CN110660222B - Intelligent environment-friendly electronic snapshot system for black-smoke road vehicle - Google Patents
Intelligent environment-friendly electronic snapshot system for black-smoke road vehicle Download PDFInfo
- Publication number
- CN110660222B CN110660222B CN201911060713.6A CN201911060713A CN110660222B CN 110660222 B CN110660222 B CN 110660222B CN 201911060713 A CN201911060713 A CN 201911060713A CN 110660222 B CN110660222 B CN 110660222B
- Authority
- CN
- China
- Prior art keywords
- vehicle
- black smoke
- license plate
- camera
- smoke
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 239000000779 smoke Substances 0.000 title claims abstract description 200
- 238000012544 monitoring process Methods 0.000 claims abstract description 89
- 238000001514 detection method Methods 0.000 claims abstract description 48
- 230000001502 supplementing effect Effects 0.000 claims abstract description 44
- QVFWZNCVPCJQOP-UHFFFAOYSA-N chloralodol Chemical compound CC(O)(C)CC(C)OC(O)C(Cl)(Cl)Cl QVFWZNCVPCJQOP-UHFFFAOYSA-N 0.000 claims abstract description 30
- 238000011161 development Methods 0.000 claims abstract description 29
- 238000004891 communication Methods 0.000 claims abstract description 11
- 230000006855 networking Effects 0.000 claims abstract description 9
- 238000000034 method Methods 0.000 claims description 72
- 230000003287 optical effect Effects 0.000 claims description 38
- 230000008569 process Effects 0.000 claims description 24
- 230000033001 locomotion Effects 0.000 claims description 18
- 238000007781 pre-processing Methods 0.000 claims description 12
- 239000013307 optical fiber Substances 0.000 claims description 11
- 238000012549 training Methods 0.000 claims description 11
- 238000012545 processing Methods 0.000 claims description 10
- 230000006870 function Effects 0.000 claims description 9
- 230000003993 interaction Effects 0.000 claims description 9
- 238000004364 calculation method Methods 0.000 claims description 8
- 238000012552 review Methods 0.000 claims description 8
- 230000008859 change Effects 0.000 claims description 7
- 230000009466 transformation Effects 0.000 claims description 7
- OKTJSMMVPCPJKN-UHFFFAOYSA-N Carbon Chemical compound [C] OKTJSMMVPCPJKN-UHFFFAOYSA-N 0.000 claims description 6
- 238000000605 extraction Methods 0.000 claims description 5
- 238000012423 maintenance Methods 0.000 claims description 4
- 238000007619 statistical method Methods 0.000 claims description 4
- 238000012935 Averaging Methods 0.000 claims description 3
- 238000003708 edge detection Methods 0.000 claims description 3
- 230000007613 environmental effect Effects 0.000 claims description 3
- 238000009432 framing Methods 0.000 claims description 3
- 238000007689 inspection Methods 0.000 claims description 2
- 238000012795 verification Methods 0.000 claims description 2
- 230000001131 transforming effect Effects 0.000 claims 1
- 239000000523 sample Substances 0.000 description 14
- 239000013598 vector Substances 0.000 description 8
- 238000010586 diagram Methods 0.000 description 7
- 238000004422 calculation algorithm Methods 0.000 description 6
- 239000000284 extract Substances 0.000 description 4
- 230000005540 biological transmission Effects 0.000 description 3
- 238000013527 convolutional neural network Methods 0.000 description 3
- 210000003128 head Anatomy 0.000 description 3
- 230000008901 benefit Effects 0.000 description 2
- 125000004122 cyclic group Chemical group 0.000 description 2
- 230000007547 defect Effects 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000010606 normalization Methods 0.000 description 2
- 238000000926 separation method Methods 0.000 description 2
- 238000012360 testing method Methods 0.000 description 2
- 230000006978 adaptation Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000002238 attenuated effect Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 235000019504 cigarettes Nutrition 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 230000008451 emotion Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000003384 imaging method Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000001737 promoting effect Effects 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 230000001960 triggered effect Effects 0.000 description 1
Abstract
The invention relates to an intelligent environment-friendly electronic snapshot system for a black-smoke road vehicle, which comprises a high-definition license plate recognition camera, a stroboscopic light supplementing lamp, an environment-friendly monitoring center, a high-brightness light supplementing lamp, a night full-color monitoring camera, an environment detector, a photoelectric smoke measuring telescope and a LORA development board; the cameras of the high-definition license plate recognition cameras are opposite to the head of an incoming vehicle, the cameras of the night full-color monitoring cameras and the lenses of the photoelectric smoke measuring telescope face the tail of the vehicle, the environment detector is arranged around the two cameras, the night full-color monitoring cameras are connected with the LORA development board, and the LORA development boards of the night full-color monitoring cameras of all roads are constructed to form a local area network to realize networking communication among the cameras; the environment-friendly monitoring center comprises a terminal monitoring host and an upper computer, and the LORA development board can receive data transmitted by the terminal monitoring host. The system can detect the blackness of the vehicle tail gas under different natural environments, effectively improves the blackness detection accuracy of the black smoke vehicle tail gas, and improves the supervision efficiency.
Description
Technical Field
The invention relates to a road black smoke vehicle snapshot system, in particular to an intelligent environment-friendly electronic road black smoke vehicle snapshot system.
Background
In recent years, as the number of automobiles continues to increase, the problem of exhaust emission from automobiles has been receiving more and more attention. And many old diesel vehicles such as farm vehicles, trucks and the like generate more black smoke during starting. However, in the process of supervising the black-smoke vehicle, since such vehicles generally travel at a relatively late time, a great deal of manpower is required. Even with a monitoring system, it is difficult to accurately detect a black smoke vehicle due to light. In this case, intelligent black smoke vehicle recognition can be performed according to different light environments (daytime, evening, sunny day, cloudy day, etc.) in the daytime, and the recognition becomes a hot spot for the development of domestic and foreign enterprises.
At present, the black smoke vehicle monitoring system has some problems, and under different weather conditions, the detection of the black smoke vehicle is affected to a certain extent, and the detection is also affected by factors such as emotion, fatigue degree and the like of people only through manual monitoring. At present, the detection of the black smoke vehicle at night is immature and incomplete, and an effective means for monitoring the black smoke vehicle is lacked, so that the black smoke vehicle cannot be monitored well. Meanwhile, the monitoring of the road black smoke vehicle only relates to the identification of the black smoke at the tail part of the vehicle, the positions of smoke outlets arranged at the two sides and the top end of the vehicle (because the smoke outlets of some vehicles are arranged between the head part and the hopper or exhaust tail gas from the bottom of the vehicle to the two sides) cannot be identified, the testing effect is limited by the environment, the speed of the vehicle and the operation level of technicians, and the testing precision is required to be improved.
The black smoke electronic snapshot system which is online at the present stage has the following defects that: the first and current-stage black smoke electronic snapshot system can identify black smoke vehicles, but has a certain error such as omission in detection of low-concentration black smoke. The second and black smoke electronic snapshot system can record and track black smoke vehicles, but in practice, the distance between cameras is generally far greater than the measurable range, a complete system is not formed between the cameras, so that the cameras repeatedly record the black smoke vehicles, or the mutual communication between the cameras is realized by using optical fibers, and the storage space and funds are wasted due to high optical fiber laying cost.
Disclosure of Invention
Aiming at the defects of the prior art, the invention aims to provide an intelligent environment-friendly electronic snapshot system for a black smoke road vehicle. The black smoke vehicle electronic snapshot system can detect the blackness of vehicle tail gas under different natural environments, particularly, the blackness detection accuracy of the black smoke vehicle tail gas is effectively improved aiming at the black smoke vehicle tail gas with low concentration, the black smoke vehicle is found as accurately as possible through detection, the replacement of the black smoke vehicle is accelerated by combining related departments, the pollution of the black smoke of the motor vehicle to the atmosphere is reduced, the use of human resources and cost is reduced, and the supervision efficiency is improved. The networking of the camera is realized, and the current situations of long evidence obtaining time, high speed, long moving distance and great difficulty in law enforcement and evidence obtaining of a black smoke vehicle are simultaneously solved; the license plate information of the detected vehicle is automatically identified in an intelligent sensing mode, whether black smoke is emitted from the vehicle is identified, and once the black smoke vehicle is detected by the system, automatic video recording and snapshot are carried out.
In order to solve the technical problems, the invention adopts the technical scheme that the intelligent environment-friendly electronic snapshot system for the black smoke road vehicle comprises a high-definition license plate recognition camera, a stroboscopic light supplementing lamp and an environment-friendly monitoring center, and is characterized by further comprising a high-brightness light supplementing lamp, a night full-color monitoring camera, an environment detector, a photoelectric smoke measuring telescope and a LORA development board; the high-definition license plate recognition camera, the stroboscopic light supplementing lamp, the highlighting light supplementing lamp, the night full-color camera, the environment detector, the photoelectric smoke measuring telescope and the LORA development board are all arranged on a portal frame of a road, the camera of the high-definition license plate recognition camera is opposite to the head of an incoming vehicle, the camera of the night full-color monitoring camera and the lens of the photoelectric smoke measuring telescope face to the tail of the vehicle, the environment detector is arranged around the two cameras, the night full-color monitoring camera is connected with the LORA development board, and the LORA development boards of the night full-color monitoring cameras of all roads are constructed to form a local area network to realize networking communication among the cameras; the environment-friendly monitoring center comprises a terminal monitoring host and an upper computer, and the LORA development board can receive data transmitted by the terminal monitoring host;
the high-definition license plate recognition camera and the stroboscopic light supplementing lamp are combined to shoot license plate information and the face of a driver,
The night full-color monitoring camera and the high-brightness light supplementing lamp are combined to detect black smoke around the vehicle;
The environment detector detects real-time environment change and is used for adjusting the light supplementing brightness and acquiring the influence of the environment on tail gas;
The high-definition license plate recognition camera, the stroboscopic light supplementing lamp, the night full-color monitoring camera, the high-definition light supplementing lamp and the environment detector form a front-end snapshot subsystem;
the collected information of the high-definition license plate recognition camera, the stroboscopic light supplementing lamp, the night full-color monitoring camera, the high-definition light supplementing lamp and the environment detector is respectively transmitted to the terminal monitoring host through the switch and the optical fiber transceiver which are connected, the terminal monitoring host is used for carrying out data processing and storage on the collected information, the specific data processing and storage process comprises vehicle recognition, black smoke area recognition, black smoke blackness detection and vehicle tracking and positioning,
And the terminal monitoring host transmits the processed data to the upper computer, and the upper computer observes the blackness level of the vehicle tail gas acquired by the front-end snapshot subsystem and stores license plate information of the corresponding vehicle with the blackness not reaching the standard.
Compared with the prior art, the invention has the beneficial effects that:
the invention has the substantial characteristics that:
According to the invention, when the blackness of the tail gas of the black smoke vehicle is identified, the black smoke area is detected by adopting a mode of combining an optical flow method and an optical attenuation method, the blackness of the black smoke area is detected by utilizing the photoelectric smoke measuring telescope, and the grade of black smoke is determined by comparing with an electronic ringelman table in the photoelectric smoke measuring telescope, so that the detection accuracy is effectively improved under the condition of low black smoke concentration; on the other hand, in the innovation of networking cameras, a local area network is built among the cameras, the data transmission among the cameras is carried out by using a local area network mode, the information of the detected black smoke vehicles is shared into the cameras which are mutually connected, only the position information of the vehicles is provided for the reappeared detected vehicles, the black smoke area and the grade judgment are not carried out any more, the detection time is reduced when the vehicles are identified as the black smoke vehicles, and meanwhile, the running interval of the black smoke vehicles is determined.
The remarkable progress of the invention is:
a) Compared with the training method of the traditional convolutional neural network, the cascade classifier training used in vehicle identification can better extract the characteristics of the vehicle. Because convolutional neural networks compress full-size pictures into one vector containing classifications during learning, spatial information is lost during vector expansion during vehicle recognition. The cascade classifier extracts and classifies all the characteristics of the vehicle, then matches the characteristics when the vehicle is identified, and can be determined as the vehicle when all the characteristic classifications are satisfied, thereby improving the robustness of vehicle identification.
B) And improving the black smoke detection accuracy. The existing black smoke detection method mainly adopts an optical flow method, but the movement direction of black smoke will be influenced first in windy environment, and then the detection error is caused. Therefore, the photoelectric smoke measuring telescope is overlapped with the optical flow method, so that not only is the movement direction and the area of black smoke detected, but also the black smoke is imaged on the photoelectric probe, and the detection is performed through electronic ringeman image comparison. The detection accuracy of the system is improved through the combination of the image and the optical flow method.
C) The invention not only can provide the tracking and positioning functions of the existing cameras, but also realizes the information communication among the cameras through the local area network in a mode of constructing the local area network by LORA, and for the determined black-smoke vehicles, the black-smoke vehicle information is mutually 'informed' through the information transmission among the cameras, and the running direction of the black-smoke vehicles and the positioning of the cameras are combined, so that the running route of the black-smoke vehicles is better tracked, and the traffic management department is helped to stop the black-smoke vehicles from going on the road as soon as possible. The problem that in the prior art, dead zones exist in the distance between cameras (the distance between the cameras in the urban road is larger than the shooting distance by which the cameras can shoot clearly, so that the shot dead zones exist between the two cameras) so that vehicles are repeatedly detected and the processing time and the memory resource are wasted is avoided.
D) The front-end snapshot subsystem adopts a night full-color monitoring camera (black light), mimics the human eye imaging principle, adopts a double-photosensitive element (sensor) framework, is responsible for collecting color information, is responsible for collecting brightness information, is simultaneously perceived in visible light and infrared light, has greatly improved photosensitive capacity, can ensure full-color image quality under a low-light environment, realizes accurate identification of black smoke, license plates and the like, and solves the problems that a traditional monitoring camera cannot clearly monitor road black smoke vehicles at night and cannot normally identify black smoke. Compared with the combination of the camera and the light supplementing lamp in the current stage, the camera used by the invention is more applicable in a low-light environment.
E) The black smoke vehicle monitoring system can meet the requirement of 24-hour uninterrupted monitoring, can clearly shoot samples at night, changes the condition that the traditional monitoring is poor in shooting quality at night, and improves the quality of black smoke vehicle detection.
F) The platform has high processing efficiency and high running speed, and the number of the traversed pixels is reduced in the process of detecting black smoke and license plates, so that the traversed pixels have pertinence, and the algorithm can embody the advantage of real-time performance in the running process.
G) And the license plate recognition can be carried out, and for vehicles with black smoke exceeding standards, the system can also rapidly track and shoot the vehicles, so that more sufficient evidence is provided for repairing the black smoke vehicles.
H) The client interface is clear in classification and complete in information, so that a vehicle owner can inquire vehicle information in real time, and a user interaction function can be provided.
Drawings
FIG. 1 is a flow chart of a hardware system of the present invention.
FIG. 2 is a schematic flow chart of the method of the present invention.
Fig. 3 is a background separation diagram of the invention, wherein the left side of the diagram is a diagram of an unseparated snap vehicle, the middle of the diagram is a background diagram after separating the background, and the right side of the diagram is a foreground diagram after separating the background.
Fig. 4 is a host vehicle identification image.
FIG. 5 is an optical flow field of the exhaust of the host vehicle.
Fig. 6 is a vehicle exhaust gas map of the present invention.
Fig. 7 is a license plate extraction image of the present invention.
Fig. 8 is a black smoke vehicle measurement result panel according to the present invention.
Fig. 9 is an interface of the black smoke vehicle electronic snapshot system of the present invention.
Detailed Description
The present application is further explained below with reference to examples and drawings, but is not limited thereto.
The intelligent environment-friendly electronic snapshot system for the black-smoke road vehicle comprises a high-definition license plate recognition camera, a stroboscopic light supplementing lamp, a highlight light supplementing lamp, a full-color night monitoring camera, an environment detector, a photoelectric smoke measuring telescope, a LORA development board and an environment-friendly monitoring center; the high-definition license plate recognition camera, the stroboscopic light supplementing lamp, the highlighting light supplementing lamp, the night full-color camera, the environment detector, the photoelectric smoke measuring telescope and the LORA development board are all arranged on a portal frame of a road, the camera of the high-definition license plate recognition camera is opposite to the head of an incoming vehicle, the camera of the night full-color monitoring camera and the lens of the photoelectric smoke measuring telescope face to the tail of the vehicle, the environment detector is arranged around the two cameras, the night full-color monitoring camera is connected with the LORA development board, and the LORA development boards of the night full-color monitoring cameras of all roads are constructed to form a local area network to realize networking communication among the cameras; the environment-friendly monitoring center comprises a terminal monitoring host and an upper computer, and the LORA development board can receive data transmitted by the terminal monitoring host;
the high-definition license plate recognition camera and the stroboscopic light supplementing lamp are combined to shoot license plate information and the face of a driver,
The night full-color monitoring camera and the high-brightness light supplementing lamp are combined to detect black smoke around the vehicle;
The environment detector detects real-time environment change and is used for adjusting the light supplementing brightness and acquiring the influence of the environment on tail gas;
The high-definition license plate recognition camera, the stroboscopic light supplementing lamp, the night full-color monitoring camera, the high-definition light supplementing lamp and the environment detector form a front-end snapshot subsystem;
the collected information of the high-definition license plate recognition camera, the stroboscopic light supplementing lamp, the night full-color monitoring camera, the high-definition light supplementing lamp and the environment detector is respectively transmitted to the terminal monitoring host through the switch and the optical fiber transceiver which are connected, the terminal monitoring host is used for carrying out data processing and storage on the collected information, the specific data processing and storage process comprises vehicle recognition, black smoke area recognition, black smoke blackness detection and vehicle tracking and positioning,
The terminal monitoring host transmits the processed data to the upper computer, and the upper computer observes the blackness level of the vehicle tail gas collected by the front-end snapshot subsystem and stores license plate information of the corresponding vehicle with the blackness not reaching the standard;
The upper computer interface is provided with a video monitoring window, a bayonet snapshot window, a manual review sending window, a review sending notice management, a blacklist management, a comprehensive query, a statistical analysis, equipment and network management, data interaction, system setting and help;
the video monitoring window has two functions of real-time monitoring and video playback window, so that real-time snapshot of a road can be met, and the shot video can be played back.
The bayonet snap window records images of all vehicles passing through the highway lane, recognizes license plates of the vehicles in real time, and uploads the license plates to the data center. The recognition result can be exported for sharing with public security and traffic police.
The manual review sending window can be used for manually uploading the corresponding video to detect again under the condition that the intelligent snapshot result is disagreeable, and the result verification is completed.
The inspection notification management is to associate the black smoke car subjected to manual review with the information of car owners, contact phones, running certificates, administrative areas to which registration and registration belong and the like in the environment protection department, and generate a file of a motor vehicle tail gas detection notification or a motor vehicle black smoke emission punishment notification.
The blacklist management can perform centralized management on all blacklist vehicles, search is performed by inputting license plate numbers and snapshot times in a search field, and search results are displayed in a search result field. The method supports that according to the information of the vehicle owners, a driver archive management page can be entered to manage drivers of the blacklisted vehicles.
The comprehensive inquiry window function comprises a punishment identification and punishment mode specified by the country about the black smoke car.
The statistical analysis window supports the condition of displaying the snapshot accuracy rate in a cake pattern mode and the condition of rechecking the black smoke car by selecting monitoring points and statistical time ranges which need to be counted. The method supports the display of the proportion of the black smoke car to the yellow standard car and the proportion of the black smoke car to the green standard car in a cake pattern.
The equipment and network management window supports configuration and maintenance of the equipment information of the snapshot point and remote information maintenance of the equipment. And supporting the online networking condition of each device displayed on the GIS map.
The data interaction window can realize the data interaction between the snapshot system and the traffic control department and the data interaction between the snapshot system and the vehicle owner.
The intelligent road black smoke vehicle electronic snapshot system collects detection targets through the intelligent perceived front-end snapshot subsystem, and then transmits collected information to a terminal monitoring host through an accessed switch and an optical fiber transceiver. In this process, the terminal monitoring host performs data processing and storage. The processed data are transmitted to a local area network constructed by an upper computer and an LORA through a core switch, the blackness level of the vehicle tail gas collected by a front-end snapshot subsystem is observed in the upper computer, an alarm is triggered after the blackness of the vehicle tail gas reaches a certain threshold value, and the storage equipment stores corresponding information such as vehicle license plates and the like so as to provide basis for later auditing management.
The vehicle identification process comprises the following steps of;
in a video shot by a high-definition license plate recognition camera, haar feature extraction is carried out on each frame of image containing a complete vehicle, training of a cascade classifier is carried out aiming at the extracted features, and recognition of a vehicle sample is carried out through the trained vehicle features;
meanwhile, each frame of image in the video shot by the high-definition license plate recognition camera is subjected to vehicle shadow detection (the shadow detection process can be directly realized by adopting the existing algorithm), the vehicle shadow detection result and the vehicle sample recognition result are subjected to logical AND operation, the recognized vehicle is subjected to rectangular marking, and a rectangular area marked by the vehicle is recorded; the accuracy of vehicle identification can be improved by adopting two detection results to carry out logical AND operation.
The cascade classifier is a deep learning method, and can directly extract the characteristics of positive and negative samples (positive samples are pictures of vehicles and negative samples are pictures without vehicles) to generate files, and then directly call the files, so that the running time of a program is reduced. The phenomenon that information is lost or parameters are more and overfitting possibly occurs through a convolutional neural network is avoided.
The black smoke region identification process comprises the following steps:
1) Marking a rectangular area marked by a vehicle in a video image shot by a full-color monitoring camera at night after the vehicle is identified in the video shot by the high-definition license plate identification camera, circularly traversing around the vehicle by taking the marked vehicle as the center in the video shot by the full-color monitoring camera at night, and stopping traversing when the traversed length of a sub-window is twice the length of the rectangular marked vehicle;
2) In the traversed pixels, calculating the motion trail of the vehicle tail gas by a gradient-based optical flow method, finding a black smoke motion area, and automatically threshold-dividing the black smoke motion area to divide the black smoke motion area into vehicle tail gas black smoke areas;
3) Meanwhile, the light attenuation condition of the night full-color monitoring camera under the environment of the visible light of the image is detected by a light attenuation method, and the area where the light is absorbed is marked as a black smoke area;
4) And superposing the black smoke areas marked by the optical flow method and the optical attenuation method to obtain the target black smoke area.
Further, in the cyclic traversal method mentioned in step 1), the center pixel of the rectangular frame of the marked vehicle is used as the origin, the expansion is performed to 8 surrounding pixel points, the value of the expanded pixel point is extracted, and the calculation of the optical flow method is performed. And then traversing values of 72 pixels around the calculated 9 pixel points serving as centers and calculating by an optical flow method. The loop is repeated until the loop is traversed until the number of pixels reaches twice the length of the vehicle marking rectangle.
Further, the gradient-based optical flow method is to assign a velocity vector initial value to each pixel point of the image, so as to form an image motion field, and at a specific moment of object motion, points on the image correspond to points of the three-dimensional object one by one. When an object in the image moves relatively, the velocity vector in the image changes relatively, and a moving area image of the optical flow is formed.
The light attenuation method is to detect the attenuation area of the emitted light, namely the black smoke area by utilizing the principle that the light can be attenuated in the black smoke.
The combination of the gradient-based optical flow method and the optical attenuation method used in the step can effectively improve the detection stability of the black smoke vehicle tail gas under the condition of low concentration, and the combination of the photoelectric smoke measuring telescope can accurately measure the blackness of the tail gas, so that the invention is the innovation and improvement of the detection method.
The vector recognition engine based on the optical flow field analyzes black smoke through the consistency of motion vectors, meanwhile, the high-definition license plate recognition camera adopts RGB visible light to carry out license plate recognition, the night full-color monitoring camera adopts white light to carry out light supplementing shooting, so that the definition of night videos is ensured, various interference conditions of light change, leaf shadow shaking, automobile shadow change and mutual shielding of automobiles are accurately recognized, dynamic information such as square running, speed and the like of the automobiles is output, license plate recognition shooting functions in different directions are realized, and the black smoke recognition efficiency can be improved by about 20%.
The black smoke blackness detection is used for measuring the black smoke grade of vehicle tail gas with black smoke, and the specific process is as follows:
The blackness detection of the tail gas of the vehicle is carried out on the marked area by the optical flow method and the optical attenuation method by utilizing the photoelectric smoke measuring telescope, the blackness calculation is carried out on the tail gas of the black smoke vehicle by utilizing the photoelectric smoke measuring telescope through an internal self-contained algorithm, and then the blackness level is determined by utilizing the ringeman blackness level standard;
the video image of the black smoke area calculated by the optical flow method is subjected to image preprocessing, and the specific image preprocessing process is as follows:
converting the color space of each frame of image, and drawing a corresponding gray level histogram; dividing the background and the foreground according to the peak value of the histogram; performing operations such as edge detection, histogram equalization and the like on the image to obtain better image information;
Extracting pixel values after image preprocessing, wherein the level of the ringelman blackness corresponding to the pixel values is the blackness level of black cigarettes;
And determining the blackness level of the tail gas by averaging the blackness level corresponding to the detection result of the photoelectric smoke measuring telescope and the blackness level result detected by the pixel value calculation method.
The license plate recognition comprises the following specific processes:
Calling a video image shot by a high-definition license plate recognition camera after marking the vehicle marking rectangle, searching and framing license plates from the vehicle marking rectangle of the video image in a color matching and area matching mode;
Intercepting the license plate part selected by the frame, and preprocessing license plate images of the intercepted part; and the identified vehicle is used as a vehicle with the standard exceeding black smoke level and is uploaded to an upper computer interface.
The preprocessing process of the license plate image comprises the following steps: the license plate is changed into a front view which is convenient for matching the license plate through perspective transformation; dividing characters on a license plate; and (3) recognizing characters in a template matching mode, and finally completing recognition of license plates.
According to the perspective transformation, the license plate part obtained by cutting the rectangular area of the vehicle mark is transformed into the front view angle, so that characters on the license plate can be more clearly seen, the characters can be conveniently matched, and the recognition accuracy of the license plate is improved.
Comparing the blackness level of the current vehicle determined according to the blackness detection of the black smoke with a standard blackness threshold value, and if the existing blackness level is higher than the standard blackness threshold value, indicating that the current vehicle is a tail gas black smoke exceeding vehicle; and recording the license plate of the current exceeding vehicle, and uploading the black smoke grade, license plate information and related vehicle information of the vehicle to an upper computer interface.
The vehicle tracking and positioning is used for tracking and recording vehicles with black smoke exceeding standards and updating data among cameras, and the specific process is as follows;
When the detected black smoke vehicle is detected to be out of standard, the information of the vehicle is transmitted to the upper computer in an optical fiber mode for displaying and alarming, clear out-of-standard vehicle information and position information are provided for traffic departments, and the information is transmitted to the next path of camera in a local area network mode, so that when the next path of high-definition license plate recognition camera recognizes the transmitted out-of-standard vehicle information, the tail gas black smoke region recognition and black smoke blackness detection process is skipped, the relevant information (including the position, license plate, blackness level, driver information and the like of the vehicle) of the out-of-standard vehicle is directly transmitted to the upper computer through the optical fiber, and the vehicle information is also transmitted to the next path of camera through the local area network; the repeatedly-appearing black smoke vehicles provide position information, and the position information is transmitted to an environment-friendly monitoring center in real time, so that tracking and video recording of the black smoke exceeding vehicles are realized;
the local area network is connected with the LORA development board through each camera (namely the camera of the night full-color monitoring camera or the camera of both the night full-color monitoring camera and the high-definition license plate recognition camera), and the local area network is connected with each other through a given communication protocol between the development boards, so that the network communication between the cameras is realized.
The invention has the innovation points that each single camera is built into a simple camera network, the position of the repeatedly appeared black smoke vehicle is locked, the timely management of the traffic management department is more convenient, and the tracking video recording of the locked black smoke vehicle can provide more basis for improving the accuracy of the system.
Examples
The related hardware used by the intelligent environment-friendly road black smoke vehicle electronic snapshot system of the embodiment comprises: 700 ten thousand pixels high definition license plate recognition cameras, night full color monitoring cameras, terminal monitoring hosts, stroboscopic light supplementing lamps, high brightness light supplementing lamps, QT210B type photoelectric smoke measuring telescope and RA-02SX1278 LORA development board; installing the set of hardware on a portal frame for monitoring the road at intervals of a planned distance; in this embodiment, the high-definition license plate recognition camera and the night full-color monitoring camera on each portal frame are connected together to form a LORA development board, and the LORA development boards on the road are connected with each other through a predetermined communication protocol to form a LORA network, so that networking communication between the cameras is realized.
Step one, hardware layout of a front-end snapshot subsystem:
the 700 ten thousand pixel high-definition license plate recognition camera and the stroboscopic light supplementing lamp are used as a combination to shoot license plate information and the face of a driver;
and detecting the tail part and black smoke of the vehicle by using a combination of a night full-color monitoring camera and a high-brightness light supplementing lamp (a black light shield integrated machine).
Detecting real-time environmental changes by using an environmental detector, and adjusting the light supplementing brightness and acquiring the influence of the environment on tail gas;
In order to better find the black smoke vehicle, a road section with a certain gradient and moderate vehicle speed is preferably selected. And the position suggestion of the front end of the electronic snapshot of the black smoke car is uniformly planned according to points, lines and planes.
1) The point: the system comprises large-scale freight transportation and passenger transport vehicles such as a freight transportation logistics center, a bus/bus station access, a dock access, a high-speed toll station and the like;
2) A wire: the method comprises the steps of entering and exiting a main road in a city and controlling the traffic road with emphasis;
3) And (3) surface: according to administrative areas, urban roads which are frequently appeared in other large-scale diesel vehicles are controlled.
Step two, preprocessing an image of the full-color monitoring camera at night:
Processing a vehicle video image shot by a night full-color monitoring camera, and preprocessing the acquired vehicle image to obtain better image information;
Performing color space conversion on the image, converting the RGB image into a Gray space image, and drawing a corresponding Gray histogram;
Determining wave crests and wave troughs according to the gray level images, and reading wave crest values in the histogram;
dividing the background and the foreground according to the peak value of the histogram;
and performing operations such as edge detection, histogram equalization and the like on the image.
The process of converting the color space from the RGB space to the Gray space comprises the following steps:
1) Converting the original image into a gray image through cvtColor;
2) Selecting a plurality of points in the road part to measure gray values;
3) Calculating probability density of each gray level (Nk is the number of gray values of k, n is the total number) and then carrying out gray histogram normalization and equalization on the image;
in this embodiment, the foreground and background separation results are shown in fig. 3.
Step three, identifying the vehicle;
1) In a video shot by a high-definition license plate recognition camera, haar feature extraction is carried out on each frame of image containing a complete vehicle, training of a weak classifier is carried out aiming at the extracted features, then the trained weak classifier is trained into a strong classifier, pixel traversal is carried out in the image when the vehicle passes through the image, the detected region conforming to the features of all the strong classifiers is marked as the vehicle, and recognition of a vehicle sample is realized;
2) And meanwhile, detecting the shadow of the vehicle for each frame of image in the video shot by the high-definition license plate recognition camera: the method comprises the steps of converting an image in an RGB space into a binary image, and selecting a vehicle area through screening the binary image by an automatic threshold value and a trapezoid mask;
3) And the identification detection result of the vehicle sample and the vehicle shadow detection result are mutually overlapped to carry out rectangular marking on the identified vehicle, so that the accuracy of vehicle identification is improved, and the area becomes a vehicle marking area.
The specific steps of vehicle identification are as follows:
1) And extracting features, namely solving the gray level change between pixels by a Haar feature method, and mainly applying linear features and edge features. The calculation formula is as follows: v=sum white-Sum black. Where H is a Haar eigenvalue, sum white and Sum black are the pixel sums of the white and black parts, respectively, in the eigenvalue template.
The training classifier, the Adaboost cascade classifier, is a weak classifier, and the classification of the features in the image is completed by continuously adjusting the weight of the training sample for iteration. The formula for training the classifier is as follows:
a) The given training sample is (x 1,y1)、(x2,y2)……(xn,yn); where x i denotes the i-th sample, y i =0 denotes a negative sample, and y i =1 denotes a positive sample. n is the total number of training samples, i=1 to n.
B) The best weak classifier is obtained. The formula is as followsWherein θ is a set optimal threshold, p is bias, x represents the input picture sub-window, and f j (x) is the value of the jth feature on the x sample; θ j is the classifier optimum threshold of the j-th template; p j is the bias of the j-th template, i.e. the direction of the inequality sign of the j sample, if the classification result is greater than the threshold value, it is-1, otherwise it is +1, so as to ensure that the direction of the inequality sign is unchanged.
C) Initializing the weight of positive and negative samples, wherein the weight of the negative sample is w fi =1\ (2 m), and the weight of the positive sample is w zi =1\ (2L). Where m is the number of negative samples, L is the number of positive samples, i ε (1, n), n is any positive integer, and the total number of samples.
D) Normalization of weights: where i e (1, n), q represents the normalized weight, D represents the weight of the training data, and t represents the number of iterations.
E) Calculating a weighted error rate of the weak classifier: εt= Σ iqi|h(xi,f,p,θ)-yi |, where h (x, f, p, θ) is the weak classifier for each feature, x is the sub-window of the input picture, f is the value of the image feature, θ is the threshold of the classifier.
F) Weight updating, weight reassignment is carried out according to the error rate, and the assigned formula is as follows:
g) Correctly classified: d t+1(i)＝Dt(i)βt, misclassified: d t+1(i)＝Dt (i), wherein β t＝εt/1-εt,βt is a weight update coefficient.
H) Obtaining a strong classifier, wherein the obtained formula is as follows:
Representing a strong classifier, h t (x) is a weak classifier, a i＝㏑((1-εt)/εt).
The image feature classification can be completed through the steps, and then the identification of the vehicle is completed.
When a vehicle passes through the image, pixel traversal is performed in the image, and the lower half of the image is selected as a traversal region when the pixel traversal is performed. Because the camera is fixed in position, it is in the lower half of the image when the vehicle is present within the camera's field of view. Therefore, the number of pixel traversal can be reduced, and the algorithm speed is improved.
The effect of identifying a rectangular image of a vehicle mark in the present invention is shown in fig. 4.
Step four, license plate recognition is carried out on the vehicle and the license plate recognition is uploaded to an upper computer interface;
1) Searching and framing license plates in marked vehicle marked area frames in a color matching and area matching mode;
2) Intercepting a license plate part selected by the frame, and converting the intercepted part into a gray level image;
3) The license plate is changed into a front view which is convenient for matching the license plate through perspective transformation;
4) Dividing characters on a license plate;
5) Character recognition is carried out in a template matching mode, and finally recognition of license plates is completed;
the perspective transformation specifically comprises the following sub-steps:
a) And reading the internal reference and the external reference of the high-definition license plate recognition camera, and eliminating distortion of the read image.
B) And the acquisition frame selects four corner coordinates of the license plate part.
C) The transformation matrix is obtained by a function getPerspectiveTransform, and then the image after the perspective change is obtained by a function WARPPERSPECTIVE.
The template of the character in the invention comprises 34 provinces, short for direct administration, uppercase of 24 English letters and L, O is removed. 10 digits 0-9.
The result of the extraction of the license plate of the vehicle in the present invention is shown in fig. 5.
Fifthly, detecting black smoke areas around the identified vehicles;
1) Marking a rectangular area for marking vehicles in the video image shot by the full-color night monitoring camera after the pretreatment in the step two after the vehicle identification is finished in the video shot by the high-definition license plate identification camera, circularly traversing around the marked rectangular area for marking vehicles in the video shot by the full-color night monitoring camera after the pretreatment, and stopping traversing when the traversed length of the sub-window is twice of the rectangular area for marking vehicles;
2) In the traversed pixels, calculating the motion trail of the vehicle tail gas by an optical flow method, finding a black smoke motion area, and automatically dividing the black smoke motion area into black smoke areas of the vehicle tail gas by threshold values;
3) Meanwhile, detecting the attenuation condition of light in the environment of the image visible light of the preprocessed night full-color monitoring camera by using an optical attenuation method, and marking the area where light is absorbed as a black smoke area;
4) And superposing the black smoke areas marked by the optical flow method and the optical attenuation method to obtain the target black smoke area.
The cyclic traversal mode mentioned in the step 1) in the black smoke area detection is to expand 8 surrounding pixel points by taking the marked central pixel of the rectangular frame of the vehicle mark as an origin, and extract the value of the expanded pixel point and calculate the related optical flow method. And then traversing the values of surrounding 72 pixels and the calculation of a related optical flow method by taking the calculated 9 pixel points as the center. The loop is repeated until the loop is traversed until the number of pixels reaches twice the length of the rectangular frame of the vehicle marking.
The gradient-based optical flow calculation is to assign a velocity vector initial value to each pixel point of the image, so that an image motion field is formed, and at a specific moment of object motion, points on the image correspond to points of the three-dimensional object one by one. When an object in the image moves relatively, the velocity vector in the image changes relatively, and a moving area image of the optical flow is formed.
In the present invention, an optical flow field image of the vehicle tail gas and a vehicle tail gas mark image are shown in fig. 6 and 7.
Step six, measuring the tail gas black smoke grade of the vehicle with black smoke by utilizing the ringeman blackness grade;
1) The darkness detection of the tail gas of the vehicle is carried out by utilizing the QT210B type photoelectric smoke measuring telescope, and the areas marked by the optical flow method and the optical attenuation method are the areas to be detected by the photoelectric telescope. The photoelectric smoke measuring telescope calculates the blackness of a black smoke area of the out-of-standard vehicle tail gas through an internal algorithm, and then determines the blackness level by utilizing a ringeman blackness level standard, and marks as H1;
2) Extracting pixel values of a region of the black smoke vehicle tail gas calculated by an optical flow method and a black smoke region marked by an optical attenuation method, wherein the level of ringelman blackness corresponding to the pixel values is the blackness level of black smoke and is marked as H2;
3) And determining the blackness level of the tail gas by averaging the detection result of the photoelectric smoke measuring telescope and the detection result of the pixel value calculating method, namely H=H2+H2.
Step seven, tracking and video recording vehicles with black smoke exceeding standards, and updating data among cameras;
1) Tracking video recording is carried out on the identified black-smoke vehicle for 5 seconds, and license plate information of the black-smoke vehicle and road section information where the license plate information is located are uploaded;
2) The black smoke vehicle information with the exceeding tail gas is transmitted between cameras in a Local Area Network (LAN) constructed by LORA;
3) The black smoke vehicle information with the exceeding tail gas is uploaded to the upper computer in an optical fiber mode at the same time, and is used for displaying and alarming the black smoke vehicle information;
4) The repeatedly occurring black smoke vehicle will provide location information to transmit its location information to the monitoring center in real time.
The proposed construction of the local area network has the following sub-steps:
a) Each camera is connected to and numbered from the LORA development board.
B) The LORA development boards are connected with each other through a communication protocol inside the development boards.
C) And after the black smoke vehicle is captured, transmitting the signals recognized by the black smoke vehicle to a LORA development board with a corresponding number to complete the transmission of the shooting information of the corresponding camera, wherein the built local area network is used for transmitting data information among cameras of different road segments.
The detection evidence and the vehicle information detection result in the present invention are shown in fig. 8.
Therefore, by fully utilizing and improving the existing algorithms in the fields of target detection and the like, the intelligent black-smoke vehicle monitoring system which can adapt to the complex natural environment is developed, the method is high in efficiency, the road segments are continuously subjected to snapshot analysis in 24 hours, powerful evidence can be provided for repairing black-smoke vehicles, the biggest advantage can be achieved by taking a snapshot at night under the condition that traffic is not affected, and a pair of clear 'eyes' are provided for black-smoke vehicle detection at night.
For the difficult problems of night monitoring and black smoke position monitoring, the intelligent environment-friendly black smoke vehicle electronic monitoring system for monitoring is an important application of combining an Artificial Intelligence (AI) technology with environment-friendly service, and can effectively realize the recognition of black smoke vehicles and the monitoring of black smoke grades after the system is implemented. The vehicle tail gas blackness is monitored by taking the ringeman blackness as a unified blackness grade standard, so that the standard can be unified, the persuasion of monitoring results can be ensured, the method has important significance for promoting the promotion of the atmospheric pollution control work, and plays a vital role in the special treatment of high-emission motor vehicles.
The foregoing description is only illustrative of the preferred embodiment of the present invention, and is not to be construed as limiting the invention, but is to be construed as limiting the invention to any and all simple modifications, equivalent variations and adaptations of the embodiments described above, which are within the scope of the invention, may be made by those skilled in the art without departing from the scope of the invention.
The invention is applicable to the prior art where it is not described.
Claims (6)
1. The intelligent environment-friendly electronic snapshot system for the black-smoke road vehicle comprises a high-definition license plate recognition camera, a stroboscopic light supplementing lamp and an environment-friendly monitoring center, and is characterized by further comprising a high-brightness light supplementing lamp, a full-color night monitoring camera, an environment detector, a photoelectric smoke measuring telescope and a LORA development board; the high-definition license plate recognition camera, the stroboscopic light supplementing lamp, the highlighting light supplementing lamp, the night full-color camera, the environment detector, the photoelectric smoke measuring telescope and the LORA development board are all arranged on a portal frame of a road, the camera of the high-definition license plate recognition camera is opposite to the head of an incoming vehicle, the camera of the night full-color monitoring camera and the lens of the photoelectric smoke measuring telescope face to the tail of the vehicle, the environment detector is arranged around the two cameras, the night full-color monitoring camera is connected with the LORA development board, and the LORA development boards of the night full-color monitoring cameras of all roads are constructed to form a local area network to realize networking communication among the cameras; the environment-friendly monitoring center comprises a terminal monitoring host and an upper computer, and the LORA development board can receive data transmitted by the terminal monitoring host;
the high-definition license plate recognition camera and the stroboscopic light supplementing lamp are combined to shoot license plate information and the face of a driver,
The night full-color monitoring camera and the high-brightness light supplementing lamp are combined to detect black smoke around the vehicle;
The environment detector detects real-time environment change and is used for adjusting the light supplementing brightness and acquiring the influence of the environment on tail gas;
The high-definition license plate recognition camera, the stroboscopic light supplementing lamp, the night full-color monitoring camera, the high-definition light supplementing lamp and the environment detector form a front-end snapshot subsystem;
The system comprises a high-definition license plate recognition camera, a stroboscopic light supplementing lamp, a night full-color monitoring camera, a high-brightness light supplementing lamp and an environment detector, wherein acquisition information of the high-definition license plate recognition camera, the stroboscopic light supplementing lamp, the night full-color monitoring camera, the high-brightness light supplementing lamp and the environment detector is transmitted to a terminal monitoring host through an accessed switch and an optical fiber transceiver respectively, and the terminal monitoring host is used for carrying out data processing and storage on the acquisition information, wherein specific data processing and storage processes comprise vehicle recognition, black smoke area recognition, black smoke blackness detection and vehicle tracking and positioning;
the vehicle identification process is that;
in a video shot by a high-definition license plate recognition camera, haar feature extraction is carried out on each frame of image containing a complete vehicle, training of a cascade classifier is carried out aiming at the extracted features, and recognition of a vehicle sample is carried out through the trained vehicle features;
Meanwhile, carrying out vehicle shadow detection on each frame of image in the video shot by the high-definition license plate recognition camera, carrying out logical AND operation on a vehicle shadow detection result and a vehicle sample recognition result, carrying out rectangular marking on the recognized vehicle, and recording a vehicle marking rectangular area;
the black smoke region identification process comprises the following steps:
1) Marking a rectangular area marked by a vehicle in a video image shot by a full-color monitoring camera at night after the vehicle is identified in the video shot by the high-definition license plate identification camera, circularly traversing around the vehicle by taking the marked vehicle as the center in the video shot by the full-color monitoring camera at night, and stopping traversing when the traversed length of a sub-window is twice the length of the rectangular marked vehicle;
2) In the traversed pixels, calculating the motion trail of the vehicle tail gas by a gradient-based optical flow method, finding a black smoke motion area, and automatically threshold-dividing the black smoke motion area to divide the black smoke motion area into vehicle tail gas black smoke areas;
3) Meanwhile, the light attenuation condition of the night full-color monitoring camera under the environment of the visible light of the image is detected by a light attenuation method, and the area where the light is absorbed is marked as a black smoke area;
4) Superposing black smoke areas marked by an optical flow method and an optical attenuation method to obtain a target black smoke area;
The vehicle tracking and positioning is used for tracking and recording vehicles with black smoke exceeding standards and updating data among cameras, and the specific process is as follows;
When the detected black smoke vehicle is detected to be out of standard, the information of the vehicle is transmitted to the upper computer in an optical fiber mode for displaying and alarming, clear out-of-standard vehicle information and position information are provided for traffic departments, and the information is transmitted to the next path of camera in a local area network mode, so that when the next path of high-definition license plate recognition camera recognizes the transmitted out-of-standard vehicle information, the tail gas black smoke region recognition and black smoke blackness detection process is skipped, the relevant information of the out-of-standard vehicle is directly transmitted to the upper computer through the optical fiber, and the vehicle information is transmitted to the next path of camera through the local area network; the repeatedly-appearing black smoke vehicles provide position information, and the position information is transmitted to an environment-friendly monitoring center in real time, so that tracking and video recording of the black smoke exceeding vehicles are realized;
and the terminal monitoring host transmits the processed data to the upper computer, and the upper computer observes the blackness level of the vehicle tail gas acquired by the front-end snapshot subsystem and stores license plate information of the corresponding vehicle with the blackness not reaching the standard.
2. The snapshot system according to claim 1, wherein the interface of the upper computer is provided with a video monitoring window, a bayonet snapshot window, a manual review feeding window, a review notification management, a blacklist management, a comprehensive query, a statistical analysis, a device and network management, a data interaction, a system setting and a help;
the video monitoring window has two functions of real-time monitoring and video playback window;
The bayonet snapshot window records images of all vehicles passing through the highway lane, recognizes license plates of the vehicles in real time, uploads the license plates to the data center, and derives recognition results to be shared with public security and traffic police;
the manual review sending window is used for manually uploading the corresponding video to detect again under the condition that the intelligent snapshot result is disagreeable, and completing result verification;
The inspection notification management is to correlate the black smoke car subjected to manual review, if the environmental protection department has the information of the administrative district to which the car owner, the contact phone, the running certificate and the registration belong, and generate a file of 'automobile exhaust detection notification' or 'automobile black smoke emission punishment notification';
The blacklist management can perform centralized management on all blacklist vehicles, search is performed by inputting license plate numbers and snapshot times in a search bar, and search results are displayed in a search result bar; the method comprises the steps of supporting entering a driver archive management page according to owner information to manage drivers of blacklisted vehicles;
The comprehensive inquiry window function comprises a punishment identification and punishment mode specified by the country about the black smoke car;
The statistical analysis window supports the condition of displaying the snapshot accuracy rate in a cake pattern mode by selecting monitoring points and statistical time ranges which need to be counted, and the condition of rechecking the black smoke car; supporting to display the proportion of the black smoke car to the yellow standard car in a cake graph mode and the proportion of the black smoke car to the green standard car;
The equipment and network management window supports configuration and maintenance of equipment information of the snapshot point and remote information maintenance of equipment; supporting to display the online networking condition of each device on a GIS map;
The data interaction window can realize the data interaction between the snapshot system and the traffic control department and the data interaction between the snapshot system and the vehicle owner.
3. The snapshot system according to claim 1, wherein the black smoke darkness detection is used for determining a black smoke level of a vehicle tail gas in which black smoke exists, and the specific process is as follows:
The blackness detection of the tail gas of the vehicle is carried out on the marked area by the optical flow method and the optical attenuation method by utilizing the photoelectric smoke measuring telescope, the blackness calculation is carried out on the tail gas of the black smoke vehicle by utilizing the photoelectric smoke measuring telescope, and then the blackness level is determined by utilizing the ringeman blackness level standard;
Performing image preprocessing on the video image of the black smoke area calculated by the optical flow method, and extracting pixel values after the image preprocessing, wherein the level of the ringelmann blackness corresponding to the pixel values is the blackness level of the black smoke;
And determining the blackness level of the tail gas by averaging the blackness level corresponding to the detection result of the photoelectric smoke measuring telescope and the blackness level result detected by the pixel value calculation method.
4. A snapshot system according to claim 3, wherein the image preprocessing process is:
Converting the color space of each frame of image, and drawing a corresponding gray level histogram; dividing the background and the foreground according to the peak value of the histogram; and performing edge detection and histogram equalization operation on the image to obtain better image information.
5. The snapshot system according to claim 1, wherein the specific process of license plate recognition is:
Calling a video image shot by a high-definition license plate recognition camera after marking the vehicle marking rectangle, searching and framing license plates from the vehicle marking rectangle of the video image in a color matching and area matching mode;
Intercepting the license plate part selected by the frame, and preprocessing license plate images of the intercepted part; and the identified vehicle is used as a vehicle with the standard exceeding black smoke level and is uploaded to an upper computer interface.
6. A snapshot system according to claim 5, wherein the preprocessing of the license plate image comprises: the license plate is changed into a front view which is convenient for matching the license plate through perspective transformation; dividing characters on a license plate; character recognition is carried out in a template matching mode, and finally recognition of license plates is completed;
the perspective transformation is used for transforming the license plate part obtained by cutting the rectangular area of the vehicle mark into the angle of the front view.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911060713.6A CN110660222B (en) | 2019-11-01 | Intelligent environment-friendly electronic snapshot system for black-smoke road vehicle |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911060713.6A CN110660222B (en) | 2019-11-01 | Intelligent environment-friendly electronic snapshot system for black-smoke road vehicle |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110660222A CN110660222A (en) | 2020-01-07 |
| CN110660222B true CN110660222B (en) | 2024-05-10 |
Family
ID=
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103808723A (en) * | 2014-02-27 | 2014-05-21 | 中国科学技术大学 | Exhaust gas blackness automatic detection device for diesel vehicles |
| CN104794486A (en) * | 2015-04-10 | 2015-07-22 | 电子科技大学 | Video smoke detecting method based on multi-feature fusion |
| CN104899843A (en) * | 2015-06-30 | 2015-09-09 | 西南石油大学 | Real-time haze-eliminating displayer and haze-eliminating display method thereof |
| CN105467003A (en) * | 2015-12-31 | 2016-04-06 | 天津云视科技发展有限公司 | Environment-friendly monitoring system based on Internet of things |
| EP3082116A1 (en) * | 2013-12-11 | 2016-10-19 | TNK Corporation Ltd. | Guidance system |
| CN110097571A (en) * | 2019-04-28 | 2019-08-06 | 重庆大学 | The vehicle collision prediction technique of quick high accuracy |
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP3082116A1 (en) * | 2013-12-11 | 2016-10-19 | TNK Corporation Ltd. | Guidance system |
| CN103808723A (en) * | 2014-02-27 | 2014-05-21 | 中国科学技术大学 | Exhaust gas blackness automatic detection device for diesel vehicles |
| CN104794486A (en) * | 2015-04-10 | 2015-07-22 | 电子科技大学 | Video smoke detecting method based on multi-feature fusion |
| CN104899843A (en) * | 2015-06-30 | 2015-09-09 | 西南石油大学 | Real-time haze-eliminating displayer and haze-eliminating display method thereof |
| CN105467003A (en) * | 2015-12-31 | 2016-04-06 | 天津云视科技发展有限公司 | Environment-friendly monitoring system based on Internet of things |
| CN110097571A (en) * | 2019-04-28 | 2019-08-06 | 重庆大学 | The vehicle collision prediction technique of quick high accuracy |
Non-Patent Citations (3)
| Title |
|---|
| The effects of deterioration and technological levels on pollutant emission factors for gasoline light-duty trucks;Qingyu Zhang, Juwang Fan, Weidong Yang, Bixin Chen, Lijuan Zhang, Jiaoyu Liu, Jingling Wang, Chunyao Zhou and Xuan Chen;《Journal of the Air & Waste Management Association》;20170519;第67卷(第07期);814-823 * |
| 基于机器视觉的在用柴油车烟度测量方法;梁荆璞，文国军，甘露，吴丹，王玉丹;《计算机测量与控制》;20190425;第27卷(第04期);32-35+45 * |
| 柴油车尾气光学烟度计;陈玉琼;《中国集成电路》;20120505;第21卷(第05期);75-77+86 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112101433B (en) | Automatic lane-dividing vehicle counting method based on YOLO V4 and DeepSORT | |
| CN104751634B (en) | The integrated application method of freeway tunnel driving image acquisition information | |
| CN102867417B (en) | Taxi anti-forgery system and taxi anti-forgery method | |
| CN103093249B (en) | A kind of taxi identification method based on HD video and system | |
| US20140028842A1 (en) | Calibration device and method for use in a surveillance system for event detection | |
| CN104951775A (en) | Video technology based secure and smart recognition method for railway crossing protection zone | |
| TW202013252A (en) | License plate recognition system and license plate recognition method | |
| CN104978567B (en) | Vehicle checking method based on scene classification | |
| WO2004042673A2 (en) | Automatic, real time and complete identification of vehicles | |
| CN113947731A (en) | Foreign matter identification method and system based on contact net safety inspection | |
| CN112651293B (en) | Video detection method for road illegal spreading event | |
| CN114170580A (en) | Highway-oriented abnormal event detection method | |
| CN109165602A (en) | A kind of black smoke vehicle detection method based on video analysis | |
| CN112465854A (en) | Unmanned aerial vehicle tracking method based on anchor-free detection algorithm | |
| CN110765900B (en) | Automatic detection illegal building method and system based on DSSD | |
| CN210515650U (en) | Intelligent environment-friendly electronic snapshot system for black smoke vehicle on road | |
| CN112115737B (en) | Vehicle orientation determining method and device and vehicle-mounted terminal | |
| CN111950499A (en) | Method for detecting vehicle-mounted personnel statistical information | |
| CN110660222B (en) | Intelligent environment-friendly electronic snapshot system for black-smoke road vehicle | |
| CN202887450U (en) | Taxi anti-fake system | |
| Pan et al. | Identifying Vehicles Dynamically on Freeway CCTV Images through the YOLO Deep Learning Model. | |
| CN113850112A (en) | Road condition identification method and system based on twin neural network | |
| CN113850111A (en) | Road condition identification method and system based on semantic segmentation and neural network technology | |
| CN110660222A (en) | Intelligent environment-friendly electronic snapshot system for black smoke vehicle on road | |
| KR102595910B1 (en) | Vehicle number automatic recognition system and method therefor |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant |