CN114530042A - Urban traffic brain monitoring system based on internet of things technology - Google Patents
Urban traffic brain monitoring system based on internet of things technology Download PDFInfo
- Publication number
- CN114530042A CN114530042A CN202111650637.1A CN202111650637A CN114530042A CN 114530042 A CN114530042 A CN 114530042A CN 202111650637 A CN202111650637 A CN 202111650637A CN 114530042 A CN114530042 A CN 114530042A
- Authority
- CN
- China
- Prior art keywords
- vehicle
- license plate
- internet
- things
- monitoring system
- 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.)
- Pending
Links
- 238000012544 monitoring process Methods 0.000 title claims abstract description 33
- 238000005516 engineering process Methods 0.000 title claims abstract description 28
- 210000004556 brain Anatomy 0.000 title claims abstract description 19
- 238000001514 detection method Methods 0.000 claims abstract description 63
- 238000004458 analytical method Methods 0.000 claims abstract description 37
- 238000000034 method Methods 0.000 claims abstract description 30
- 230000006870 function Effects 0.000 claims abstract description 10
- 230000033001 locomotion Effects 0.000 claims description 42
- 230000009466 transformation Effects 0.000 claims description 15
- 238000004364 calculation method Methods 0.000 claims description 11
- 230000008859 change Effects 0.000 claims description 8
- 238000003708 edge detection Methods 0.000 claims description 8
- 238000001914 filtration Methods 0.000 claims description 8
- 230000011218 segmentation Effects 0.000 claims description 6
- 238000012545 processing Methods 0.000 claims description 5
- 238000013461 design Methods 0.000 claims description 4
- 238000009434 installation Methods 0.000 claims description 3
- 230000026676 system process Effects 0.000 claims description 2
- 230000008569 process Effects 0.000 abstract description 9
- 206010039203 Road traffic accident Diseases 0.000 abstract description 6
- 230000009286 beneficial effect Effects 0.000 abstract description 4
- 230000001737 promoting effect Effects 0.000 abstract description 4
- 230000006399 behavior Effects 0.000 description 49
- 238000010586 diagram Methods 0.000 description 11
- 239000011159 matrix material Substances 0.000 description 4
- 230000002349 favourable effect Effects 0.000 description 3
- 206010000117 Abnormal behaviour Diseases 0.000 description 2
- 230000002159 abnormal effect Effects 0.000 description 2
- 238000013528 artificial neural network Methods 0.000 description 2
- 238000006243 chemical reaction Methods 0.000 description 2
- 238000006073 displacement reaction Methods 0.000 description 2
- 239000000284 extract Substances 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 230000007704 transition Effects 0.000 description 2
- PXFBZOLANLWPMH-UHFFFAOYSA-N 16-Epiaffinine Natural products C1C(C2=CC=CC=C2N2)=C2C(=O)CC2C(=CC)CN(C)C1C2CO PXFBZOLANLWPMH-UHFFFAOYSA-N 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 238000005286 illumination Methods 0.000 description 1
- 230000010365 information processing Effects 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 230000035772 mutation Effects 0.000 description 1
- 230000008447 perception Effects 0.000 description 1
- 230000002265 prevention Effects 0.000 description 1
- 230000001105 regulatory effect Effects 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 230000002194 synthesizing effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/01—Detecting movement of traffic to be counted or controlled
- G08G1/017—Detecting movement of traffic to be counted or controlled identifying vehicles
- G08G1/0175—Detecting movement of traffic to be counted or controlled identifying vehicles by photographing vehicles, e.g. when violating traffic rules
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/20—Analysis of motion
- G06T7/246—Analysis of motion using feature-based methods, e.g. the tracking of corners or segments
- G06T7/251—Analysis of motion using feature-based methods, e.g. the tracking of corners or segments involving models
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/20—Analysis of motion
- G06T7/277—Analysis of motion involving stochastic approaches, e.g. using Kalman filters
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/01—Detecting movement of traffic to be counted or controlled
- G08G1/052—Detecting movement of traffic to be counted or controlled with provision for determining speed or overspeed
- G08G1/054—Detecting movement of traffic to be counted or controlled with provision for determining speed or overspeed photographing overspeeding vehicles
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N23/00—Cameras or camera modules comprising electronic image sensors; Control thereof
- H04N23/60—Control of cameras or camera modules
- H04N23/695—Control of camera direction for changing a field of view, e.g. pan, tilt or based on tracking of objects
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N7/00—Television systems
- H04N7/18—Closed-circuit television [CCTV] systems, i.e. systems in which the video signal is not broadcast
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T10/00—Road transport of goods or passengers
- Y02T10/10—Internal combustion engine [ICE] based vehicles
- Y02T10/40—Engine management systems
Landscapes
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Signal Processing (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Theoretical Computer Science (AREA)
- Traffic Control Systems (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses an urban traffic brain control system based on the technology of the Internet of things, which relates to the technical field of the Internet of things and mainly comprises an intelligent traffic monitoring system, wherein the intelligent traffic monitoring system comprises three subsystems, namely a video-based vehicle speed detection system, a video-based vehicle behavior semantic analysis system and a city road-based license plate recognition system. The vehicle speed detection system, the vehicle behavior semantic analysis system and the license plate recognition system are combined to form an intelligent traffic monitoring system. The intelligent traffic monitoring system provided by the invention processes the video data and the image data acquired based on the Internet of things, realizes the functions of vehicle speed detection, vehicle behavior semantic analysis, license plate recognition and the like, can snapshot the violation vehicles, and then recognizes the license plates of the violation vehicles, is beneficial to promoting the management and control of intelligent traffic and reducing the occurrence of traffic accidents.
Description
The technical field is as follows:
the invention relates to the field of Internet of things, in particular to an urban traffic brain monitoring system based on the Internet of things technology.
Background art:
the internet of things (IOT) is a network connecting physical objects to the internet for acquiring and processing information from the physical objects, and is currently defined as a network that connects objects to the internet according to corresponding protocols through various information sensing devices to collect, transmit and process data, and realize intelligent identification, positioning, tracking, monitoring and management through terminals or platforms.
At present, with the popularization of the internet, the development of the internet of things technology and the trend of people to intelligent life, the internet of things technology is widely applied to various fields, in particular to the construction of smart city traffic. The intelligent traffic system adopts the technology of Internet of things, people can timely acquire the conditions of urban roads through sensing equipment arranged on the urban roads, can select reasonable travel routes according to road conditions, and reduces road congestion, and traffic managers can reasonably adjust the time length of traffic lights at traffic intersections with the help of an intelligent traffic scheduling system by utilizing the traffic road information acquired in real time, so that the urban congestion is relieved, the traffic flow is optimized, and the utilization rate of the existing traffic resources is improved to the maximum extent.
The intelligent traffic system (intelligent transport system, ITS for short) is to apply the internet of things to the intelligent traffic field, connect pedestrians, vehicles and roads by using various advanced sensor devices, network transmission technology, information processing technology and computer vision technology, establish a comprehensive, accurate and timely comprehensive management and control traffic monitoring system, and has wide and comprehensive functions.
With the increasing influence of the intelligent traffic monitoring system on the life of people, the vehicle speed detection system, the vehicle driving behavior semantic analysis system and the license plate recognition system are more and more important as part of the intelligent traffic monitoring system. Among them, the license plate recognition system is abbreviated as lpr (license plate recognition), and is widely applied to the fields of parking lots, highway toll stations, and the like. The vehicle speed detection system is widely applied to expressways and urban roads, and plays a vital role in traffic safety, traffic accident prevention and the like. The automatic recognition and semantic analysis of vehicle traffic behaviors as key parts in intelligent traffic monitoring systems have attracted much attention in recent years from research institutions of all countries around the world. By exploring the urban road traffic monitoring direction, the method is favorable for finding out the abnormal traffic behavior of the vehicle in time and rapidly sending the abnormal information to the control center, thereby being favorable for the traffic management department to take effective measures in time and preventing serious traffic accidents from happening.
The invention content is as follows:
the invention aims to solve the problems that: by processing the video and image data, the intelligent traffic monitoring system realizes the functions of vehicle speed detection, vehicle driving behavior semantic analysis, license plate recognition and the like, can snapshot the violation vehicles and then recognize the license plates of the violation vehicles, is favorable for promoting the management and control of intelligent traffic and reduces the occurrence of traffic accidents. An urban traffic brain monitoring system based on the technology of the Internet of things is provided.
The technical scheme adopted by the invention for solving the technical problems is as follows: an urban traffic brain monitoring system based on the technology of the Internet of things comprises a video-based vehicle speed detection system, a video-based vehicle behavior semantic analysis system and a city road-based license plate recognition system, wherein the intelligent traffic monitoring system processes video data and image data acquired based on the Internet of things to realize the functions of vehicle speed detection, vehicle behavior semantic analysis and license plate recognition, can capture illegal vehicles and then recognize license plates of the illegal vehicles.
The urban traffic brain monitoring system based on the technology of the Internet of things comprises the following design steps:
the method comprises the following steps: the vehicle speed detection subsystem firstly sets a virtual coil, then obtains the time required by the vehicle to pass through a fixed distance through license plate positioning and license plate recognition, and finally obtains the vehicle speed by dividing the distance by the time.
Step two: the vehicle behavior semantic analysis subsystem detects the running track of the vehicle by video detection and tracking of a moving vehicle target, realizes detection and analysis of vehicle behavior semantics by combining the relation between the running track of the vehicle and a lane line segment, judges whether the vehicle has a violation behavior or not, and finally realizes the function of violation behavior alarm.
Step three: the license plate recognition subsystem firstly carries out rough positioning on license plate characters by utilizing a maximum stable extreme value region method, and then carries out accurate positioning on the license plate by combining the license plate characteristics and the license plate character characteristics, so that an accurate license plate region can be obtained. And then, identifying the license plate number by utilizing the character segmentation and the license plate identification of the license plate.
Furthermore, a sensing layer of the Internet of things acquires data required by the system through a camera, and the camera is arranged right above an urban road or on one side of the road and is 5-10 meters in height.
The main principle of the video-based vehicle speed detection algorithm is based on the most basic physical kinematics, and the vehicle movement speed is obtained by calculating the vehicle movement position removal and the vehicle movement time, wherein a specific vehicle speed calculation formula is shown as follows.
Further, two or more virtual coils are arranged in the video, and the specific steps of the virtual coil vehicle speed detection algorithm are as follows:
the method comprises the steps of firstly, setting two appropriate vehicle speed detection virtual coils according to the installation position of a road traffic camera and the road surface condition shot by the camera, and knowing the actual distance between the two virtual coils.
And secondly, judging whether the vehicle passes through the virtual coil 1, if so, positioning and identifying the license plate number of the vehicle, recording the passing time of the vehicle, and then executing the third step, otherwise, circulating the second step.
And step three, judging whether the vehicle passes through the virtual coil 2, if so, positioning and identifying the license plate number of the vehicle, recording the passing time of the vehicle, and then executing the step four, otherwise, returning to the step two.
And fourthly, obtaining the time consumed by the same vehicle to pass through the two virtual coils after the vehicle passes through the two virtual coils, and obtaining the average speed of the vehicle passing through the virtual coils by knowing the actual distance between the virtual coils through a speed calculation formula.
Further, the feature of the semantic representation of the vehicle behavior is first analyzed, wherein the equation of the center lane line L1 is as follows,
in the formula, a is the slope of the lane line.
Furthermore, the lane lines are detected by Hough transformation based on video vehicle behavior semantic analysis for subsequent vehicle behavior semantic analysis, and the Hough transformation principle is that the linear detection problem in an image space is converted into point detection in a parameter space by using coordinate transformation and the symmetry of points and lines in two spaces.
Further, if a linear equation is used in the parameter space, when the slope of the linear in the image space is infinite, a polar equation is selected.
Further, the step of Hough transformation detection of the straight line in the image is that the color image is processed into a gray image, then the gray image is subjected to edge detection, and then Hough transformation is carried out on the result of the edge detection, and finally the straight line detection result is obtained.
Further, based on the semantic analysis of the vehicle behavior of the video, a Kalman filtering is adopted to track the moving vehicle target, and the basic principle of a Kalman algorithm is as follows: firstly, creating a prediction model to describe a random dynamic variable changing along with the change of time, introducing a linear differential equation description to a state space xERn, then observing the random variable in real time, and optimally estimating a state value by using Kalman filtering.
Further, understanding and analyzing the semantics of the vehicle behavior are realized, and the motion features of the vehicle behavior need to be extracted. The vehicle behavior can be accurately described only by selecting proper motion characteristics, such as target position, motion speed and motion reversal.
Compared with the prior art, the invention has the beneficial effects that: (1) the intelligent traffic monitoring system disclosed by the invention processes the video data and the image data acquired based on the Internet of things, so that the functions of vehicle speed detection, vehicle behavior meaning analysis, license plate recognition and the like are realized, the illegal vehicles can be captured, and then the license plates of the illegal vehicles are recognized, so that the intelligent traffic monitoring system is beneficial to promoting the management and control of intelligent traffic and reducing traffic accidents; (2) the improved virtual coil vehicle speed detection algorithm can detect the vehicle speed in real time, and solves the problems that the virtual coil algorithm cannot detect the speeds of a plurality of vehicles at the same time and the vehicle speed in a large-scale area or across areas is detected; (3) the invention can observe that the white area of the binary image continuously appears, grows and merges and goes through a process from full black to full white.
Drawings
FIG. 1 is a schematic view of the spatial position of a camera and a road vehicle.
Fig. 2 is a perspective view of a road.
Fig. 3 is a semantic representation of a vehicle driving behavior.
Fig. 4 is a schematic diagram of the vehicle motion trajectory.
FIG. 5 is a schematic diagram of the size of a license plate unified in China.
Fig. 6 is a diagram for defining a state transition matrix a and an observation matrix H.
Detailed Description
The present invention will now be described in further detail with reference to the accompanying drawings, which are simplified schematic and block schematic illustrations, and thus show only those elements relevant to the present invention.
The invention discloses an urban traffic brain monitoring system based on the technology of Internet of things, which comprises three subsystems, namely a video-based vehicle speed detection system, a video-based vehicle behavior semantic analysis system and a city road-based license plate recognition system.
In the video-based vehicle speed detection system, the algorithm for detecting the vehicle speed by the virtual coil is improved. First add virtual coils at fixed positions in the camera scene and know the actual distance between the two virtual coils. When the vehicle is detected to pass through the virtual coil, the license plate is positioned and recognized, so that the time required for the vehicle to pass through a fixed distance is obtained, and the vehicle speed is obtained by dividing the distance by the time.
In the video-based vehicle behavior semantic analysis system, a vehicle behavior semantic analysis algorithm is designed by combining the relation between a vehicle running track and a lane line, firstly, a mixed Gaussian model is combined with a self-adaptive background difference method to realize the detection of a moving vehicle, and a Kalman filter is used for tracking a vehicle target; the center of mass of the vehicle in each frame of image is obtained, the driving track of the vehicle is synthesized, and the motion characteristics of the vehicle in each frame of image can be extracted: vehicle position, vehicle speed, and vehicle direction of travel; and analyzing and understanding the vehicle behavior semantics by combining the lane line equation to finally obtain the semantic representation of the vehicle behavior. The moving vehicle target is detected and tracked through the video, the driving track of the vehicle is obtained, and the motion characteristics of the vehicle in each frame of image, including the position, the speed and the driving direction of the vehicle, can be extracted. And according to the relation between the vehicle running track and the lane line, realizing the semantic analysis of the vehicle behavior. And when the system detects abnormal behaviors of the vehicle such as illegal parking, overspeed, illegal turning and the like, the system gives an alarm in time.
The semantic analysis of the vehicle behavior, namely representing the behavior of the vehicle in the driving process by using language, is used for judging the following behaviors of the vehicle, such as lane change, turning around, reversing, overtaking, backing and the like. Vehicle behavior semantic analysis is mainly based on the detection and tracking of moving vehicles. The video-based vehicle behavior semantic analysis algorithm is mainly divided into four parts, namely vehicle target detection and tracking, vehicle motion trajectory analysis and vehicle behavior semantic analysis.
The first part is to detect the lane line segment in the road shot by the camera through Hough transformation, calculate the equation expression of the lane line segment and label the lane line segment in the camera.
And a second part, detecting the moving vehicle by using a Gaussian mixture model and combining a self-adaptive background difference method, tracking the moving vehicle by using a Kalman filter, extracting the center of mass of the vehicle, and synthesizing the driving track of the vehicle. After the foreground target image of the vehicle is successfully extracted, the vehicle needs to be tracked. In order to improve the tracking efficiency and precision, a Kalman filter is introduced to pre-estimate the vehicle target. The Kalman filter is linear recursion, carries out optimal pre-estimation on the current state by using the state at the previous moment, has the characteristics of small time complexity and strong real-time property, and has the following prediction equation:
and the third part can extract the mass center of the vehicle target by detecting and tracking the vehicle target in the video, synthesize the driving track of the vehicle, wherein the driving track of the vehicle comprises the motion characteristics of the vehicle target, such as target position, motion direction and motion speed, and analyze the driving behavior of the vehicle by combining with a lane line. Fig. 4 is a schematic diagram of a vehicle motion trajectory, and it can be seen that the vehicle motion trajectory is formed by connecting a plurality of points through line segments, and is not a whole curve.
And fourthly, combining the motion information of the vehicle target in the continuous multi-frame image sequence, namely the position, the motion direction and the motion speed of the vehicle target, obtaining parameters such as the change rate and the speed change rate of the motion direction of the vehicle, analyzing the behavior of the vehicle, combining the lane line analysis to obtain the driving behavior of the vehicle, judging whether the vehicle has the behavior of violation of rules according to the road driving regulations, and if the vehicle violates the rules, capturing the vehicle and giving an alarm. After the vehicle is captured, the illegal vehicle can be identified through a license plate identification subsystem in the system, and automatic license plate detection is realized. The system can accurately detect the vehicle driving in the highway in real time and carry out semantic analysis, and alarm and snapshot the abnormal behaviors and the violation behaviors of the vehicle, thereby being beneficial to promoting the management and control of intelligent traffic and reducing the occurrence of traffic accidents.
In the license plate recognition system based on urban roads, the invention introduces a license plate positioning algorithm fusing a maximum stable extreme area and license plate characteristics, firstly, a license plate is roughly positioned by using a maximum stable extreme area method, then, the license plate characteristics and the license plate character characteristics are combined to accurately position the license plate, then, after the license plate is divided into single characters, the license plate characters are recognized by using LBP + SVM, namely, the LBP (local binary pattern) characteristics of the characters are extracted, and the characters are input into a trained network to recognize the license plate.
The urban traffic brain monitoring system based on the technology of the Internet of things comprises the following design steps:
the method comprises the following steps: the vehicle speed detection subsystem firstly sets a virtual coil, then obtains the time required by a vehicle to pass through a fixed distance through license plate positioning and license plate recognition, and finally obtains the vehicle speed by dividing the distance by the time.
Step two: the vehicle behavior semantic analysis subsystem detects the running track of the vehicle by video detection and tracking of a moving vehicle target, realizes detection and analysis of vehicle behavior semantics by combining the relation between the running track of the vehicle and a lane line segment, judges whether the vehicle has a violation behavior or not, and finally realizes the function of violation behavior alarm.
Step three: the license plate recognition subsystem firstly carries out rough positioning on license plate characters by utilizing a maximum stable extreme value region method, and then carries out accurate positioning on the license plate by combining the license plate characteristics and the license plate character characteristics, so that an accurate license plate region can be obtained. And then, identifying the license plate number by utilizing the character segmentation and the license plate identification of the license plate.
A license plate recognition system based on urban roads is an important part of the intelligent traffic field and is also an important research subject in the computer vision field at present. The license plate recognition system mainly processes vehicle videos and images collected on urban roads by using technologies such as mode recognition, artificial intelligence technology, computer vision and the like, and can accurately convert the license plate images into character formats which can be recognized by a computer in real time and output the character formats. License plate positioning, character cutting and character recognition are three main steps of a license plate recognition system. The method comprises the following steps of positioning a license plate, dividing characters into a plurality of parts, and identifying the license plate. Technically, two indexes are mainly used for evaluating a license plate recognition system, namely license plate recognition accuracy and recognition speed. The formula of the license plate recognition accuracy is shown below.
The size of the current Chinese unified license plate size is fixed at 44 x 14cm, the size of each character is fixed at 9 x 4.5cm, and the distance between the two characters is 1.2 cm. Therefore, the feature of fixing the aspect ratio of the size of the license plate can be utilized when the license plate is positioned, and the feature of fixing the aspect ratio of 2:1 of the size of the character can be utilized when the license plate is segmented. The schematic size diagram of the license plate unified in china is shown in fig. 5.
The essence of the edge detection is to adopt a certain algorithm to extract the regions of the object with mutation in the image. The classical edge detection method mainly extracts operators such as Roberts, Prewitt and Sobel.
The gradient operator is a first derivative operator. For an image function f (x, y), its gradient is defined as a vector as follows:
the license plate positioning algorithm based on the color segmentation consists of two parts, namely color segmentation and target positioning.
Firstly, the RGB mode color image is converted into the HSI mode, and in order to reduce the time complexity of calculation, the influence of illumination change is reduced.
Sampling the color image in HSI mode, regulating the color saturation of the image by using logarithm method, and cutting the converted color image by using color neural network.
And finally, segmenting a reasonable license plate region by a projection method according to the prior knowledge of the license plate color, the aspect ratio and the like. The license plate positioning method based on color segmentation has high accuracy, but when the color of the license plate region is approximate to the color of the background, the calculation speed is low because the neural network calculation method is adopted for calculation.
The MSER-based license plate positioning is an algorithm based on an affine feature extraction region and is also called as a maximum stable extremum region. The internal gray level of the region extracted by the maximum stable extremum region method is almost unchanged in a large range, the shape is kept unchanged under multiple thresholds, and the contrast with the background is quite obvious.
The principle based on the MSER license plate positioning algorithm is that an image is grayed, and then the image is converted into a series of binary images according to a threshold value in sequence of 255. With the increase of the threshold value, the brightness of the image is also increased, and it can be observed that the white area of the binary image continuously appears, grows and merges, and goes through a process from full black to full white. In the process, a region with small area change along with the rise of the threshold value is called a maximum stable extremum region and is also called MSER +, and conversely, a region obtained in the process from white to black is called MSER-.
At present, in the intelligent transportation field, the video based on urban road does not need to use special speed measuring equipment, and the perception layer of the Internet of things only needs to be capable of acquiring data required by the system through a camera. The camera is generally arranged right above or on one side of the urban road, and the height is generally between 5 and 10 meters. The camera is shown in fig. 1 in a schematic spatial position with respect to a road vehicle.
The main principle of the video-based vehicle speed detection algorithm is still based on the most basic physical kinematics, and the vehicle movement speed is obtained by calculating the vehicle movement position removal and the vehicle movement time, wherein a specific vehicle speed calculation formula is shown as follows.
In the video-based vehicle speed detection, the acquired data are all from the camera, so that the standard diagram of the lane line of the Chinese highway is not rectangular after being seen through, but becomes a trapezoid-like shape, as shown in the left and right diagrams in fig. 2. If the actual displacement of the vehicle motion is not known in advance by the algorithm, the actual displacement of the vehicle motion can be obtained only by processing the video data and converting the coordinates of the vehicle into actual physical coordinates by coordinate conversion, wherein the coordinates of the vehicle are pixel coordinates.
The invention improves a virtual coil vehicle speed detection algorithm by combining the vehicle speed detection principle and the advantages and disadvantages of the traditional virtual coil vehicle speed detection algorithm. Two (more than two) virtual coils are arranged in a video, when the image gray scale of the virtual coil 1 changes, the number plate of a passing vehicle is directly identified, the passing time of the vehicle is recorded, and when the image gray scale of the virtual coil 2 changes, the number plate of the passing vehicle and the recording time are identified again, so that the average speed of the vehicle passing through the virtual coils can be calculated. The invention designs a vehicle speed detection algorithm flow chart, which is shown in fig. 2 and fig. 6. The improved virtual coil vehicle speed detection algorithm comprises the following specific steps:
firstly, according to the installation position of a road traffic camera and the road surface condition shot by the camera, appropriate vehicle speed detection virtual coils are set, generally two virtual coils are set, and the actual distance between the two virtual coils is known.
And secondly, judging whether the vehicle passes through the virtual coil 1, if so, positioning and identifying the license plate number of the vehicle, recording the passing time of the vehicle, and then executing the third step, otherwise, circulating the second step.
And step three, judging whether the vehicle passes through the virtual coil 2, if so, positioning and identifying the license plate number of the vehicle, recording the passing time of the vehicle, and then executing the step four, otherwise, returning to the step two.
And fourthly, obtaining the time consumed by the same vehicle to pass through the two virtual coils after the vehicle passes through the two virtual coils, and obtaining the average speed of the vehicle passing through the virtual coils by knowing the actual distance between the virtual coils through a speed calculation formula.
The improved virtual coil vehicle speed detection algorithm is mainly characterized as follows:
1) the improved virtual coil vehicle speed detection algorithm can detect the vehicle speed in real time, and solves the problems that the virtual coil algorithm cannot detect the speeds of a plurality of vehicles at the same time and the vehicle speed in a large-range area or across areas.
2) Compared with the license plate positioning vehicle speed detection method, the improved virtual coil vehicle speed detection algorithm reduces the complexity of the algorithm because the actual distance between the virtual coils is known and coordinate conversion is not needed, in addition, the advantages of the license plate positioning vehicle speed detection method are also kept, and the license plate number of the vehicle can be obtained while the vehicle speed is detected.
3) Compared with the characteristic matching vehicle speed detection method, the improved virtual coil vehicle speed detection algorithm utilizes the license plate number of the vehicle as the unique characteristic of the vehicle, the vehicle speed can be calculated only after the same vehicle completely passes through the virtual coil, and the condition that false detection occurs or the detection object is not the same vehicle is avoided.
Of course, the improved virtual coil vehicle speed detection algorithm of the invention still has limitations. The algorithm combines the virtual coil and the license plate recognition algorithm, so that the requirements on the photographing condition and the license plate recognition technology are higher. In addition, although the method can realize speed measurement in a large range, the average speed in the region is obtained.
The features of the semantic representation of the vehicle behavior are first analyzed. Wherein, let the equations of the right lane line in the (a), (b), (c) trajectory diagram or the middle lane line L1 in the (d), (e), (O trajectory diagram in FIG. 3 be as follows,
in the formula, a is the slope of the lane line.
Therefore, if the semantic representation diagram of the vehicle behavior is combined with the vehicle motion characteristics and the lane lines, the semantic representation diagram can be described by the relationship between the motion characteristics of the vehicle running and the lane, and can also be represented by an equation between the motion characteristics and the lane, and the representation of the specific vehicle running behavior type is shown in table 3.1, wherein, a) is normal running, (b) is reverse running, (c) is stop running, (d) is head turning reverse running, (e) is lane changing, and O is overtaking.
And (3) performing semantic analysis on the vehicle behaviors based on the video, and detecting a lane line by adopting Hough transformation for subsequent semantic analysis on the vehicle behaviors. The principle of Hough transform is to convert the linear detection problem in image space into point detection in parameter space by using coordinate transformation [ }20} and the symmetry of points and lines in two spaces. In image space X-Y, the set of all collinear points can be represented by the following equation.
If the linear equation is used in the parameter space, when the slope of the image space line is infinite, the accumulator becomes very large, so that the calculation complexity is overlarge.
Therefore, the step of Hough transformation detection of the straight line in the image comprises the steps of processing the color image into a gray image, carrying out edge detection on the gray image, carrying out Hough transformation on the result of the edge detection, and finally obtaining a straight line detection result.
And performing semantic analysis on the vehicle behavior based on the video, and tracking the moving vehicle target by adopting Kalman filtering. The kalman (Kaknan) filtering considers the basic characteristics of signals and measured values, and is an algorithm for solving the linear minimum mean square error estimation of the state sequence of a dynamic system. The method introduces a concept of state space, describes a system by using a state equation and an observation equation of a state, can estimate multidimensional non-equilibrium random information, and has the following basic principle of Kalman algorithm: firstly, a prediction model is created to describe a random dynamic variable changing along with the change of time, a linear differential equation description is introduced for a state space x ERn, then the random variable is observed in real time, and the state value is optimally estimated by Kalman filtering. Since the time interval between two adjacent frames is short, the state of the tracked target is considered to be basically unchanged, and therefore the target can be assumed to be at a constant speed in the period. Based on the method, a system state x of Kalman filtering is defined as a four-dimensional vector (x, Y, Vx, Vy) T, and parameters respectively represent the position and the speed of a moving target on x and Y coordinate axes. According to the above motion model of uniform velocity, a state transition matrix a and an observation matrix H are defined, as shown in the figure.
The semantic meaning of the vehicle behavior is understood and analyzed, and the motion characteristics of the vehicle behavior need to be extracted. The vehicle behavior can be accurately described only by selecting proper motion characteristics, and the common motion characteristics comprise a target position, a motion speed and a motion reversal direction.
Claims (9)
1. The urban traffic brain monitoring system based on the Internet of things technology comprises three subsystems, namely a video-based vehicle speed detection system, a video-based vehicle behavior semantic analysis system and an urban road-based license plate recognition system, and is characterized in that: the intelligent traffic monitoring system processes video data and image data acquired based on the Internet of things, so that the functions of vehicle speed detection, vehicle behavior meaning analysis and license plate recognition are realized, illegal vehicles can be captured, and then license plates of the illegal vehicles are recognized;
the urban traffic brain monitoring system based on the technology of the Internet of things comprises the following design steps:
the method comprises the following steps: the vehicle speed detection subsystem firstly sets a virtual coil, then obtains the time required by a vehicle passing through a fixed distance through license plate positioning and license plate recognition, and finally obtains the vehicle speed by dividing the distance by the time;
step two: the vehicle behavior semantic analysis subsystem detects and tracks a moving vehicle target through video detection, detects the running track of the vehicle, realizes detection and analysis of vehicle behavior semantics by combining the relation between the running track of the vehicle and a lane line segment, judges whether the vehicle has a violation behavior or not and finally realizes a violation behavior alarm function;
step three: the license plate recognition subsystem firstly carries out rough positioning on license plate characters by using a maximum stable extreme value region method, then carries out accurate positioning on a license plate by combining license plate characteristics and license plate character characteristics to obtain an accurate license plate region, and then carries out license plate character segmentation and license plate recognition to recognize a license plate number.
2. The brain monitoring system for urban traffic based on the technology of the internet of things as claimed in claim 1, wherein: a sensing layer of the Internet of things acquires data required by the system through a camera, and the camera is arranged right above an urban road or on one side of the road and is 5-10 meters in height;
the main principle of the video-based vehicle speed detection algorithm is based on the most basic physical kinematics, the vehicle motion speed is obtained by calculating the vehicle motion position and removing the vehicle motion time, a specific vehicle speed calculation formula is shown as follows,
3. the brain monitoring system for urban traffic based on the technology of the internet of things as claimed in claim 1, wherein: two or more virtual coils are arranged in a video, and the specific steps of the virtual coil vehicle speed detection algorithm are as follows:
firstly, setting two appropriate vehicle speed detection virtual coils according to the installation position of a road traffic camera and the road surface condition shot by the camera, and knowing the actual distance between the two virtual coils;
secondly, judging whether the vehicle passes through the virtual coil 1, if so, positioning and identifying the license plate number of the vehicle, recording the passing time of the vehicle, and then executing the third step, otherwise, circulating the second step;
thirdly, judging whether the vehicle passes through the virtual coil 2, if so, positioning and identifying the license plate number of the vehicle, recording the passing time of the vehicle, and then executing the fourth step, otherwise, returning to the second step;
and fourthly, obtaining the time consumed by the same vehicle to pass through the two virtual coils after the vehicle passes through the two virtual coils, and obtaining the average speed of the vehicle passing through the virtual coils by knowing the actual distance between the virtual coils through a speed calculation formula.
4. The brain monitoring system for urban traffic based on the technology of the internet of things as claimed in claim 1, wherein: first, the features of the semantic representation of the vehicle behavior are analyzed, wherein the equation of the center lane line L1 is as follows,
in the formula, a is the slope of the lane line.
5. The brain monitoring system for urban traffic based on the technology of the internet of things as claimed in claim 1, wherein: the method is characterized in that a lane line is detected by Hough transformation based on video vehicle behavior semantic analysis for subsequent vehicle behavior semantic analysis, and the Hough transformation principle is that a straight line detection problem in an image space is converted into point detection in a parameter space by using coordinate transformation and using the symmetry of points and lines in two spaces.
6. The brain monitoring system for urban traffic based on the technology of the internet of things as claimed in claim 1, wherein: if the linear equation is used in the parameter space, when the slope of the image space linear is infinite, the polar coordinate equation is selected to be used.
7. The brain monitoring system for urban traffic based on the technology of the internet of things as claimed in claim 1, wherein: the Hough transformation detection method for the straight line in the image comprises the steps of processing the color image into a gray image, carrying out edge detection on the gray image, carrying out Hough transformation on the edge detection result, and finally obtaining a straight line detection result.
8. The brain monitoring system for urban traffic based on the technology of the internet of things as claimed in claim 1, wherein: based on the semantic analysis of the vehicle behavior of the video, a Kalman filtering is adopted to realize the tracking of a moving vehicle target, and the basic principle of a Kalman algorithm is as follows: firstly, creating a prediction model to describe a random dynamic variable changing along with the change of time, introducing a linear differential equation description to a state space xERn, then observing the random variable in real time, and optimally estimating a state value by using Kalman filtering.
9. The brain monitoring system for urban traffic based on the technology of the internet of things as claimed in claim 1, wherein: the semantic meaning of the vehicle behavior is understood and analyzed, the motion characteristics of the vehicle behavior need to be extracted, the vehicle behavior can be accurately described only by selecting proper motion characteristics, and the common motion characteristics comprise a target position, a motion speed and a motion reversal direction.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111650637.1A CN114530042A (en) | 2021-12-31 | 2021-12-31 | Urban traffic brain monitoring system based on internet of things technology |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111650637.1A CN114530042A (en) | 2021-12-31 | 2021-12-31 | Urban traffic brain monitoring system based on internet of things technology |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114530042A true CN114530042A (en) | 2022-05-24 |
Family
ID=81620478
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111650637.1A Pending CN114530042A (en) | 2021-12-31 | 2021-12-31 | Urban traffic brain monitoring system based on internet of things technology |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114530042A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114758511A (en) * | 2022-06-14 | 2022-07-15 | 深圳市城市交通规划设计研究中心股份有限公司 | Sports car overspeed detection system, method, electronic equipment and storage medium |
| CN115242853A (en) * | 2022-09-20 | 2022-10-25 | 中关村科学城城市大脑股份有限公司 | Digital twin traffic order maintenance system based on three-dimensional point cloud |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20160042239A1 (en) * | 2014-08-11 | 2016-02-11 | Here Global B.V. | Variable Speed Sign Value Prediction and Confidence Modeling |
| CN105632186A (en) * | 2016-03-11 | 2016-06-01 | 博康智能信息技术有限公司 | Method and device for detecting vehicle queue jumping behavior |
| CN106408940A (en) * | 2016-11-02 | 2017-02-15 | 南京慧尔视智能科技有限公司 | Microwave and video data fusion-based traffic detection method and device |
| CN107424412A (en) * | 2017-09-21 | 2017-12-01 | 程丹秋 | A kind of traffic behavior analysis system |
| US20180137754A1 (en) * | 2015-05-18 | 2018-05-17 | Roadmetric Ltd | Detection and documentation of tailgating and speeding violations |
| CN108335489A (en) * | 2018-03-11 | 2018-07-27 | 西安电子科技大学 | Vehicle on highway behavior semantic analysis and abnormal behaviour monitoring system and method |
| CN111667507A (en) * | 2020-06-01 | 2020-09-15 | 浙江省机电设计研究院有限公司 | Method for tracking vehicle track on highway |
| CN112927535A (en) * | 2021-01-07 | 2021-06-08 | 沈阳聚德视频技术有限公司 | Intelligent temporary signal lamp based on vision technology and implementation method thereof |
| CN113593235A (en) * | 2021-08-03 | 2021-11-02 | 安徽迪卡智能控制技术有限公司 | Traffic monitoring system based on internet of things technology |
-
2021
- 2021-12-31 CN CN202111650637.1A patent/CN114530042A/en active Pending
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20160042239A1 (en) * | 2014-08-11 | 2016-02-11 | Here Global B.V. | Variable Speed Sign Value Prediction and Confidence Modeling |
| US20180137754A1 (en) * | 2015-05-18 | 2018-05-17 | Roadmetric Ltd | Detection and documentation of tailgating and speeding violations |
| CN105632186A (en) * | 2016-03-11 | 2016-06-01 | 博康智能信息技术有限公司 | Method and device for detecting vehicle queue jumping behavior |
| CN106408940A (en) * | 2016-11-02 | 2017-02-15 | 南京慧尔视智能科技有限公司 | Microwave and video data fusion-based traffic detection method and device |
| CN107424412A (en) * | 2017-09-21 | 2017-12-01 | 程丹秋 | A kind of traffic behavior analysis system |
| CN108335489A (en) * | 2018-03-11 | 2018-07-27 | 西安电子科技大学 | Vehicle on highway behavior semantic analysis and abnormal behaviour monitoring system and method |
| CN111667507A (en) * | 2020-06-01 | 2020-09-15 | 浙江省机电设计研究院有限公司 | Method for tracking vehicle track on highway |
| CN112927535A (en) * | 2021-01-07 | 2021-06-08 | 沈阳聚德视频技术有限公司 | Intelligent temporary signal lamp based on vision technology and implementation method thereof |
| CN113593235A (en) * | 2021-08-03 | 2021-11-02 | 安徽迪卡智能控制技术有限公司 | Traffic monitoring system based on internet of things technology |
Non-Patent Citations (1)
| Title |
|---|
| 杨全琪等: "基于Hough变换的智能车辆车道线检测", 《科技视界》, pages 121 - 122 * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114758511A (en) * | 2022-06-14 | 2022-07-15 | 深圳市城市交通规划设计研究中心股份有限公司 | Sports car overspeed detection system, method, electronic equipment and storage medium |
| CN114758511B (en) * | 2022-06-14 | 2022-11-25 | 深圳市城市交通规划设计研究中心股份有限公司 | Sports car overspeed detection system, method, electronic equipment and storage medium |
| CN115242853A (en) * | 2022-09-20 | 2022-10-25 | 中关村科学城城市大脑股份有限公司 | Digital twin traffic order maintenance system based on three-dimensional point cloud |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108052880B (en) | Virtual and real lane line detection method for traffic monitoring scene | |
| CN108230254B (en) | Automatic detection method for high-speed traffic full lane line capable of self-adapting scene switching | |
| CN110619279B (en) | Road traffic sign instance segmentation method based on tracking | |
| US20090309966A1 (en) | Method of detecting moving objects | |
| CN111753797B (en) | Vehicle speed measuring method based on video analysis | |
| CN106682586A (en) | Method for real-time lane line detection based on vision under complex lighting conditions | |
| CN108335489A (en) | Vehicle on highway behavior semantic analysis and abnormal behaviour monitoring system and method | |
| CN105744232A (en) | Method for preventing power transmission line from being externally broken through video based on behaviour analysis technology | |
| WO2015089867A1 (en) | Traffic violation detection method | |
| CN104378582A (en) | Intelligent video analysis system and method based on PTZ video camera cruising | |
| CN114530042A (en) | Urban traffic brain monitoring system based on internet of things technology | |
| CN104978567A (en) | Vehicle detection method based on scenario classification | |
| CN102915433A (en) | Character combination-based license plate positioning and identifying method | |
| CN114898296A (en) | Bus lane occupation detection method based on millimeter wave radar and vision fusion | |
| CN103034843A (en) | Method for detecting vehicle at night based on monocular vision | |
| CN116434159A (en) | Traffic flow statistics method based on improved YOLO V7 and Deep-Sort | |
| CN108520528B (en) | Mobile vehicle tracking method based on improved difference threshold and displacement matching model | |
| Li et al. | Automatic passenger counting system for bus based on RGB-D video | |
| Su et al. | A new local-main-gradient-orientation HOG and contour differences based algorithm for object classification | |
| Cheng et al. | Structure-aware network for lane marker extraction with dynamic vision sensor | |
| Ghahremannezhad et al. | Robust road region extraction in video under various illumination and weather conditions | |
| Liu et al. | Towards industrial scenario lane detection: vision-based AGV navigation methods | |
| Ren et al. | Automatic measurement of traffic state parameters based on computer vision for intelligent transportation surveillance | |
| Zhou et al. | Real-time traffic light recognition based on c-hog features | |
| Ding et al. | A comprehensive approach for road marking detection and recognition |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20220524 |
|
| WD01 | Invention patent application deemed withdrawn after publication |