CN103886759B - System and method for snapshotting behavior of running red light - Google Patents

Abstract

The invention provides a system and method for snapshotting the behavior of running the red light. The method for snapshotting the behavior of running the red light comprises the steps that a vehicle head picture is snapshotted by a second camera, license plate information is recognized according to the vehicle head picture, and the vehicle head picture, the license plate information and the snapshotting time are sent to an information processing device to be cached; whether a vehicle breaks rules and regulations is recognized by a first camera, and if the vehicle breaks the rules and regulations, the license plate information is recognized according to a vehicle tail picture, and a regulation violation picture, the license plate information of the vehicle breaking the rules and regulations, and the regulation violation time are sent to the information processing device to be cached; after the information processing device receives the regulation violation picture, the license plate information of the vehicle breaking the rules and regulations, and the regulation violation time, a vehicle head picture matched with the regulation violation picture is found during a snapshotting period according to which the time difference between the regulation violation time and snapshotting time of the vehicle head picture is smaller than a threshold value, and the regulation violation picture and the found vehicle head picture are combined to form a regulation violation record, wherein the regulation violation picture, the license plate information of the vehicle breaking the rules and regulations, and the regulation violation time are sent by the first camera. According to the system and method for snapshotting the behavior of running the red light, the second camera is additionally arranged for acquiring the vehicle head pictures, then head portraits of drivers are obtained, and drivers corresponding to the regulation violation records can be found.

Description

System and method for capturing images of running a red light

technical field

The invention relates to the technical field of video surveillance, in particular to a system and method for capturing images of running a red light.

Background technique

As the flow of road traffic increases year by year, for the management of road traffic and the supervision of violations, the electronic police system came into being (hereinafter referred to as "electric police system"). The electric police system is usually installed on signal-controlled intersections and road sections. It is mainly responsible for continuous automatic detection and recording of violations of motor vehicles in designated lanes such as running red lights, pressing lines, and not driving in designated lanes. The electric police system is generally composed of a capture camera, a front-end processing terminal (optional when the front-end camera does not have image processing functions), and a central platform.

Currently, the implementation of the electric police system is shown in Figure 1. Among them, the snap camera is generally installed behind the stop line, aimed at the rear of the vehicle, and facing the traffic lights at the same time, to obtain images when the vehicle is driving to multiple positions on the road, thus recording the entire process of the vehicle running the red light. Due to the high definition of the current capture camera, the license plate information (ABC) of the vehicle can be distinguished by taking a photo. When there is a vehicle violation, the system will provide the traffic control department with the combination of the previously captured vehicle picture and the violation record as evidence of the violation. Please refer to Figure 2 for the violation record. The three pictures in the upper left corner, upper right corner and lower left corner are images of the process of vehicle violation, and the picture in the lower right corner is a close-up of the license plate of the vehicle, which is usually called "passing car image".

However, in this shooting method, since the shooting direction is behind the vehicle, the driver’s head portrait information cannot be viewed. As a result, the traffic police can only confirm the violation of the vehicle by the violation record, but cannot confirm the violation of the driver driving the vehicle. The driver will deduct points. For public vehicles belonging to enterprises and institutions, there may be multiple drivers taking turns driving. Therefore, the traditional electric police system can only fine public vehicles and cannot deduct points for drivers, which is not conducive to the management of some drivers.

Contents of the invention

In view of this, the present invention provides a red light running capture system and method, which can capture the head portrait of the violating driver.

The red light running capture system includes a first camera, a second camera and an information processing device;

The second camera is used to capture a photo of the front of the car, recognize the license plate information according to the photo of the front of the car, and send the photo of the front of the car, the license plate information and the snapshot time to the cache of the information processing device;

The first camera is used to identify whether the vehicle violates the regulations, and if it violates the regulations, recognize the license plate information according to the violation photos including the rear image, and send the violation photos, violation license plate information and violation time to the information processing device;

The information processing device is configured to pass the license plate information of the violating vehicle after receiving the photo of the violation, the license plate information of the violation and the time of violation sent by the first camera. Searching for a vehicle front photo that matches the violation photo within the capture time of the threshold, and merging the violation photo and the found vehicle front photo into a violation record.

Cut out and enlarge the front photo with the driver's head portrait and license plate from the found photo of the car that matches the photo of the violation, and merge it with the photo of the violation to form a record of violation.

The information processing device is set in the first camera or in the front-end processing terminal or on the platform server.

There are three photos of the violation, including the photo of the vehicle body not crossing the stop line, the photo of the vehicle body having crossed the stop line, and the photo of the vehicle body crossing the stop line and continuing to drive towards the traffic lights, and the traffic light images are also included in the violation photos .

If the time difference between the violation time and the snapshot time of the front photo is less than the threshold, the front photo matching the violation photo is not found, then the car with the rear image and license plate information is extracted from the violation photo. The tail photo, and the said tail photo and the violation photo are combined into one violation record.

Based on the same idea, the present invention provides a method for snapping photos of running a red light, comprising steps:

The second camera captures the picture of the front of the car, recognizes the license plate information according to the picture of the front of the car, and sends the picture of the front of the car, the license plate information and the capture time to the cache of the information processing device;

The first camera identifies whether the vehicle violates the regulations, and if it violates the regulations, it recognizes the license plate information according to the violation photos containing the rear image, and sends the violation photos, the violation license plate information and the violation time to the information processing device;

After receiving the illegal photo, license plate information and time of violation sent by the first camera, the information processing device uses the license plate information of the violating vehicle to search within the time when the time difference between the time of violation and the snapshot time of the front photo is less than a threshold value. For the photo of the front of the car that matches the photo of violation, through the license plate information of the violation, search for the photo of the front of the car that matches the photo of violation in the snapshot time when the time difference between the time of violation and the time of capturing the photo of the front of the car is less than the threshold, and store the photo of violation of the law And the found photos of the front of the car are combined into one violation record.

Cut out and enlarge the front photo with the driver's head portrait and license plate from the found photo of the car that matches the photo of the violation, and merge it with the photo of the violation to form a record of violation.

The information processing device is set in the first camera or in the front-end processing terminal or on the platform server.

There are three photos of the violation, including the photo of the vehicle body not crossing the stop line, the photo of the vehicle body having crossed the stop line, and the photo of the vehicle body crossing the stop line and continuing to drive towards the traffic lights, and the traffic light images are also included in the violation photos .

If the time difference between the violation time and the snapshot time of the front photo is less than the threshold, the front photo matching the violation photo is not found, then the car with the rear image and license plate information is extracted from the violation photo. The tail photo, and the said tail photo and the violation photo are combined into one violation record.

Compared with the prior art, the present invention can obtain the front photo by adding a second camera, and obtain the head portrait information of the violating driver from the snapped front photo. When there are multiple drivers using the same car, the violating driver can be effectively found, and the violating driver can be managed and punished. Therefore, it makes up for the lack of the driver's head portrait in the violation records of the traditional electric police system, and enhances the comprehensiveness of traffic management.

Description of drawings

Fig. 1 is the schematic diagram of a kind of electric police system in the prior art;

Fig. 2 is a schematic diagram of a violation record in the prior art;

Fig. 3 is a schematic structural diagram of a red-light snapping system provided by the present invention;

Fig. 4 is a schematic structural diagram of a red light running capture system in an embodiment of the present invention;

Fig. 5 is a schematic structural diagram of another red light capture system in an embodiment of the present invention;

Fig. 6 is a schematic diagram of a violation record in the embodiment of the present invention.

detailed description

In order to solve the problem that the electric police system cannot capture the driver's head portrait, a system and method for capturing traffic violating drivers are provided in the embodiment of the present invention.

Please refer to FIG. 3 , which is a schematic structural diagram of a red light running capture system provided by the present invention, wherein the system includes a first camera, a second camera and an information processing device;

Wherein, the second camera is used to capture a picture of the front of the car, recognize the license plate information according to the picture of the front of the car, and send the picture of the front of the car, the license plate information and the time of the snapshot to the cache of the information processing device. Photos of the front of the car include the driver's headshot, license plate, etc. Because it is difficult to capture human faces in a moving vehicle, the capture camera needs to have a higher resolution, and the resolution should be selected according to the size of the shooting area set by the capture camera. A camera with a general resolution is usually only responsible for capturing For one lane, a high-resolution camera can be used to capture multiple lanes.

The first camera is used to identify whether the vehicle is violating the rules, if it is violating the rules, the information of the illegal license plate is identified according to the illegal photo containing the image of the rear of the vehicle, and the illegal photo, the illegal license plate information and the time of violation are sent to the information processing device; The above-mentioned first camera is mainly responsible for photographing the rear of the vehicle and reference objects on the lane, such as traffic lights (traffic lights) and stop lines, through which the driving position of the vehicle in the lane can be compared.

The information processing device is configured to pass the license plate information of the violating vehicle after receiving the photo of the violation, the license plate information of the violation and the time of violation sent by the first camera. Searching for a vehicle front photo that matches the violation photo within the capture time of the threshold, and merging the violation photo and the found vehicle front photo into a violation record.

It can be seen that, the present invention can obtain the front photo by adding a second camera, and obtain the head portrait information of the violating driver from the snapped front photo. When there are multiple drivers using the same car, the violating driver can be effectively found, and the violating driver can be managed and punished. Therefore, it makes up for the lack of the driver's head portrait in the violation records of the traditional electric police system, and enhances the comprehensiveness of traffic management.

In order to make it easier for the traffic police to see the head portrait of the driver in the violation record, a more preferred implementation is to cut out and enlarge the head photo with the driver's head portrait and license plate from the found photos matching the violation photo, and compare it with the Violation photos are combined into one violation record. This makes violation records more convincing.

In practical applications, the application scheme of the information processing device is not unique (it can be realized by a software program with the above-mentioned information processing function, or by a hardware device that can replace the software function). The device is usually applied to the first camera, or to a front-end processing terminal, or to a system platform.

In the embodiment of the present invention, when the first camera has strong computing and processing capabilities, the information processing device application can be set on the first camera, and other cameras will send the photos taken by themselves to the On the first camera with information processing function, he will further process the captured photos and identified information. It is worth noting that, due to the limited processing power and cache capacity of the first camera, even if the above-mentioned preset time is used to reduce the cache burden, there are still many photos captured by the snap camera on roads with many vehicles, so , the present invention can also further set a preset buffer size (such as 10) for the buffer size, then when the information processing device in the first camera analyzes the buffered images, it only needs to match the photos within 10 and analysis, thereby reducing the burden on the processor and improving work efficiency.

In addition, when the configuration of the first camera is relatively low and cannot process images, the information processing device can also be set on the front-end processing terminal connected to the first camera, and the first camera and the second camera will take pictures by themselves Both the photos and the identified information are sent to the front-end processing terminal with information processing function, and he will further process the captured photos.

If in an electric police system that is not equipped with a front-end processing terminal, the information processing device can also be used on the system platform server, and the first camera and the second camera will send the photos taken by themselves and the information identified to the information processing device. On the system platform with various functions, he will further process the captured photos and identification information.

In a preferred embodiment, there are three photographs of violations captured by the first camera of the present invention. According to relevant laws and regulations, the photographs of violations respectively include photos of the vehicle body not crossing the stop line, and photos of the vehicle body having crossed the stop line. And the photo of the vehicle body continuing to drive towards the traffic light after crossing the stop line, and the traffic light image is also included in the said violation photo. According to the license plate information in the violation photo identified by the first camera, the information processing device searches the cache for a headshot with the same license plate information according to the license plate information, and the time when the front photo is taken and the violation time A front photo is selected from the front photos whose difference is less than a preset time threshold, and the front photo and the violation photo are combined into one violation record. It should be noted that the setting of the time threshold needs to be set according to the size of the shooting area, according to factors such as the length of the actual road, the average speed of the vehicle, and the distance from the stop line. Usually, the rear camera is installed at a distance of tens of meters from the traffic lights. If the resolution of the camera is higher, it can be installed at a farther distance. According to the average value of current tests, the preset time threshold is generally set within 3 minutes. Then, if there is no special requirement, the front photos in the cache that exceed the preset time threshold can be deleted, thereby saving cache space and reducing the computational burden.

Considering that when the capture camera of the system breaks down, it is possible that the front photo cannot be obtained at some moments, therefore, the information processing device is also used to find the front photo that matches the illegal photo within the time threshold, Then from one of the violation photos, a rear photo with a license plate image and a rear image is extracted to replace the front photo, and the rear photo and the violation photo are combined into a violation record. That is to say, if the second camera responsible for taking photos of the front of the car breaks down and the photos of the front of the car cannot be obtained, in order not to affect the normal record of violations, in an alternative solution, the photos of the rear of the car in the photos of violations can also be Continue to use it as a substitute photo to ensure the stability of system services.

In addition to the above-mentioned one abnormal situation, due to the influence of many objective factors in practical applications, the system may encounter various problems. For example, when some vehicles do not have a license plate, or when the license plate is blocked or damaged, the license plate information in the image acquired by the system is incomplete, and this type of image cannot be used for vehicle violation records, so the information described in the present invention When the processing device finds that there is an image with incomplete vehicle license plate information in the images in the cache, it will mark the vehicle image with incomplete vehicle information as an abnormal image and save it in the cache, or delete it. In this way, the interference of abnormal images can be solved, and useless images can be deleted to save cache space.

Please refer to the accompanying drawing 4 to describe the embodiment of the present invention in detail below.

Shown in the figure is a one-way lane, and a stop line is set at the front end of the lane, and there are traffic lights and pedestrian crossings in front of the stop line, and vehicles drive from left to right. The system of the present invention is set on the gantry, wherein the snap camera a is the second camera, and the snap camera b is the first camera, assuming that the information processing device is set on the system platform connected to the snap camera superior.

When a driver drives a vehicle with a license plate of "ABC" from the left end of the lane to position 1, the capture camera a will capture the front of the vehicle due to receiving a trigger signal, and obtain the front photo a1 of the vehicle at 12:01 :01, and the license plate number "ABC" is sent to the cache of the system platform; the vehicle continues to move to the right, and when it travels to position 2, the snap camera b receives a trigger signal to snap a photo of the rear of the vehicle b1, The time is 12:01:03, and the license plate number "ABC" is also placed on the cache of the system platform. When the vehicle continues to drive, the snap camera b will also snap a picture of the vehicle located at the stop line (position three) and when driving out of the stop line. (Position 4) photos of the rear of the car (assumed to be b2 and b3 respectively, and the time is 12:01:05 and 12:01:07 respectively), the photos b2 and b3 and their shooting time and license plate information also exist in the system In the cache of the platform; wherein the rear photo also includes the information of the traffic light, assuming that the light is green at this time, the vehicle whose license plate is "ABC" driven by the driver a is judged by the snap camera b, and there is no violation.

If on the same day, another driver two drives the same license plate as "ABC" vehicle on this lane, as shown in Figure 5. According to the above-mentioned capture process, when the vehicle travels to the position one, the capture camera a captures the front photo a2, and the shooting time is 18:01:01; when the vehicle travels to the position two, the capture camera b Take a snapshot to obtain the rear photo b4 of the vehicle at 18:01:03; similarly, when the vehicle continues to drive, the snap camera b will also capture the rear photos at positions three and four respectively (Assume they are b5 and b6 respectively, and the times are 18:01:05 and 18:01:07 respectively). Assuming that it is a red light at this time, then the snap camera b will analyze that the vehicle whose license plate is "ABC" has a violation of running a red light at 18:01:04 according to the rear photos b4, b5, b6, and then the violation will be Photos b4, b5, and b6 are sent to the cache of the system platform.

Therefore, the system platform cache includes the above-mentioned images a2, b4, b5, b6, and a1, as well as the capture time of the images, the violation time of the vehicle, and the license plate information. The information processing device in the system platform finds the front photos a1 and a2 corresponding to the license plate information corresponding to the illegal photo according to the car number information. Violation Because a vehicle was found to have a violation of the red light at 18:01:04, assuming that the preset threshold time is 3 minutes, then it is only necessary to analyze the photos taken between 17:58:04 and 18:04:04 . Therefore, the information processing device can analyze that the front photo a2 is a passing image, and according to the analysis of the profile picture information in the passing image, determine that the violating driver is driver 2. Then extract the driver's profile picture in a2 and the violation photos (b4, b5, b6) to synthesize the violation record. After the synthesis is successful, the traffic control department will punish the driver two accordingly according to the violation records synthesized by this system (as shown in Figure 6).

It can be seen that the present invention can distinguish the violations of different drivers of the same vehicle. Compared with the prior art, the present invention can obtain the front photo of the car by adding a second camera, thereby obtaining the head portrait of the driver, so that the record of violations can be accurate to the driver, making up for the fact that there is no head portrait of the driver in the record of violations of the traditional electric police system. Insufficient, enhance the comprehensiveness of traffic management.

The above descriptions are only preferred embodiments of the present invention, and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention shall be included in the present invention. within the scope of protection.

Claims (10)

1. A red light running capture system, characterized in that: comprising a first camera, a second camera and an information processing device; The second camera is used to capture a picture of the front of the vehicle when the vehicle enters the capture area, recognize the license plate information according to the picture of the front of the vehicle, and send the picture of the front of the vehicle, the license plate information and the time of capture to the buffer of the information processing device; The first camera is used to identify whether the vehicle violates the regulations, and if it violates the regulations, it recognizes the license plate information according to the illegal photos including the rear image of the vehicle, and sends the illegal photos, the illegal license plate information and the violation time to the information processing device; The shooting direction of the first camera is opposite to that of the second camera; The information processing device is configured to pass the license plate information of the violating vehicle after receiving the photo of the violation, the license plate information of the violation and the time of violation sent by the first camera. Searching for a vehicle front photo that matches the violation photo within the capture time of the threshold, and merging the violation photo and the found vehicle front photo into a violation record. 2. The red light capture system as claimed in claim 1, characterized in that: in the found photo of the head of the vehicle matched with the photo of the violation, the photo of the head of the car with the driver's head portrait and the license plate is enlarged, and combined with the photo of the violation of regulations to form A violation record. 3. The red light running capture system according to claim 1, characterized in that: the information processing device is set in the first camera or in the front-end processing terminal or on the platform server. 4. The system for snapping photos of running a red light according to claim 1, characterized in that: there are three photos of violations, including photos of the vehicle body not crossing the stop line, photos of the vehicle body having crossed the stop line, and the vehicle body facing the traffic light after crossing the stop line Continue driving photos, and the traffic light images are also included in the violation photos. 5. A method for snapping photos of running a red light according to claim 1, characterized in that: if the time difference between the time of violation and the snapping time of the photo of the front of the car is less than a threshold, no matching front of the car is found within the snapping time photo, then extract the rear photo with the rear image and license plate information from the violation photo, and combine the rear photo and the violation photo into a violation record. 6. A method for snapping photos of running a red light, comprising the steps of: The second camera captures a picture of the front of the vehicle when the vehicle enters the capture area, recognizes the license plate information according to the picture of the front of the vehicle, and sends the picture of the front of the vehicle, the license plate information and the time of capture to the cache of the information processing device; The first camera identifies whether the vehicle violates the regulations, if it violates the regulations, it recognizes the license plate information according to the violation photos containing the rear image, and sends the violation photos, violation license plate information and violation time to the information processing device cache; the first camera Opposite to the shooting direction of the second camera; After receiving the illegal photo, license plate information and time of violation sent by the first camera, the information processing device uses the license plate information of the violating vehicle to search within the time when the time difference between the time of violation and the snapshot time of the front photo is less than a threshold value. The photo of the head of the vehicle matched with the photo of the violation, and the photo of the violation of the regulations and the photo of the head of the car found are combined into one violation record. 7. The method for snapping photos of running a red light according to claim 6, characterized in that: cut out and amplify the front photo with the driver's head portrait and license plate at the front photo matched with the illegal photo, and merge with the illegal photo to form A violation record. 8. The method according to claim 6, characterized in that: the information processing device is set in the first camera or in the front-end processing terminal or on the platform server. 9. The method for snapping photos of running a red light according to claim 6, characterized in that: there are three photos of the violation, which respectively include a photo of the vehicle body not crossing the stop line, a photo of the vehicle body having crossed the stop line, and a photo of the vehicle body facing the traffic light after crossing the stop line Continue driving photos, and the traffic light images are also included in the violation photos. 10. The snapping method for running a red light as claimed in claim 6, characterized in that: if the time difference between the violation time and the snapping time of the front photo is less than a threshold, no front photo matching the illegal photo is found within the snapping time, Then extract the rear photo with the rear image and license plate information from the violation photo, and combine the rear photo and the violation photo into a violation record.