JP2006092097A - Vehicle sensing device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a vehicle sensing device for outputting a video for measuring a traffic flow and a video for monitoring safety on a road with naked eyes while reducing the number of cameras to be used. <P>SOLUTION: In a vehicle sensing device, a CPU 20 makes an ITV camera unit 1 photograph an image with different incident light quantity L1 and L2, and performs vehicle judgement from the image photographed with the incident light quantity L1, and outputs the result of the vehicle judgement through an I/O 23 to an external device. Also, the CPU 20 outputs the image photographed by the incident light quantity L2 through an I/O 23 to an external device as a static image. That is, the CPU 20 outputs the result of the vehicle judgement to the external device every 0.2 second, and outputs the static image to the external device. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to a vehicle sensing device, and in particular, detects the number of vehicles existing in a certain area of the road based on an image of the certain area on the road photographed by a camera, and monitors the road with human eyes. The present invention relates to a vehicle sensing device that outputs an image for the purpose.

  Conventionally, images taken by a camera such as an ITV (Industrial Television) camera have been used for measuring traffic flow on a road. Specifically, a vehicle sensing device that measures a traffic flow by processing an image captured by a camera has been conventionally proposed. In such a vehicle sensing device, generally, a headlight or a vehicle contour ( (Hereinafter referred to as “vehicle body”) is a detection target, and the presence / absence of the vehicle and the number of vehicles are measured based on the detection result.

  In such a vehicle sensing device, since the headlight is bright at night when the illumination is low, the vehicle body is dark, so that the headlight of the vehicle is recognized with the incident light amount of the ITV camera reduced. The vehicle was detected by

And in patent document 1, in such a vehicle sensing apparatus, the problem in the vehicle detection in both the cases where the incident light quantity of the ITV camera is not reduced and when it is reduced is mentioned, and in order to solve these The vehicle is discriminated based on the images of the same road taken by changing the amount of incident light of the ITV camera, and the vehicle detection result obtained by removing the overlap of the detected vehicles from these discrimination results Is disclosed as a technique.
JP-A-8-235492

  A camera image that captures roads for traffic monitoring is also required for traffic flow measurement by recognizing headlights as described above. It is also required to monitor the above safety.

  However, as described above, an image used for recognizing a headlight or the like for traffic flow measurement needs to be photographed with a reduced incident light amount at night or the like. In other words, the image used for traffic flow measurement may be taken with the amount of incident light reduced, and thus may not be suitable as an image to be monitored by the human eye. Therefore, in order to obtain an image for a person to monitor with the naked eye, it is necessary to newly provide a camera different from the camera for measuring the traffic flow and photograph the road.

  In this way, when more cameras are needed, it is not only costly, but the driver of a car traveling on the road will be photographed with so many cameras, so it is given to the driver “being photographed” It is thought that the feeling of discomfort increases.

  The present invention has been devised in view of the above-described circumstances, and its purpose is to monitor safety on the road with traffic flow measurement images and the naked eye while reducing the number of cameras used. It is providing the vehicle detection apparatus which can output the image | video of this.

  A vehicle sensing device according to the present invention discriminates a vehicle on the basis of a camera installed on the side of a road and photographing a road, a light amount adjusting means for adjusting an incident light amount of the camera, and an image photographed by the camera. A vehicle discrimination means; and an image for vehicle discrimination, which is an image captured by causing the light intensity adjustment means to make the incident light quantity of the camera the first light quantity, to the vehicle discrimination means; Output control means for outputting an output image, which is an image taken with the incident light amount set to a second light amount different from the first light amount, to the outside.

  In the vehicle sensing device according to the present invention, the light amount adjusting means adjusts the incident light amount by changing at least one of a lens aperture, a lens system transmittance, and a shutter speed in the camera. It is preferable.

  In the vehicle sensing device according to the present invention, the first light amount is a light amount for photographing only the headlight of the vehicle at night, and the second light amount can be visually recognized with a human eye. It is preferable that the amount of light is for taking an image.

  In the vehicle sensing device according to the present invention, it is preferable that the output control means outputs the vehicle discrimination image and the output image at regular intervals.

  In the vehicle sensing device according to the present invention, it is preferable that the output control means outputs the output image as a still image.

  In the vehicle sensing device according to the present invention, the output control means may output the output image taken in a frame other than a frame that causes the camera to take the vehicle discrimination image as a moving image. preferable.

  In addition, the vehicle sensing device according to the present invention further includes an image memory for accumulating images, and the output control means includes the frame in all frames other than the frame that causes the camera to capture the vehicle discrimination image. The image for output is photographed, the image for output is output to the outside as a moving image, the image for output captured in the frame immediately before the frame in which the image for vehicle determination is captured is stored in the image memory, It is preferable that the output image stored in the image memory is output to the outside simultaneously with outputting the vehicle determination image to the vehicle determination means.

  According to the present invention, the vehicle sensing device can perform vehicle discrimination based on an image captured by a camera and output an image captured by the camera to the outside. The camera can take an image for vehicle discrimination and an image output to the outside with different amounts of incident light. Therefore, the vehicle sensing device can make an image output to the outside suitable for monitoring a road with the naked eye of a human, unlike an image for vehicle discrimination.

  In other words, the vehicle sensing device of the present invention can capture an image for vehicle discrimination and an image for monitoring a road with the naked eye of a person with the same camera, so that the number of cameras to be used is reduced, It is possible to output a video for measuring traffic flow and a video for monitoring safety on the road with the naked eye.

  In addition, according to the present invention, the vehicle sensing device can output a vehicle discrimination image and an output image (an image output to the outside for monitoring) at regular intervals. As a result, a vehicle discrimination result can be obtained at regular time intervals, and a monitoring video can be obtained at regular time intervals.

  According to the present invention, the vehicle sensing device can output a moving image as an image output to the outside for monitoring. This makes it easier to monitor the safety on the road.

  In addition, the vehicle sensing device also determines the timing at which the moving image output to the outside for monitoring corresponds to a frame in which the image is not captured, that is, the timing corresponding to the frame in which the vehicle sensing image is captured. , Image information will be included. Note that the image information is an output image of a frame immediately before the timing, which is stored in the image memory. As a result, the image (moving image) output to the outside for monitoring is made more uncomfortable to the human eye.

  Below, an example of the vehicle sensing device which is one embodiment of the present invention is explained.

  FIG. 1 is a conceptual diagram of installation of a vehicle sensing device. FIG. 1A is a plan view showing the vicinity of a place where the vehicle sensing device is installed, and FIG. 1B is an elevation view showing the vicinity of the place.

  Referring to FIG. 1, the ITV camera unit 1 is provided at the upper part of a pole (a place where the height from the ground is about 10 m) installed at the intersection of a four-lane road on one side. Moreover, the control part 2 is provided in the lower part of the said pole. The ITV camera unit 1 includes an ITV camera. The ITV camera monitors an area of about 150 m length on a four-lane road.

  FIG. 2 is a block diagram showing an electrical configuration of the vehicle sensing device of the present embodiment.

  Referring to FIG. 2, the vehicle sensing device includes an ITV camera unit 1 and a control unit 2.

  The control unit 2 is acquired from a CPU (central processing unit) 20 that controls the control unit 2 as a whole, an image memory 21 that temporarily stores an image signal input from the ITV camera unit 1, and the ITV camera unit 1. An A / D (analog / digital) converter 22 that converts the image signal into a digital signal, and an input / output device (I / O) 23 that performs processing for transmitting and receiving information between the control unit 2 and the outside. A D / A (digital / analog) converter 24 for converting an image signal (digital signal) stored in the image memory 21 into analog data, and a ROM (read only) for storing various information such as a program executed by the CPU 20 memory) 25.

  The ITV camera unit 1 shoots by changing the shutter speed electronically by changing the charge accumulation time of a CCD (charge-coupled device).

  Assuming that 1/30 second is a unit and one unit is one frame, the ITV camera unit 1 shoots at a rate of twice every six frames as shown in FIG. FIG. 3 is a diagram showing the relationship between the timing at which the ITV camera unit 1 performs shooting and the amount of incident light in the ITV camera unit 1 at each timing. The ITV camera unit 1 shoots with the light amount limited by shortening the shutter speed at the first time out of 6 frames. The amount of light in this case is L1. In addition, the ITV camera unit 1 takes a picture at the second time by making the shutter speed slower than the first time to make the light amount limit smaller than the first time. The light quantity in this case is L2. That is, the ITV camera unit 1 outputs two image signals having different incident light amounts, L1 and L2, every 6 frames, that is, every 0.2 seconds.

  The amount of incident light in the ITV camera unit 1 is changed when the CPU 20 controls the ITV camera unit 1 via the I / O 23 to change the shutter speed.

  In addition, the CPU 20 can perform processing for discriminating the vehicle with respect to the image signal photographed by the ITV camera unit 1 and stored in the image memory 21, or can output the image signal as a still image to an external device. . Specifically, the CPU 20 performs a vehicle discrimination process based on an image signal that is captured and generated with the incident light amount L1, and the image signal that is captured and generated with the incident light amount L2 is used as a still image. Output to an external device via / O23. The CPU 20 also outputs the result of the vehicle determination process to an external device via the I / O 23.

  Here, the vehicle discrimination process will be described.

  FIG. 4 is a diagram illustrating an example of an image captured with the incident light amount set to L1 for vehicle discrimination. In the image 200 shown in FIG. 4, the vehicle body cannot be detected because the amount of incident light is limited, but the headlights of the vehicles 201 to 207 appear to be separated even when the vehicle is far from the intersection.

  The CPU 20 determines the presence of a vehicle within the monitoring range of the ITV camera of the ITV camera unit 1 based on an image photographed by the ITV camera unit 1 like the image 200. Specific processing for this determination will be described below.

  First, the CPU 20 sets measurement sample points so that the images taken by the ITV camera unit 1 have a grid pattern, and performs a differential operation on the image constituted by the measurement sample points. The image is edge-enhanced, and the result is binarized. The image obtained by binarizing is applied with a mask of the same size as the vehicle body, and based on the number of edge pixels in the mask. Extract car body candidates. Then, the vehicle body of the vehicle is determined by determining feature amounts such as the vehicle width, the vehicle length, and the headlight interval based on the extracted profile of the vehicle body candidate (see Japanese Patent Laid-Open No. 5-307695).

  If the vehicle is determined using the image 200, the vehicles 201 to 207 are determined based on the headlights in order from the side closer to the intersection in the image 200 (near the lower end of the image 200). The In this way, the number of vehicles existing within the monitoring range of the ITV camera and the congestion length can be detected. Also, based on the results of discrimination using a plurality of images taken at different timings, the vehicle speed can be measured by tracking the vehicle between the different timings, and the moving vehicle and the stopped vehicle can be distinguished. (For the speed measurement method, refer to Japanese Patent Laid-Open No. 5-31818).

  In the present embodiment described above, the CPU 20 causes the ITV camera unit 1 to capture images with different incident light amounts of L1 and L2. Then, the CPU 20 determines the vehicle for the image captured with the incident light amount L1, and outputs the result of the vehicle determination to an external device via the I / O 23. Further, the CPU 20 outputs an image photographed with the incident light quantity L2 as a still image to an external device via the I / O 23. In other words, in the present embodiment, the CPU 20 outputs the result of vehicle discrimination to an external device and outputs a still image to the external device every 0.2 seconds.

  In the present embodiment, the incident light amount L1 is an incident light amount that is limited by the ITV camera, and the incident light amount L2 is an incident light amount that is limited so that the light amount is larger than the incident light amount L1. However, it is not limited to this. The incident light amount L1 may be an incident light amount for photographing an image suitable for vehicle detection. For example, when photographing is performed at night, the incident light amount L1 is a light amount for photographing only the headlight of the vehicle. . Further, the incident light amount L2 may be a light amount such that the image captured by the ITV camera unit 1 has a contrast that allows the human eye to visually recognize the state within the monitoring range of the ITV camera. For this reason, for example, in the case of shooting when the sunlight is strong, when the light amount is limited so that the incident light amount L2 <incident light amount L1 in the ITV camera at the timing when the image of the incident light amount L2 is captured. There is also.

  In this embodiment, the amount of incident light of the ITV camera is controlled by changing the shutter speed. However, the present invention is not limited to this, or by changing the aperture of the lens of the ITV camera, or The amount of incident light may be controlled by changing the transmittance of the lens system by electrically fogging the lens.

  Further, in the present embodiment, the CPU 20 outputs the still image captured with the incident light amount L2 to the outside every 0.2 seconds, but the present embodiment is not limited to this. That is, the CPU 20 may output the still image to the outside only at a timing when an external request input via the I / O 23 is received.

  In the present embodiment, the photographing with the incident light amount L2 is performed only in the second frame of every six frames, but the present invention is not limited to this. The CPU 20 may cause the ITV camera unit 1 to perform photographing with the incident light amount L2 for each frame other than a frame where photographing with the incident light amount L1 is performed. In other words, in this case, as shown in FIG. 5, the ITV camera unit 1 performs shooting with the incident light amount L1 for vehicle discrimination in the first frame in groups of six frames, and in the second frame. In the fourth and sixth frames, photographing is performed with an incident light quantity L2 for output to the outside for monitoring. As a result, the CPU 20 can output an image photographed for monitoring every 2/30 seconds, that is, every 1/15 seconds. The monitoring image output in this way may be transmitted as a still image or may be output as a moving image via the D / A converter 24.

  Further, in the present embodiment, the CPU 20 may cause the ITV camera unit 1 to perform photographing with the incident light amount L2 in all frames that have not been photographed with the incident light amount L1. Further, the CPU 20 accumulates an image taken immediately before the ITV camera unit 1 takes an image with the incident light amount L1 in the image memory 21, and an image taken with the incident light amount L2 for photographing the incident light amount L1. When the image cannot be output, the image stored immediately before in the image memory 21 may be output to the outside.

  That is, in this case, as shown in FIG. 6, the ITV camera unit 1 performs shooting with the incident light quantity L1 for vehicle discrimination in the first frame in groups of six frames, and from the second frame. In the frames up to the sixth frame, photographing is performed with an incident light quantity L2 for output to the outside for monitoring. Further, the CPU 20 outputs an image photographed with the incident light amount L2 in the immediately preceding frame of the first frame of each group (sixth frame of the previous group) to the outside and stores it in the image memory 21. The stored image is output to an external device for monitoring at a timing corresponding to the frame.

  Thus, the CPU 20 can output the result of vehicle discrimination every 0.2 seconds and can output a continuous monitoring image every 1/30 seconds. In this case, the image output for monitoring may be output as a moving image.

  In this case, in the image output for monitoring (image taken with the incident light amount L2), as shown in FIG. 6, for the group every 6 frames, the first frame and the previous group 6 The frame is the same image. However, if it is this level, it is thought that it can be visually recognized without a sense of incongruity with the human naked eye.

  The embodiment disclosed this time should be considered as illustrative in all points and not restrictive. The scope of the present invention is defined by the terms of the claims, rather than the description above, and is intended to include any modifications within the scope and meaning equivalent to the terms of the claims.

It is an installation conceptual diagram of the vehicle sensing device which is one embodiment of the present invention. It is a block diagram which shows the electric constitution of the vehicle sensing apparatus of FIG. It is a figure which shows the relationship between the timing which the ITV camera unit of the vehicle detection apparatus of FIG. 1 image | photographs, and incident light quantity. FIG. 2 is a diagram illustrating an example of an image captured with the amount of incident light adjusted for vehicle discrimination in the vehicle sensing device of FIG. 1. It is a figure which shows the modification of the relationship between the timing shown by FIG. 3, and incident light quantity. It is a figure which shows another modification of the relationship between the timing shown by FIG. 3, and incident light quantity.

Explanation of symbols

  1 ITV camera unit, 2 control unit, 20 CPU, 21 image memory, 22 A / D converter, 23 input / output device, 24 D / A converter, 25 ROM.

Claims (7)

A camera installed on the side of the road and shooting the road;
A light amount adjusting means for adjusting the incident light amount of the camera;
Vehicle discriminating means for discriminating a vehicle based on an image taken by the camera;
A vehicle discrimination image, which is an image captured by setting the incident light amount of the camera as the first light amount to the light amount adjustment unit, is output to the vehicle determination unit, and the incident light amount of the camera is output to the light amount adjustment unit. A vehicle sensing device comprising: output control means for outputting an output image, which is an image captured with a second light amount different from the first light amount, to the outside.   2. The vehicle sensing device according to claim 1, wherein the light amount adjusting unit adjusts an incident light amount by changing at least one of a lens aperture, a lens system transmittance, and a shutter speed in the camera. The first light amount is a light amount for photographing only a vehicle headlight at night,
3. The vehicle sensing device according to claim 1, wherein the second light amount is a light amount for capturing an image whose contrast can be visually recognized with a human naked eye.   4. The vehicle sensing device according to claim 1, wherein the output control unit outputs the vehicle determination image and the output image at regular intervals. 5.   The vehicle detection device according to claim 1, wherein the output control unit outputs the output image as a still image.   5. The output control unit according to claim 1, wherein the output control unit outputs the output image captured in a frame other than a frame that causes the camera to capture the vehicle determination image as a moving image. 6. Vehicle sensing device. An image memory for storing images;
The output control means includes
Causing the camera to capture the output image in all frames other than the frame in which the vehicle determination image is captured;
The output image is output to the outside as a movie,
Storing the output image captured in a frame immediately before the frame in which the vehicle determination image is captured, in the image memory;
5. The vehicle sensing device according to claim 1, wherein the output image stored in the image memory is output to the outside simultaneously with outputting the vehicle determination image to the vehicle determination unit. .

Images