WO2020107306A1

WO2020107306A1 - Intelligent collection system for information of vehicle running red light and method therefor

Info

Publication number: WO2020107306A1
Application number: PCT/CN2018/118095
Authority: WO
Inventors: 高峡; 高月仁; 谷明静; 耿华; 吕晓栓; 李罡; 杨海涛; 康海凤; 李海新; 成景坤; 史守帆; 祁志雷; 马桂泽; 王鹏
Original assignee: 唐山曹妃甸联城科技有限公司
Priority date: 2018-11-29
Filing date: 2018-11-29
Publication date: 2020-06-04

Abstract

An intelligent collection system for information of a vehicle running a red light and a method therefor, which are used for collecting information of a vehicle that runs a red light at an intersection corresponding to a signal lamp; the collection system comprises an induction coil set (3), a signal lamp status detection device (4), an image detection device (5) and a DSP processor (8); the induction coil set (3) is used for sensing whether a vehicle is crossing a lane of an intersection that corresponds to a signal lamp; the signal lamp state detection device (4) is used for detecting the current state of the signal lamp; according to the sensing result of the induction coil set (3) and the detection result of the signal lamp state detection device (4), the DSP processor (8) controls image collecting and image processing operations of the image detection device (5) so as to obtain information of a vehicle running a red light. The collection system and the method therefor further comprise using a system for collection.

Description

Intelligent information collection system and method for vehicle running through red light

Technical field

The invention relates to the field of traffic supervision, and in particular to a method for intelligently collecting vehicle information through red lights.

Background technique

In order to realize the effective detection, punishment and maintenance of traffic safety of vehicle violations, a large number of illegal vehicle information collection systems are installed at various important traffic intersections to be applied to road traffic management. However, the information collection system for vehicles running through red lights in the prior art has the following limitations: First, emergency vehicles such as ambulances and fire trucks cannot be identified in real time, and these emergency vehicles are still photographed through red lights, wasting limited system resources. It brings interference to the investigation and punishment of red light violations; second, it is generally through the analysis of an intersection image to judge the detection of red light violations of multiple lanes at the same time. In fact, the detection of the vehicle information of the currently allowed lanes is meaningless and the system is also wasted. Resources.

Therefore, there is a need for a method for intelligently collecting information on vehicles running through red lights, which can improve the utilization rate of electronic resources of the information collecting system for vehicles running through red lights, and complete interference removal of emergency vehicles at the shooting site in real time to ensure the accuracy of information collection. At the same time, the intelligence level and working efficiency of the collection system are improved.

Summary of the invention

In order to solve the above-mentioned problems, the present invention provides a method for intelligently collecting information on vehicles running through red lights, using pre-stored image templates of emergency passing vehicles and image matching technology to remove the interference of vehicles passing through vehicles passing through red lights in emergency, and detecting the current red light in real time. On the basis of the light lane, the parallel image processing technology is used to abandon the violation detection of the green light lane, and only information collection is carried out on the illegal vehicles in the red light lane, which makes the information collection of the red light vehicle of the present invention more targeted.

According to one aspect of the present invention, there is provided an intelligent red light vehicle information collection system for collecting red light vehicle information at intersections corresponding to signal lights, the collection system includes an induction coil set, a signal light state detection device, an image detection device, and DSP processing The induction coil set is used to sense whether a vehicle passing through the traffic lane corresponding to the signal light is detected, the signal light state detection device is used to detect the current state of the signal light, and the DSP processor is based on the induction result and the position of the induction coil set The detection result of the signal light state detection device, and controlling the image acquisition and image processing operations of the image detection device to obtain the information of the vehicle running through the red light;

The invention also provides a method for intelligently collecting information on vehicles running through red lights. More specifically, in the intelligently collecting information system on vehicles running through red lights, it further includes: a set of oscillation circuits, including a plurality of oscillation circuits, the number of oscillation circuits and signal lights The number of lanes at the corresponding intersection is the same. Multiple oscillation circuits correspond to multiple induction coils in the induction coil set. Each oscillation circuit is connected to the corresponding induction coil. When a vehicle passes a predetermined distance in front of the parking line of the lane where the corresponding induction coil is located, The mutual induction between the corresponding induction coil and the vehicle body causes the oscillation frequency of the oscillation circuit to change. On the contrary, when no vehicle passes, the oscillation frequency of the oscillation circuit remains unchanged; the frequency detection circuit is connected to a plurality of oscillation circuits of the oscillation circuit set, Detect whether the oscillation frequency of each oscillating circuit has changed, and send out the lane number of the induction coil corresponding to the oscillating circuit whose oscillation frequency has changed, the number sequence of the lane number is from left to right; the storage device, use In advance, two emergency passing vehicle image templates are stored in advance, the emergency passing vehicle image template is an image pre-photographed for a reference emergency passing vehicle, the reference emergency passing vehicle is a reference ambulance or a reference fire truck, and the storage device is also in advance The upper limit gray threshold of the lane reference line, the lower limit gray threshold of the lane reference line, the upper limit gray threshold of the vehicle, and the lower limit gray threshold of the vehicle are stored. The upper limit gray threshold of the lane reference line and the lower limit gray threshold of the lane reference line are used To separate the lane reference line and the background in the image, the upper vehicle grayscale threshold and the lower vehicle grayscale threshold are used to separate the vehicle from the background in the image; the wireless communication device is used to receive one or more red lights Vehicle information and send the one or more red light vehicle information to the server of the local traffic management platform through a wireless communication link, so that the server's hard disk stores the one or more red light vehicle information; the induction coil set Including multiple induction coils, the number of induction coils is the same as the number of lanes corresponding to the intersection of signal lights, multiple induction coils correspond to multiple lanes one by one, each induction coil is buried under a predetermined distance in front of the parking lane of the corresponding lane; The signal light state detection device is respectively connected to a plurality of signal light units of the signal light, the number of the signal light units is the same as the number of lanes of the corresponding intersection of the signal light, the plurality of signal light units correspond to the plurality of lanes one by one, the signal light state detection device is based on the plurality of signal light units Whether it is currently a red light outputting the lane number of the lane corresponding to the signal light unit currently being red light; the DSP processor and the frequency detection circuit, the signal light state detection device, the image detection device and the wireless communication device Connect respectively to find the lane numbers with the same value output by both the frequency detection circuit and the signal light state detection device as the same lane number and send them to the image detection device, and the DSP processor also receives the output of the image detection device. One or more red light vehicle information, and forward the one or more red light vehicle information to the wireless communication device; the image detection device It includes an image acquisition unit, a vehicle type recognition unit, a wavelet filter unit, a grayscale processing unit, an image segmentation unit, and multiple parallel image processing units. The number of parallel image processing units is the same as the number of lanes at the intersection corresponding to the signal light. Multiple parallel images The serial number of the processing unit corresponds one-to-one with multiple lanes. The image acquisition unit is connected to the DSP processor and is used to capture an image of the intersection at a predetermined distance in front of the corresponding intersection of the signal light when the same lane number is received. The wavelet filter unit is The image acquisition unit is connected to perform wavelet filtering processing on the intersection image based on a Haar wavelet filter to obtain a filtered image, and the vehicle type identification unit is connected to the wavelet filter unit and the storage device, Match the vehicle pattern in the filtered image with the two emergency passing vehicle image templates one by one. When there is an emergency passing vehicle image template that matches successfully, the emergency passing vehicle signal is output; otherwise, the non-emergency passing vehicle signal is output. The processing unit is connected to the vehicle type identification unit, and when the non-emergency passing vehicle signal is received, the filtered image is gray-scaled to obtain a gray-scale image. The image segmentation unit and the gray-scale processing The unit, the wavelet filtering unit and the storage device are respectively connected to identify pixels in the grayscale image between the upper grayscale threshold of the lane reference line and the lower grayscale threshold of the lane reference line And form multiple lane reference line sub-images, and divide the filtered image into multiple lane filter sub-images based on the positions of the multiple lane reference line sub-images in the grayscale image, the number of the lane filter sub-images The number of lanes corresponding to the intersection of the signal light is the same, multiple lane filter sub-images correspond to multiple lanes, and each parallel image processing unit is connected to the image segmentation unit, the storage device, and the DSP processor, each The parallel image processing unit exits the power saving mode to start parallel image processing when the corresponding lane number exists in the received same lane number, and enters the power saving mode when the corresponding lane number does not exist in the received same lane number Mode, the parallel image processing includes receiving a lane filtering sub-image corresponding to a lane, and identifying pixel values of the received lane filtering sub-image between the vehicle upper gray threshold and the vehicle lower gray threshold And form a vehicle sub-image, identify the license plate information in the vehicle sub-image, and combine the lane number of the corresponding lane and the recognized license plate information to output as red light vehicle information; wherein, the wireless communication link is GPRS mobile communication One of a link, a 3G mobile communication link, or a 4G mobile communication link.

More specifically, in the intelligent red light vehicle information collection system: the image collection unit includes a CMOS vision sensor.

More specifically, in the intelligent red light vehicle information collection system: the image collection unit further includes a lens and a filter, the filter is located in front of the lens, and the lens is located in front of the CMOS vision sensor.

More specifically, in the red light vehicle information intelligent collection system: the CMOS vision sensor is a high-definition vision sensor, and the resolution adopted is 1920×1080.

More specifically, in the intelligent red light vehicle information collection system, the collection system further includes a brightness detection device for detecting the ambient brightness of the surrounding environment of the collection system; an auxiliary lighting source is located on the image detection device. Above the image acquisition unit, it is connected to the brightness detection device, and determines whether to provide auxiliary lighting for shooting of the image acquisition unit based on the ambient brightness.

BRIEF DESCRIPTION

In order to more clearly explain the technical solutions of the embodiments of the present invention, the drawings required for the description of the embodiments will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present invention. Those of ordinary skill in the art can obtain other drawings based on these drawings without creative efforts.

The embodiments of the present invention will be described below with reference to the drawings, in which:

FIG. 1 is a structural block diagram of a red light traffic information intelligent collection system according to an embodiment of the present invention.

2 is a block diagram showing the structure of an image detection device for a red light traffic information intelligent collection system according to an embodiment of the present invention.

detailed description

In order to make the objectives, technical solutions and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention, and are not intended to limit the present invention.

The implementation of the red light traffic information intelligent collection system and method of the present invention will be described in detail below with reference to the drawings.

Red light running refers to the behavior of a motor vehicle crossing a stop line and continuing to drive in violation of the red light no-traffic regulations at signal-controlled intersections and sections.

Running a red light will cause the driver to have the following adverse reactions:

1. When driving at a red light, the driver's attention is on the traffic lights, and he cannot fully and correctly perceive the changes inside and outside the car, which causes difficulties for the driver's response and measures;

2. When running through a red light, the speed of the car usually passes through the intersection is relatively fast, which weakens the driver's cognitive ability of space;

3. When running through a red light, the driver's ability to judge various traffic environments is reduced, which sometimes makes the driver's thinking or operation mistakes;

4. The intersection is a meeting point and a conflict point of traffic flow. Running a red light increases the insecurity of the meeting point and the conflict point itself, resulting in a decrease in the safety factor;

5. Running a red light affects the driver's timely, accurate and effective control of the vehicle;

6. Running a red light affects the safe passing performance of vehicles, and also reduces the ability of other traffic participants to respond to and respond to strains;

7. Practice has proved that 90% of heavy traffic accidents occur at intersections, so the consequences of accidents caused by running red lights are usually more serious;

8. Under traffic lights, if the driver's unconsciousness appears, it is extremely harmful.

In order to suppress the violation of the red light and avoid serious traffic accidents, it is necessary to obtain real-time information on the red light vehicles to severely punish the holders of the red light vehicles. The existing red light vehicle information collection system cannot remove the interference caused by emergency vehicles, the system redundancy is too high, and the work efficiency is average.

The present invention builds an intelligent information collection system for vehicles running through red lights. The collection system is built on the crossbar where the signal lights are located and is used to collect information about vehicles running through red lights corresponding to the intersections of the signal lights. Multiple image processing technologies such as filtering, image graying processing and image segmentation complete the collection of vehicle information of running red lights efficiently and reasonably, and the structure of the entire collection system is more compact.

1 is a structural block diagram of an intelligent collection system for red light traffic vehicles according to an embodiment of the present invention. The collection system is disposed on a crossbar where a signal light is located, and is used to collect information about a red light traffic vehicle at an intersection corresponding to the signal lights. Including frequency detection circuit 1, oscillation circuit set 2, induction coil set 3, signal status detection device 4, image detection device 5, storage device 6, wireless communication device 7 and DSP processor 8, DSP processor 8 and frequency detection circuit 1 The signal light state detection device 4, the image detection device 5, the storage device 6, and the wireless communication device 7 are respectively connected, and the oscillation circuit set 2 is connected to the frequency detection circuit 1 and the induction coil set 3, respectively.

Among them, the induction coil set 3 is used to sense whether a vehicle passing through the lane corresponding to the signal light is detected, the signal light state detection device 4 is used to detect the current state of the signal light, and the DSP processor 8 is based on the induction coil set 3 The sensing result and the detection result of the signal light state detection device 4 control the image acquisition and image processing operations of the image detection device 5 to obtain the red light vehicle information.

Next, the specific structure of the red light vehicle information intelligent collection system of the present invention will be further explained.

The oscillating circuit set 2 includes a plurality of oscillating circuits, the number of oscillating circuits is the same as the number of lanes corresponding to the intersections of the signal lights, the oscillating circuits correspond one-to-one with the plurality of induction coils in the induction coil set 3, and each oscillating circuit is connected Induction coil, when a vehicle passes a predetermined distance in front of the parking line where the corresponding induction coil is located, the corresponding induction coil and the vehicle body have a mutual inductance, causing the oscillation frequency of the oscillation circuit to change. Conversely, when no vehicle passes, the oscillation frequency of the oscillation circuit remains constant.

The frequency detection circuit 1 is connected to a plurality of oscillation circuits of the oscillation circuit set 2 respectively, detects whether the oscillation frequency of each oscillation circuit has changed, and sends out the lane number of the induction coil corresponding to the oscillation circuit whose oscillation frequency has changed, The numbering sequence of the lane numbers is from left to right.

The storage device 6 is used to pre-store two emergency passing vehicle image templates, the emergency passing vehicle image template being an image pre-photographed for a reference emergency passing vehicle, the reference emergency passing vehicle being a reference ambulance or a reference fire truck, the The storage device 6 also pre-stores the lane reference line upper limit gray threshold, the lane reference line lower limit gray threshold, the vehicle upper limit gray threshold and the vehicle lower limit gray threshold, the lane reference line upper limit gray threshold and the lane reference line The lower limit grayscale threshold is used to separate the lane reference line and the background in the image, and the upper vehicle grayscale threshold and the lower vehicle grayscale threshold are used to separate the vehicle and the background in the image.

The wireless communication device 7 is used to receive one or more red light vehicle information and send the one or more red light vehicle information to a server of a local traffic management platform through a wireless communication link, so that the hard disk of the server stores the one Or multiple vehicles running through red lights.

The induction coil set 3 includes a plurality of induction coils. The number of induction coils is the same as the number of lanes corresponding to the intersections of the signal lights. The plurality of induction coils correspond to the plurality of lanes one by one. Each induction coil is buried in front of the parking lane of the corresponding lane. Below the distance.

The signal light state detection device 4 is respectively connected to a plurality of signal light units of the signal light. The number of the signal light units is the same as the number of lanes of the corresponding intersection of the signal light. The plurality of signal light units correspond to the plurality of lanes. The signal light state detection device 4 is based on Whether the plurality of signal light units are currently red lights outputs the lane number of the lane corresponding to the signal light unit currently being red lights.

The DSP processor 8 is connected to the frequency detection circuit 1, the signal light state detection device 4, the image detection device 5, and the wireless communication device 7 respectively, looking for the frequency detection circuit 1 and the signal light state The lane numbers with the same value output by the detection devices 4 are sent to the image detection device 5 as the same lane numbers. The DSP processor 8 also receives one or more red light vehicle information output by the image detection device 5 and sends The one or more red light vehicle information is forwarded to the wireless communication device 7.

As shown in FIG. 2, the image detection device 5 includes an image acquisition unit 51, a vehicle type recognition unit 53, a wavelet filter unit 52, a gradation processing unit 54, an image segmentation unit 56, and a plurality of parallel image processing units 55. There are n parallel image processing units, and n is the same as the number of lanes at the intersection corresponding to the signal light. Multiple parallel image processing units 55 have serial numbers corresponding to multiple lanes.

In the image detection device 5, the image acquisition unit 51 is connected to the DSP processor, and is used to take an image of a junction at a predetermined distance in front of the intersection corresponding to the signal light when the same lane number is received, and the wavelet filter unit 52 is The image acquisition unit 51 is connected to perform wavelet filtering processing on the intersection image based on a Haar wavelet filter to obtain a filtered image, the vehicle type recognition unit 53 and the wavelet filter unit 52 and the storage device Connect respectively, and match the vehicle pattern in the filtered image with the two emergency passing vehicle image templates one by one. When there is a matching emergency passing vehicle image template, the emergency passing vehicle signal is output; otherwise, the non-emergency passing vehicle signal is output. The grayscale processing unit 54 is connected to the vehicle type identification unit 53, and when the non-emergency passing vehicle signal is received, the filtered image is grayscaled to obtain a grayscale image, and the image segmentation unit 56 is The gradation processing unit 54, the wavelet filtering unit 52 and the storage device are respectively connected to connect the gradation value in the gradation image to the upper limit gradation threshold of the lane reference line and the lower limit of the lane reference line The pixels between the gray thresholds identify and compose multiple lane reference line sub-images, and divide the filtered image into multiple lane filter sub-images based on the positions of the multiple lane reference line sub-images in the gray-scale image, The number of the lane filtering sub-images is the same as the number of lanes corresponding to the intersections of the signal lights, and the multiple lane filtering sub-images correspond to the multiple lanes one-to-one.

Each parallel image processing unit 55 is connected to the image segmentation unit 56, the storage device, and the DSP processor. When each parallel image processing unit 55 receives the corresponding lane number in the same lane number, Exit the power saving mode to start parallel image processing, and enter the power saving mode when the corresponding lane number does not exist in the same lane number received. The parallel image processing includes receiving the lane filtering sub-image of the corresponding lane, which will be received The pixels in the lane filter sub-image whose gray value is between the upper gray threshold of the vehicle and the lower gray threshold of the vehicle identify and form a vehicle sub-image, identify the license plate information in the vehicle sub-image, and will correspond The lane number of the lane and the recognized license plate information are combined and output as the red light vehicle information.

Wherein, the wireless communication link is one of GPRS mobile communication link, 3G mobile communication link or 4G mobile communication link.

The image acquisition unit 51 may include a CMOS vision sensor, and the image acquisition unit 51 may further include a lens and a filter. The filter is located in front of the lens, and the lens is located in front of the CMOS vision sensor. The CMOS vision sensor is a high-definition vision sensor with a resolution of 1920×1080; the collection system may further include a brightness detection device for detecting the ambient brightness of the surrounding environment of the collection system, including an auxiliary lighting source, located at the Above the image acquisition unit 51 of the image detection device 5, it is connected to the brightness detection device, and determines whether to provide auxiliary lighting for shooting of the image acquisition unit 51 based on the ambient brightness.

In addition, CMOS (Complementary Metal-Oxide-Semiconductor), Chinese name is complementary metal oxide semiconductor, he was originally an important chip in the computer system, which saved the most basic information of the system guide. The manufacturing technology of CMOS is no different from that of general computer chips. It mainly uses semiconductors made of silicon and germanium to coexist on the CMOS with N (charge-charged) and P (charged + charge) grades. In semiconductors, the current generated by these two complementary effects can be recorded and interpreted as an image by the processing chip. Later, it was discovered that CMOS can also be used as an image sensor in digital photography after processing.

For portable applications independent of the power grid, CMOS technology, known for its low power consumption, has a clear advantage: CMOS image sensors are designed for 5V and 3.3V supply voltages. The CCD chip requires a power supply voltage of about 12V, so a voltage converter has to be used, resulting in increased power consumption. In terms of total power consumption, the integration of control and system functions into the CMOS sensor will bring another benefit: he removes all external connections to other semiconductor components. Its high-power drivers are now deprecated, because the energy consumed for communication inside the chip is much lower than the external implementation through the PCB or substrate.

CMOS sensors can also be subdivided into passive pixel sensors (Passive Pixel Sensor CMOS) and active pixel sensors (Active Pixel Sensor CMOS).

Passive pixel sensor (PPS), also called passive pixel sensor, is composed of a reverse-biased photodiode and a switch tube. The photodiode is essentially a PN junction composed of a P-type semiconductor and an N-type semiconductor. It can be equivalent to a reverse biased diode and a MOS capacitor in parallel. When the switch tube is turned on, the photodiode is connected to the vertical column line (Column bus). The charge integration amplifier at the end of the column line maintains the column line voltage at a constant value. When the signal charge stored in the photodiode is read out, its voltage is reset to the column line voltage level, and at the same time The charge proportional to the optical signal is converted into a charge output by the charge integration amplifier.

Active pixel sensor (Active Pixel Sensor, APS for short) is also called active pixel sensor. Almost at the same time that the CMOS PPS pixel structure was invented, people quickly realized that introducing a buffer or amplifier into the pixel can improve the performance of the pixel. In CMOS APS, each pixel has its own amplifier. The amplifying transistor integrated on the surface reduces the effective surface area of the pixel element and reduces the "packing density", so that 40% to 50% of incident light is reflected. Another problem with this type of sensor is how to better match the multi-channel amplifiers of the sensor, which can be better achieved by reducing the residual pattern of fixed pattern noise. Since each amplifier in the CMOS APS pixel is only activated during this readout, the power consumption of the CMOS APS is smaller than that of the CCD image sensor.

By adopting the intelligent red light vehicle information collection system of the present invention, in view of the technical problems that the existing red light vehicle information collection system may erroneously collect and the system is not compact, through a series of image processing, emergency traffic vehicles such as ambulances that allow temporary red light traffic to pass are avoided Vehicles and fire engines were collected by mistake. At the same time, vehicle information was collected only for the red light channel, which improved the reliability and intelligence of the system.

It can be understood that although the present invention has been disclosed as the preferred embodiments above, the above embodiments are not intended to limit the present invention. For anyone skilled in the art, without departing from the scope of the technical solutions of the present invention, the technical contents disclosed above can be used to make many possible changes and modifications to the technical solutions of the present invention, or to be modified to equivalent changes, etc.实实施例。 Effective embodiments. Therefore, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention without departing from the technical solution of the present invention still fall within the protection scope of the technical solution of the present invention.

Claims

A red light vehicle information intelligent collection system for collecting red light vehicle information of intersections corresponding to signal lights, characterized by comprising an induction coil set, a signal light state detection device, an image detection device and a DSP processor, the induction coil set is used for Whether there is a vehicle passing through the lane corresponding to the induction signal light, the signal light state detection device is used to detect the current state of the signal light, the DSP processor is based on the induction result of the induction coil set and the detection result of the signal light state detection device, The image acquisition and image processing operations of the image detection device are controlled to obtain red light vehicle information.
The intelligent collection system for red light vehicle information according to claim 1, wherein the collection system further comprises:

The oscillating circuit set includes a plurality of oscillating circuits, the number of oscillating circuits is the same as the number of lanes corresponding to the intersections of the signal lights, the oscillating circuits correspond one-to-one with the plurality of induction coils in the induction coil set, and each oscillation circuit is connected to the corresponding induction coil;

The frequency detection circuit is connected to a plurality of oscillation circuits of the oscillation circuit set, detects whether the oscillation frequency of each oscillation circuit has changed, and sends out the lane number of the induction coil corresponding to the oscillation circuit whose oscillation frequency has changed. The numbering sequence of the lane numbers is from left to right;

The storage device is used to pre-store two emergency passing vehicle image templates, and the emergency passing vehicle image template is an image pre-photographed for a reference emergency passing vehicle, and the reference emergency passing vehicle is a reference ambulance or a reference fire engine, the The storage device also pre-stores the lane reference line upper limit gray threshold, the lane reference line lower limit gray threshold, the vehicle upper limit gray threshold and the vehicle lower limit gray threshold, the lane reference line upper limit gray threshold and the lane reference line lower limit The grayscale threshold is used to separate the lane reference line and the background in the image, and the upper vehicle grayscale threshold and the lower vehicle grayscale threshold are used to separate the vehicle from the background in the image;

A wireless communication device, configured to receive one or more red-lighting vehicle information and send the one or more red-lighting vehicle information to a server of a local traffic management platform through a wireless communication link, so that the hard disk of the server stores the one Or multiple vehicles running through red lights.
The system and method for intelligently collecting red light vehicle information according to claim 1, wherein the induction coil set includes a plurality of induction coils, and the number of induction coils is the same as the number of lanes corresponding to the intersections of the signal lights. The coils correspond to multiple lanes one by one. Each induction coil is buried under a predetermined distance in front of the parking lane of the corresponding lane. When a vehicle passes the predetermined distance in front of the lane of the lane where the corresponding induction coil is located, the corresponding induction coil and the vehicle body Mutual inductance causes the oscillation frequency of the oscillation circuit to change. On the contrary, when no vehicle passes, the oscillation frequency of the oscillation circuit remains unchanged.
A method for intelligently collecting information about vehicles running through red lights, characterized in that the signal light state detection device is connected to a plurality of signal light units of the signal light, the number of the signal light units is the same as the number of lanes corresponding to the intersections of the signal lights, In one correspondence, the signal light state detection device outputs the lane number of the lane corresponding to the signal light unit which is currently red light according to whether the plurality of signal light units are currently red lights;

The DSP processor is respectively connected to the frequency detection circuit, the signal light state detection device, the image detection device and the wireless communication device, and finds that the frequency detection circuit and the signal light state detection device both output the same value The lane number is sent to the image detection device as the same lane number, and the DSP processor also receives one or more red light vehicle information output by the image detection device, and forwards the one or more red light vehicle information to the The wireless communication device; the image detection device includes an image acquisition unit, a vehicle type recognition unit, a wavelet filter unit, a grayscale processing unit, an image segmentation unit and a plurality of parallel image processing units, the number of parallel image processing units corresponding to the signal lights The number of lanes at the intersection is the same, and the serial numbers of multiple parallel image processing units correspond to multiple lanes one-to-one;

The image acquisition unit is connected to the DSP processor, and is used to take an image of the intersection at a predetermined distance in front of the intersection corresponding to the signal light when the same lane number is received;

The wavelet filtering unit is connected to the image acquisition unit, and is configured to perform wavelet filtering processing on the intersection image based on a Haar wavelet filter to obtain a filtered image;

The vehicle type identification unit is connected to the wavelet filter unit and the storage device respectively, and the vehicle pattern in the filtered image is matched with the two emergency passing vehicle image templates one by one, and there is an emergency passing vehicle image that matches successfully When the template is output, the emergency passing vehicle signal is output; otherwise, the non-emergency passing vehicle signal is output. The graying processing unit is connected to the vehicle type identification unit, and the filtered image is grayed when the non-emergency passing vehicle signal is received. Processing to obtain grayscale images;

The image segmentation unit is connected to the grayscale processing unit, the wavelet filtering unit and the storage device respectively, and the grayscale value in the grayscale image is at the upper limit grayscale threshold of the lane reference line and the The pixels between the lower limit grayscale threshold of the lane reference line identify and compose multiple lane reference line sub-images, and divide the filtered image into multiple lanes based on the positions of the multiple lane reference line sub-images in the grayscale image Filtered sub-images, the number of the lane-filtered sub-images is the same as the number of lanes at the intersections corresponding to the traffic lights, multiple lane-filtered sub-images correspond to multiple lanes, each parallel image processing unit and the image segmentation unit, The storage device is connected to the DSP processor, and each parallel image processing unit exits the power saving mode to start parallel image processing when the corresponding lane number exists in the same lane number received in the same lane number received When there is no corresponding lane number, enter the power saving mode, the parallel image processing includes receiving the lane filtering sub-image of the corresponding lane, and the gray value of the received lane filtering sub-image is above the vehicle upper limit gray threshold and Pixels between the lower gray threshold of the vehicle identify and form a vehicle sub-image, identify the license plate information in the vehicle sub-image, and combine the lane number of the corresponding lane with the identified license plate information to output as red light vehicle information;

Wherein, the wireless communication link is one of GPRS mobile communication link, 3G mobile communication link or 4G mobile communication link.