pixels in the overlap between edge frames)/(Total number of pixels in the edge of the previous frame) When this 7 overlap D exceeds a certain value, it is determined that the vehicle is stopped.

[Problem to be solved by the invention]

However, according to this method, if the edges of a stopped vehicle exist in two consecutive frames so that they almost overlap, it can be reliably determined, but for example, if there is a person getting on or off the vehicle in front of the vehicle, When part of the edge of the vehicle disappears, the value of the degree of overlap D becomes small, and it may not be possible to determine whether the vehicle has stopped.

The present invention has been made in view of the above points, and its purpose is to prevent part of the edge of the vehicle from disappearing due to the presence of an obstacle, etc. in front of the vehicle. Another object of the present invention is to provide a vehicle detection device that can reliably detect the stoppage of a vehicle.

[Means to solve the problem]

The vehicle detection device of the present invention includes an imaging means for photographing a monitoring area to obtain an image signal, an A/D means for digitizing the image signal obtained by the imaging means, and an A/D means for digitizing the image signal obtained by the imaging means. a spatial differentiator that performs spatial first-order differential processing and outputs a differential absolute value; a comparison device that extracts an edge image by comparing the differential absolute value with a preset setting value; and a histogram of the edge image. a histogram measurement means that calculates the main edges by performing threshold processing in each of the horizontal and vertical directions; a previous image memory that stores the main edges of the previous image; and a previous image memory that stores the main edges of the previous image; a detection mask creation means for creating a detection mask; and a determination means for calculating an exclusive OR of the main edges of the current image and the main edges of the previous image for each detection mask, and determining whether the vehicle has stopped based on the result. Further, the detection mask creation means labels the main edge and expands it by a predetermined number of pixels for each labeling to create a detection mask, Further, the detection mask creation means is configured to use a quadrilateral whose two adjacent pieces have lengths in both directions determined by projecting the main edge in the vertical and horizontal directions as the detection mask. It is characterized by:

[Effect]

The vehicle detection device of the present invention obtains a histogram of an edge image, and performs threshold processing in each of the horizontal and vertical directions to apply a detection mask to each of the main edges. The previous image and current image are compared for each detection mask, and among the multiple detection masks,
When a predetermined number of matches occur, it is determined that the vehicle is stopped.

〔Example〕

Hereinafter, one embodiment of the present invention will be described based on the drawings. 1st
FIG. 1 is a block diagram showing a schematic structure of an embodiment of the present invention. Reference numeral 1 denotes an imaging means, which is comprised of, for example, a television camera using a CCD imaging device, a vidicon, or the like. 2 is A/D
It is a means for digitizing the image signal obtained by the imaging means 1 to obtain a one-bit or multiple-bit digital signal for each pixel. 3 is a spatial differentiation means,
The digitized image data is subjected to spatial first-order differential processing, and the differential absolute value is output.

FIG. 2 is an explanatory diagram of the operation of the spatial differentiator 3.

As shown in the figure, when the luminance value of each pixel of the local block S in the image is expressed as A-1, the luminance change in the vertical direction is expressed as ΔV= (A+B+C)-CG+H+1),
The luminance change in the horizontal direction is expressed as ΔH=(A+D+G)−(C+F+I). From this, the differential absolute value let is t e l = (Δ
Hz+Δv! )0.4i. This differential absolute value 1el
The magnitude of the brightness change, that is, the edge strength, can be determined.

The differential absolute value let obtained in this manner is with respect to the center pixel width of the local block S. If this local block S is scanned, the differential absolute value te of all pixels
Therefore, l can be found.

4 is a comparison means, which compares the differential absolute value output for each pixel from the 1,000-stage spatial differential 3 with a preset value, and generates a binary image (edge image) as shown in FIG. 3(a). Convert to

Reference numeral 5 denotes a current image memory, which temporarily stores the edge image output from the comparing means 4.

6 is the hiss]・gram measurement means, and the current image memo 'J
As shown in FIG. 3(b), the edge image outputted from 5 is projected in the horizontal direction and vertical direction, respectively, and then threshold processing is performed, as shown in FIG. 3(c).
Extract major edges.

Reference numeral 7 denotes a detection mask control means, which generates a plurality of blocks including main edges from the edge image, and uses these blocks as a detection mask.

FIG. 4 shows an example of a method for controlling a detection mask.

The main edge image shown in FIG. 3(C) is subjected to labeling processing in which a unique number is assigned to each continuous edge pixel as shown in FIG. 4(a). Next, each label ■～■
, a block expanded by a predetermined number of pixels is used as a detection mask. For example, in the case of the label ■, the range indicated by the thick line in FIG. 4(b) becomes the detection mask M for the label ■ (shaded area). Similarly, a detection mask is created for each label.

Next, another example of the detection mask creation means is shown in FIG.

For the main edge image shown in FIG. 3(C), for example, first extract the leftmost main edge Ed (the part exceeding the threshold existing at the left end of the vertical histogram in FIG. 3(b)), The length X in the X direction ({ti! corresponding to the number of pixels) is determined.

Next, the length YI (value corresponding to the number of pixels) in the Y direction is determined by measuring a horizontal histogram for this main edge E4. In this way, for the main edge E, a rectangle with two adjacent pieces having a length X in the X direction and a length Y in the Y direction is set as the detection mask M for the main edge E4. . This process is performed on all other major edges. Compared to the above-described creation means, this detection mask creation means does not require labeling processing, so the processing is simpler.

A previous image memory 8 stores the main edge image and detection mask of one frame before.

Determining means 9 determines whether the vehicle has stopped, and calculates the exclusive OR of all pixels for each detection mask for the previous image and the current image to determine the degree of coincidence. That is, in the detection mask M shown in FIG. 4(b), the diagonally shaded portion corresponding to the label (3) is set to ゛1', and the other portions are calculated as ゛0°'. If edges of the same shape exist at the same position within the detection mask M, the number of pixels for which the calculation result is O increases. The degree of matching is determined based on the ratio of pixels whose calculation result is 0 in the detection mask M to the total number of pixels in the detection mask M.

The above processing is performed for all detection masks. Then, when the number of matching detection masks exceeds a predetermined number, it is determined that the vehicle is stopped.

According to the above embodiment, even if the edge image is not completely undefeated, the determination is made based on the number of matching detection masks, so if there is an obstacle such as a person in front of the vehicle, Even if there is a discrepancy in some areas of the detection area, it is possible to reliably detect a stopped vehicle.

〔Effect of the invention〕

As described above, according to the vehicle detection device of the present invention, a histogram of an edge image is obtained, and for the main edges obtained by performing threshold processing in each of the horizontal and vertical directions, A detection mask is created for each major edge, the previous image and the current image are compared for each detection mask, and when a predetermined number of detection masks match, it is determined that the vehicle is stopped. This makes it possible to provide a vehicle detection device that can reliably detect the stoppage of a vehicle even if part of the edge of the vehicle disappears due to the presence of an obstacle in front of the vehicle. Ta.

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing an embodiment of the present invention, FIG. 2 is a schematic diagram showing a spatial differentiation means according to the above, FIG. 3 is an explanatory diagram of main edge extraction according to the above, and FIG. FIG. 2 is an explanatory diagram showing an example of creating a detection mask according to the example above. FIG. 5 is an explanatory diagram showing another example of creating a detection mask. 1 - Imaging means 2 - A/D means 3 - Spatial differentiation means 4... Comparison means 5 - Current image memory 6 -
Histogram measurement 1,000 steps 71 detection mask method

Claims (3)

[Claims] (1) Imaging means for photographing a monitoring area and obtaining an image signal;
A/D means for digitizing the image signal obtained by the imaging means; spatial differentiation means for performing spatial first differentiation processing on the digitized image data and outputting a differential absolute value; a comparison means for extracting an edge image by comparing the value with a preset setting value; and a comparison means for extracting an edge image by comparing the edge image with a preset setting value; a histogram measurement means for determining the main edges of the previous image; a previous image memory for storing the main edges of the previous image; a detection mask creation means for creating a detection mask using the main edges; A vehicle detection device comprising determining means for calculating an exclusive OR of main edges of a previous image and determining whether a vehicle has stopped based on the result. (2) Claim 1, wherein the detection mask creation means labels the main edge and expands it by a predetermined pixel for each labeling to create a detection mask.
Vehicle detection device described. (3) The detection mask creation means uses a quadrilateral whose two adjacent pieces have lengths in both directions determined by projecting the main edge in the vertical and horizontal directions as the detection mask. The vehicle detection device according to claim 1.