CN204884166U - Regional violating regulations parking monitoring devices is stopped to traffic taboo - Google Patents

Abstract

The utility model discloses a monitoring device for illegal parking in a traffic prohibited area, which comprises a high-definition camera, a supplementary light, an industrial control computer, network equipment, and a computer monitoring device; Above, the industrial control computer is set and installed in the power supply box beside the road. The high-definition camera, supplementary light and the industrial control computer are electrically connected through the network cable. The industrial control computer is electrically connected with the computer monitoring device through the network equipment. The data captured by the camera is used to recognize the license plate of illegal vehicles at the same time; the beneficial effect is that, due to the use of industrial control computers, the interference of trees, pedestrians and other motor vehicles can be effectively filtered out, and the anti-noise performance is good; in addition, the optimized image can be simultaneously Monitor the illegal parking behavior in the yellow grid line area of the four lanes, with high recognition rate, accuracy rate and real-time performance.

Description

A monitoring device for illegal parking in traffic prohibited areas

technical field

The utility model relates to a monitoring device for traffic violation parking, in particular to a monitoring device for traffic violation parking within yellow grid lines.

Background technique

Traffic violation parking behavior is one of the main causes of traffic accidents and traffic congestion, and it is also one of the key points of traffic management and punishment. The yellow grid line is a kind of ground traffic indicator line, which is used to mark the area where parking for any reason is prohibited, and it can be set at intersections, entrances and exits, and other necessary positions where temporary parking is likely to cause congestion, as required. Yellow grid lines are drawn at the exits and entrances of some important units such as administrative agencies, hospitals, schools, bus hubs, and residential areas. When the parties drive through the yellow grid lines, they must not park their vehicles on the grid lines for any reason. (including parking and waiting for traffic lights), so as not to affect the passage of vehicles and cause traffic jams. In response to this situation, the traffic management department has planned yellow grid line markings in the central part of the road intersection. The traffic control department in some areas stipulates that vehicles should not stay in the yellow grid line area for more than 10 seconds, so as not to affect traffic. Normal driving of vehicles in other directions. In order to avoid traffic jams and potential safety hazards caused by traffic jams, it is very necessary to establish an automatic monitoring device for traffic violations and parking in the yellow grid line area. The illegal parking monitoring method based on video surveillance has many advantages such as high efficiency, good real-time performance, low cost, and easy collection of evidence. At present, the target extraction algorithm based on video surveillance does not filter out pedestrians and other motor vehicles, which largely leads to an increase in the false alarm rate.

Contents of the invention

The technical problem to be solved by the utility model is to provide a monitoring device for illegal parking in a traffic parking prohibited area.

The technical scheme adopted by the utility model is a monitoring device for illegal parking in a traffic prohibited area, including a high-definition camera, a supplementary light, an industrial control computer, network equipment and a computer monitoring device; the high-definition camera and supplementary light The lights are set and installed on the gantry of the monitoring road section, the industrial control computer is set and installed in the power supply box beside the road, the high-definition camera, the supplementary light and the industrial control computer are electrically connected through the network cable, and the industrial control computer is connected through the network. The equipment is electrically connected to the computer monitoring device through a network cable, and the industrial control computer receives the shooting data of the high-definition camera, and simultaneously performs license plate recognition of illegal vehicles.

The network equipment described is an optical transceiver and a network; vehicles on four lanes can be monitored by using one high-definition camera.

The beneficial effect of the utility model is that, due to the use of industrial control computers, the interference of trees, pedestrians and other motor vehicles can be effectively filtered out, and the anti-noise performance is good; in addition, the optimized image can monitor the yellow grid lines of four lanes at the same time The illegal parking behavior in the area has high recognition rate, accuracy rate and real-time performance.

Description of drawings

Fig. 1 is a block diagram of the utility model illegal parking monitoring device;

Fig. 2 is a composition block diagram of the utility model illegal parking monitoring software system;

Fig. 3 is a schematic diagram of the utility model using background difference to detect vehicle targets;

Fig. 4 is the schematic diagram based on image matching in the process of eliminating image shaking of the utility model;

Fig. 5a, 5b are the flow charts of the vehicle violation identification algorithm of the present invention;

Fig. 6 is a flow chart of the matching tracking algorithm of the present invention;

Fig. 7 is a flow chart of the utility model establishing an array of suspected violations.

In the picture:

10. HD camera 20. Fill light

30. Industrial control computer 40. Network equipment

50. Computer monitoring device.

Detailed ways

Below in conjunction with accompanying drawing and specific embodiment the utility model is described in further detail:

As shown in Figure 1, the traffic violation parking monitoring device in the yellow grid line area of the utility model includes a high-definition camera 10, a supplementary light 20, an industrial control computer 30, a network device 40 and a computer monitoring device 50; the high-definition camera 10 and supplementary light 20 are set and installed on the gantry of the monitoring road section, and the industrial control computer 30 is set and installed in the power supply box beside the road. The industrial control computer 30 is electrically connected with the computer monitoring device 50 through the network equipment 40 through the network cable. The industrial control computer 30 receives the shooting data of the high-definition camera 10, and at the same time performs the license plate recognition of the violating vehicle. The high-definition camera 10 includes shooting a distant view and a close view, the distant view takes the entire yellow grid line area as the main scene, and the close view includes a front or rear image of a license plate for a motor vehicle that violates regulations. The focal length of the camera should be adjusted to meet the clear shooting of the area at the border of the yellow grid, so as to clearly capture the license plate of the vehicle; ensure that the field of view of the surveillance can cover the driving lane. The network device 40 is an optical transceiver and a network; a high-definition camera 10 can be used to monitor the vehicles on the four lanes.

Fig. 2 is the software composition block diagram of monitoring device for parking violations of the present invention, as shown in Fig. 2, the software that described computer monitoring device 50 adopts mainly comprises, camera control module, parking violation detection function module and partition function module; Wherein, The camera control module collects the real-time pictures of the location to be detected, and transmits the image data to the display card and memory of the industrial control computer 30, and can directly display the image on the screen or display the image on the computer screen through a bitmap function. The illegal parking detection function module processes the collected images, including background extraction and update, image segmentation and recognition, and illegal detection and recording. Partition function module, which divides any one of the pictures into blocks, and each block will be used as a subsequent image processing and recognition unit, which can realize simultaneous detection of four lanes and multiple areas at the same time; it also includes the general control process module, picture encryption and annotation functions module and violation record generation module.

The overall control process module is mainly used to call the control and illegal parking detection function modules of the high-definition camera 10 during working hours, and detects the capacity change of the target hard disk in real time, and clears outdated data files in time if the hard disk capacity is insufficient.

The picture partition function module mainly completes the picture partition operation. Divide each picture into a certain number of rectangular blocks, and each block will be used as a subsequent image processing and recognition unit. This scheme can not only effectively improve the recognition efficiency, but also realize simultaneous detection of multiple areas in four lanes.

The image encryption and annotation function module mainly completes the invisible watermark encryption and annotation of time and place on the final accessed violation records to prevent the violation records from being tampered with later. For the font part at the top, the program automatically reads the location in the configuration file and generates a Hanzi dot matrix in TXT format.

The method for realizing the illegal parking monitoring device within the yellow grid line includes the following steps,

A, background extraction and update, use the histogram method to extract the background of a series of moving image frames;

B. Carry out image segmentation and recognition,

Perform grayscale processing on the image in A, that is, only grayscale images are processed;

The background frame is subtracted from the current frame to obtain the moving object area, and the moving vehicle and the existing shadow area are marked with a rectangular frame;

Convert the difference map to a binary map;

Use the binary morphology algorithm to remove 8-connected regions with less than 800 white pixels;

C. To detect and record violations,

Perform pattern matching on the target according to B to obtain the pattern matching area;

When the next image is input, judge the change of the pixels in the pattern matching area. If the number of changes in the pixels of the target vehicle is less than 30%, it is determined that the vehicle is in a stopped state, and the vehicle information is added to the violation linked list array, and the counter is updated; when the time counter counts When T<10, if the position of the vehicle moves, clear the violation file information of the vehicle; when the time counter counts T>10 seconds, set the violation flag to 1, record the violation time, mark the violation vehicle, and finally clear the violation list record ; When the vehicle is about to leave the area, start the close-range view of the camera, capture the vehicle license plate, and save the dynamic video and close-range pictures of the 15-second illegal parking of the vehicle in the hard disk.

Use the histogram method to extract the background of a series of moving image frames. When a motor vehicle is driving on a normal road, it is unlikely that a certain point on the road surface in the image will be covered by the motor vehicle for a long time, and the possibility of different brightness is high. Therefore, count the number of occurrences of different brightnesses on each pixel within a period of time, and the brightness value with the most occurrences, that is, the maximum value in the histogram is the brightness value of the pixel itself on the road surface; the histogram display is intuitive and easy to analyze .

Fig. 3 is a schematic diagram of the utility model for detecting vehicle targets by using background difference. Let B _i be the image background, f _i be the current frame image, and the difference image be D _i , then:

D _i (x,y)＝|f _i (x,y)-B _i-1 (x,y)|

Let R _i be the post-difference binarized image. Carry out connectivity analysis on R _i , when the area of a certain connected area is greater than a certain threshold, it is considered that the detected target appears, and this connected area is considered to be the detected target image.

Among them, T is the set threshold.

First, grayscale processing is performed on the image, that is, only grayscale images are processed. In order to identify and analyze the target, it is necessary to separate and extract the relevant area, and further use the area on this basis. The technology and process of dividing an image into regions with different characteristics and extracting objects of interest is image segmentation. The background difference method of the present utility model is based on motion video, at first, utilizes the histogram method to obtain the background of the motion image, then, subtract the background frame from the current frame to obtain the moving object area, and use a rectangular frame to mark the area of interest, that is, the moving vehicle And possible shadows, in the subsequent calculations, only for this small part of the image, not only intuitive and clear, but also greatly reduce the amount of calculation; then, use the iterative threshold segmentation method to convert the difference image into a binary image; After the binary image is obtained, there are many disturbances on the actual road, including natural scenes (leaves, etc.) and pedestrians, etc. The noise in the binary image has a great influence. For this, the algorithm of binary morphology is used, that is, the number of white pixels is removed Less than 800 8-connected regions to remove noise interference.

Fig. 4 is a schematic diagram of the utility model based on image matching in the process of eliminating image jitter; as shown in Fig. 4 , the image jitter in the vehicle monitoring system will seriously affect the identification and tracking of the vehicle. The utility model utilizes software to eliminate image shaking. In order not to affect the real-time performance of the system, the utility model adopts a method based on image matching, which has high detection accuracy for image sequences that only contain translation and small rotation.

Each frame of color image needs to be converted to grayscale image before dithering. Assume that the number of pixels in the horizontal and vertical directions of each frame of image is S and T respectively. Establish the reference frame coordinate system OXY, the origin is located in the upper left corner of the reference frame, the X axis is horizontal to the right, and the Y axis is vertically downward. The solid box represents the matching area in the reference frame, the dashed box represents the search box, and the search box moves within a certain range to find the matching area. Before starting processing, save the pixel gray value of the matching area in the reference frame and the coordinates (x ₀ , y ₀ ) of the upper left corner to the memory, and set the offset to (δ _x , δ _y ). The initial value is ( 0,0), the movement trend is (β _x , β _y ), when β _x and β _y take +1, they represent rightward and downward movement, respectively, and when they take -1, they represent leftward and upward movement respectively; the initial movement trend The value can be chosen arbitrarily, assuming (+1, +1). After starting the processing, assuming that (δ _x , δ _y ) and (β _x , β _y ) of the matching area in the previous frame have been obtained for any frame of image, the steps to search for the best match are:

① Predict the search starting point according to the offset of the previous frame. Make the coordinates (X, Y) of the upper left corner of the first sub-image in the current frame coordinate system satisfy X=x ₀ +δ _x , Y=y ₀ +δ _y and then calculate its similarity with the matching area according to the absolute difference sum method The value of the degree measure function, that is,

$\mathrm{<span\; class="notranslate">\; D.</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; Y</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 0</span>}}{\overset{\mathrm{<span\; class="notranslate">\; m</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{<span\; class="notranslate">\; \&Sigma;</span>}}\underset{\mathrm{<span\; class="notranslate">\; the\; y</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 0</span>}}{\overset{\mathrm{<span\; class="notranslate">\; N</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{<span\; class="notranslate">\; \&Sigma;</span>}}<span\; class="notranslate">\; |</span>\mathrm{<span\; class="notranslate">\; g</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; the\; y</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; g</span>}}_{\mathrm{<span\; class="notranslate">\; 0</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; the\; y</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; |</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>$

In the formula, (X, Y) is the coordinate of the upper left corner of the sub-image in the current frame coordinate system; D (X, Y) is the similarity measure function value between the sub-image and the matching area; M, N are the matching area level, The number of pixels in the vertical direction; (X, Y) are the coordinates of the corresponding pixels in the sub-image and the matching area in their respective coordinate systems; g(x, y) and g ₀ (x, y) are the coordinates of the sub-image and the matching area, respectively. corresponds to the gray level of the pixel.

② Search horizontally. Y keeps the value of the previous step unchanged, and predicts X of the next subgraph according to β _x , that is, X=X+β _x , and recalculates the value of the measure function according to formula (1). If the value of the measure function becomes smaller or unchanged, it means that the search direction is correct, and the search is continued according to the direction specified by β _x until the value of the measure function is about to increase. If the value of the measure function becomes larger, it means that the search direction is wrong, return to the original position, β _x = -β _x , and search according to the direction specified by the new β _x until the value of the measure function is about to become larger. If it exceeds the search range during the search, it will stop, and the search will be considered as a failure.

③ Search in the vertical direction. Similar to step ②, the difference is that X, Y and β _x

Replaced by Y, X and β _y , respectively.

④ Repeat steps ② and ③ until X and Y do not change. If the number of repetitions exceeds the given value, the search is also considered to have failed. The allowable number of repetitions needs to be comprehensively determined based on the running time and processing effect.

Figure 5a and Figure 5b are the flow charts of the vehicle violation recognition algorithm of the utility model; Position, match in the next frame, and judge whether the match is successful. When the match is successful, the counter starts counting. When it is greater than the set value, it judges the violation and outputs the result; when the counter is less than the set value, it returns to continue matching; when the match is unsuccessful, it returns The change area is cyclically identified by the above method.

After the vehicle is detected in the yellow grid line area, it is judged whether it violates the regulations. After the vehicle is judged to be in violation, the vehicle is tracked, and the vehicle is captured and photographed when it leaves the shooting area.

Fig. 6 is a flow chart of the matching tracking algorithm of the present invention; as shown in Fig. 6, the obtained binarized image is marked to obtain several target areas. Calculate the circumscribed rectangle for each target area, obtain the pattern matching area, and calculate the coordinates of the center point of the rectangle. When the next image is input, judge the change in the number of pixels in the target area. If the change in the number of pixels is less than the threshold, the vehicle is in a stopped state, start timing, and iteratively match. When the counter reaches a certain time, mark the vehicle as violating the rules, and check the vehicle Tracking, when the number of pixel changes in this technical solution is less than 30%, the vehicle is in a stopped state, and when the time counter reaches 10 seconds, set the violation flag to 1, record the violation time, and mark the violation vehicle; violations in the yellow grid line area The monitoring device uses close-up and long-range shooting to work together to track and monitor changes in the center point of the target. If the center point is close to the boundary of the yellow grid, turn on the camera close-up mode to take a close-up view of the front of the vehicle, perform license plate recognition, and save the results as a basis for punishing the vehicle that caused the accident. Violation records include: violation location, violation time, violation type, violation vehicle license plate number, a dynamic panoramic video of violations and close-up images of front or rear license plates.

The high-definition camera 10 is aimed at the yellow grid line area where parking is prohibited at the intersection, and the monitoring image is collected through the camera, and any one of the pictures is divided into blocks by the partition function module, and each block will be used as a subsequent image processing and recognition unit, which can effectively improve Recognition efficiency, and can realize simultaneous detection of multiple areas in four lanes at the same time. Under the partition operation interface, the program automatically generates 7 partitions or 11 partitions. If the sample partition does not meet the requirements, the user can manually rewrite the partition mode, enter the number of partitions, and drag the box to just cover the detection area. After the partition is completed, the program automatically saves the partition information and writes it into the configuration file.

Any one of the pictures is divided into blocks by the partition function module, and each block will be used as a subsequent image processing and recognition unit. Assuming that the number of partitions is set to N, an N-dimensional linked list array R(N) is established in the program, wherein R(i) is the information file of illegal vehicles in the i-th partition, and the target area (rectangular area) is reserved first in the linked list. The location coordinates, pixel mean value and vehicle residence time T of the first occurrence.

The specific process of obtaining the position coordinates and pixel mean value of the original detection area is as follows: first, use the cvFindContours() function provided by the OpenCV library to obtain the contour of the binary image, and then use the cvBoundingRect() function to calculate the "circumscribed rectangle" of the contour , this rectangle is recorded as rect, then the position, length and width of this rectangle are respectively X=rect.x, Y=rect.y, M=rect.width, N=rect.height. Position coordinates Select the coordinates of the center point of the rectangular area, record the position coordinates as (p _x , p _y ), the calculation formula is as in formula 2

$\begin{array}{c}{\mathrm{<span\; class="notranslate">\; p</span>}}_{\mathrm{<span\; class="notranslate">\; x</span>}}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; +</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; *</span>\mathrm{<span\; class="notranslate">\; m</span>}\\ {\mathrm{<span\; class="notranslate">\; p</span>}}_{\mathrm{<span\; class="notranslate">\; the\; y</span>}}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; Y</span>}<span\; class="notranslate">\; +</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; *</span>\mathrm{<span\; class="notranslate">\; N</span>}\end{array}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>$

The calculation of the average value C of pixels in a rectangular area is shown in Equation 3, where g(x, y) is the pixel value at point (x, y).

$\mathrm{<span\; class="notranslate">\; C</span>}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; m</span>}<span\; class="notranslate">\; *</span>\mathrm{<span\; class="notranslate">\; N</span>}}\underset{\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; x</span>}}{\overset{\mathrm{<span\; class="notranslate">\; m</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; x</span>}}{<span\; class="notranslate">\; \&Sigma;</span>}}\underset{\mathrm{<span\; class="notranslate">\; the\; y</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; Y</span>}}{\overset{\mathrm{<span\; class="notranslate">\; N</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; Y</span>}}{<span\; class="notranslate">\; \&Sigma;</span>}}\mathrm{<span\; class="notranslate">\; g</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; the\; y</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 3</span>}<span\; class="notranslate">\; )</span>$

The program loops to calculate the position coordinates and pixel mean value of the target area of each frame image. If the pixel mean value changes less than 30%, the vehicle is considered to be in a stopped state, and the time count T is incremented by 1 until T=10.

When T>10, mark the vehicle and judge it as a violation. When the center point of the vehicle is close to the edge of the yellow grid line area, start the camera to take a close-up photo, and clear the linked list information that retains the vehicle information; carry out license plate recognition, and the license plate information Save it to the static character array set in the program so that the rest of the partitions can be compared to avoid repeated identification; the program automatically generates a folder named after the time and place of the violation and the license plate number of the violation. Copy a specified number of images and pictures of the violation process as the basis for law enforcement afterwards; the program automatically generates a picture composed of four pictures during the vehicle violation process, as shown in Figure 7, which can be used as a post-event verification picture. The four pictures are the close-up picture including the license plate, the picture of the vehicle starting to stop illegally, the picture of confirming the violation (the 10th second), and the picture of the middle moment of the violation.

Because the utility model adopts an industrial control computer, it can effectively filter out the interference of trees, pedestrians and other motor vehicles, and has good anti-noise performance; in addition, the optimized image can simultaneously monitor the illegal parking behavior in the yellow grid line area of four lanes , High recognition rate, accuracy rate and real-time performance.

Claims (2)

1. A monitoring device for violating parking in traffic no-parking areas, comprising, a high-definition video camera (10), a fill light (20), an industrial control computer (30), a network device (40) and a computer monitoring device (50) ); the high-definition camera (10) and supplementary light (20) are set and installed on the gantry of the monitoring road section, and the industrial control computer (30) is set and installed in the power supply box by the roadside, and the high-definition camera (10 ), the fill light (20) is electrically connected with the industrial control computer (30) through the network cable, and the industrial control computer (30) is electrically connected with the computer monitoring device (50) through the network equipment (40), and the industrial control computer (30) is electrically connected with the computer monitoring device (50) through the network cable. The computer (30) receives the shooting data of the high-definition camera (10), and simultaneously performs license plate recognition of illegal vehicles. 2. The illegal parking monitoring device in a traffic no-stop area according to claim 1, wherein said network equipment (40) is an optical transceiver and a network; a high-definition camera (10) can be used to monitor four lanes Vehicles.