16 results about "Red light running" patented technology

The invention relates to the technical field of VR panoramic road safety illegal behavior snapshot scheme design, in particular to a VR panoramic road safety illegal behavior snapshot method and system. The method comprises the following steps: acquiring an intersection panoramic video stream through a single panoramic device; illegal evidence obtaining is triggered based on space overlapping of the vehicle track and the electronic fence; synchronously judging a red light running behavior by combining a signal lamp state; identifying retrograde motion by comparing the motion vector with the lane direction vector; all-weather overspeed monitoring is realized by using a millimeter wave radar; and generating an evidence packet containing the two-dimensional panorama and the three-dimensional reconstruction model. The system is composed of a panoramic acquisition module, an electronic fence module, a behavior analysis module, an illegal evidence obtaining module and a data communication module. The method has the core advantages that all-intersection coverage is achieved through single equipment, the illegal recognition accuracy is improved through multi-source data fusion, the evidence effectiveness is enhanced through the three-dimensional reconstruction technology, and the law enforcement cost and the dispute risk are remarkably reduced.

The invention relates to the technical field of computer vision recognition, particularly discloses a vehicle illegal behavior recognition method based on computer vision, and aims to solve the problems that a traditional method is low in recognition precision, high in false alarm rate and difficult to model dynamic illegal behaviors in a complex traffic scene. The method comprises the following steps: acquiring a high-frame-rate high-definition video stream and performing time synchronization; vehicle detection and trajectory generation are realized through an improved YOLOv7 model and a multi-hypothesis tracking algorithm; fusing the vehicle track, the lane attribution, the signal lamp state and the road topology to construct a multi-dimensional space-time behavior feature map; performing end-to-end illegal behavior discrimination by using a space-time convolutional neural network and gating cycle unit fusion model; and automatically generating a structured illegal evidence packet with a digital signature. According to the technical scheme, high-precision recognition of behaviors such as red light running, line pressing and illegal lane changing can be achieved under the conditions of dense traffic flow, shielding and complex illumination, and the integrity of law enforcement data and the real-time performance of the system are remarkably improved.

The embodiment of the invention relates to a non-motor vehicle red light running behavior supervision system. The system comprises a plurality of vehicle acquisition modules, a plurality of signal lamp acquisition modules, a data receiving module, a message queue module, a data processing module, a road high-precision map and a light running evidence library. The vehicle and signal lamp acquisition modules are respectively used for acquiring non-motor vehicle and signal lamp messages; the data receiving module is used for receiving and issuing non-motor vehicle and signal lamp acquisition messages; the vehicle track processing task of the data processing module is used for updating the corresponding vehicle track queue according to the published message of each vehicle acquisition queue, and the to-be-checked vehicle processing task is used for updating the to-be-checked vehicle queue according to the published message of each vehicle track queue; and the lamp running evidence processing task is used for carrying out conjoint analysis on the to-be-checked vehicle queue, the signal lamp acquisition queue and the road high-precision map and updating the lamp running evidence library based on an analysis result. According to the invention, the conventional monitoring blind area problem can be solved, and the monitoring real-time performance can be improved.

The invention discloses a truck red light running snapshot system based on space-time consistency detection, and relates to the technical field of space-time information processing, and the system comprises a time calibration module which carries out the time calibration, and enables the time of all devices to be synchronized; the image acquisition module is used for triggering and capturing a red light running electric alarm picture when the vehicle runs the red light, and synchronously triggering and capturing a vehicle foot clamping bayonet picture; the front-end storage module is used for receiving and storing the electric police picture and the bayonet picture; the space-time retrieval module retrieves the checkpoint picture in the forced synthesis time interval by taking the illegal time of the electronic police picture as a reference; the license plate matching judgment module is used for judging whether a picture consistent with an electronic police picture license plate exists in the checkpoint pictures retrieved by the space-time retrieval module or not; and the image synthesis output module is used for outputting illegal pictures according to the matching synthesis result of the license plate matching judgment module and the retrieval result of the space-time retrieval module. According to the invention, the problem of low snapshot efficiency in the existing red light running snapshot technology of the truck is solved; the effect of standardizing the road traffic order is achieved.

The application relates to an intersection vehicle abnormal driving detection method based on machine vision detection and road passing area modeling, relates to the intelligent traffic technology field, and is characterized in that, in the real-time video picture of the electric police monitoring at the intersection of the urban road, a machine vision algorithm is used to detect the red and green light state and the spatial motion information of the vehicle; then, the road passing area is modeled based on perspective transformation, and a one-to-one mapping relationship between the video picture and the real road coordinates is obtained; whether the vehicle exists in the abnormal driving traffic events such as overspeed, abnormal parking, green light non-passing and red light running in the detection area is judged; more valuable road condition information can be provided for signal regulation and real-time scheduling of the command center; and under the development concept of the intelligent city construction, more diversified basic data can be provided for AI decision-making.

The invention relates to a pedestrian red light running detection method, and the method comprises the steps: obtaining a pedestrian image frame sequence of a target traffic scene, obtaining a continuous track point and a bounding box of a target pedestrian according to the pedestrian image frame sequence based on a YOLO target detection model and a DeepSort multi-target tracking algorithm, and obtaining a pedestrian red light running detection result. Obtaining a current coordinate and an instantaneous speed of the target pedestrian according to the bounding box and the continuous track point, determining a current state of the target pedestrian based on the current coordinate, the instantaneous speed and a preset stop line, determining a detection target according to the current state, responding to the detection target for traffic light detection, obtaining and identifying a traffic light image, and if the traffic light image is identified to be in a red light state, outputting the traffic light image. And if so, triggering red light running early warning. Through the method, the problem of inaccurate pedestrian red light running detection is solved, YOLO detection and DeepSort tracking are combined, the tracking stability under the scenes of shielding, short leaving of the visual field and the like is improved, the coordinates of the midpoint of the bottom of the bounding box serve as the coordinates of the pedestrian, and the situation that the pedestrian is easily interfered by the posture change of the pedestrian is avoided.

The present application relates to the field of data processing, and discloses a red light running judgment method, device and equipment of shared vehicle and storage medium.The method comprises the following steps: when obtaining the road condition video collected by the vehicle, detecting whether the red light appears for the first time in the road condition video; if the red light is detected for the first time, determining the pixel area corresponding to the red light in the road condition video and determining the center coordinates of the road condition video; according to the center coordinates and the pixel area, the target picture is intercepted in the road condition video; according to the target picture, it is determined whether the vehicle runs the red light; if it is determined that the vehicle runs the red light, the target picture is output.The present application reduces the difficulty of taking evidence of non-motor vehicle running the red light.

A traffic signal monitoring system and method belong to the field of intelligent traffic, and the system comprises a fusion processing module, an anomaly detection module, a control optimization module, a feedback optimization module and an anomaly recording module. According to the invention, the depth map neural network is adopted to carry out structured modeling on the intersection traffic map, the intersection feature information is dynamically extracted, the traffic jam index is calculated in real time, the accurate analysis and prediction of the traffic flow dynamic state are realized, and the intelligent level of the control strategy is improved. The motion trails of pedestrians, non-motor vehicles and motor vehicles in the video stream are analyzed based on a time sequence behavior recognition network, traffic violation behaviors such as red light running, retrogradation and road occupation are automatically recognized, and punishment suggestions are automatically generated in combination with a license plate recognition result. The current traffic state and the historical traffic mode database are utilized, the optimal signal timing strategy in different scenes is explored independently in combination with the deep reinforcement learning algorithm, real-time dynamic adjustment of the multi-phase signal lamp is achieved, and the road traffic efficiency is improved.

The present application relates to the field of traffic signal control, and particularly relates to an underground intelligent traffic light control system and method. An underground intelligent traffic light control method comprises the following steps: when the traffic light is in a road intersection prohibition phase, a traffic light signal containing a next traffic time period is broadcast to a roadway, and response information fed back by a vehicle is received, the response information comprising a vehicle identity, and vehicle position information when the vehicle receives the traffic light signal. The present application solves the problems of serious signal attenuation and large visual interference of drivers in the underground by dynamically matching signal transmission intensity and vehicle distribution, ensures stable signal reception by vehicles at a long distance, prolongs reaction time, and reduces the risk of red light running and collision; meanwhile, energy saving and consumption reduction, computing power optimization and intersection adaptive regulation and control are also achieved, and the safety, reliability and control efficiency of underground traffic operation are comprehensively improved.

The application relates to a traffic conflict risk assessment method combining multi-dimensional conflict indicators and a cloud model, which comprises the following steps: obtaining original trajectory data of a vehicle and cleaning the trajectory data; then, effectively identifying traffic conflicts based on a safety distance threshold and a time threshold; continuing to construct three-dimensional traffic conflict measurement indicators according to the effective traffic conflict trajectory data; determining the cloud model digital features of each conflict measurement indicator, establishing a traffic conflict severity assessment cloud model, calculating the mean value of the membership degree, and obtaining the weight of each conflict evaluation indicator in the comprehensive risk assessment by using the principal component analysis method; finally, based on the mean value of the membership degree and the weight value, the conflict is classified according to the weighted membership degree and the confidence degree criterion, and the traffic conflict severity classification is determined. The application can more comprehensively depict the real dynamic changes of the vehicle and the conflict risk severity, provides an efficient data processing flow and a modeling framework for the non-motor vehicle red light running conflict risk assessment, and provides scientific theoretical support and technical reference for the urban intersection traffic safety optimization and the control strategy formulation.

The invention discloses an intelligent traffic behavior recognition method based on edge calculation. The method comprises the following steps: S1, obtaining a target detection data set; s2, constructing a target detection model based on a YOLOV5 architecture, and training the target detection model to obtain a trained target detection model; s3, performing target detection on each frame of image in the video stream based on the trained target detection model, and constructing a behavior recognition data set based on a detection result; s4, constructing a traffic behavior recognition model based on the convolutional neural network, and training the behavior recognition model by using the behavior recognition data set to obtain a trained behavior recognition model; and S5, collecting a section of road monitoring video stream containing traffic lights by using monitoring equipment, uploading the road monitoring video stream to the edge server, and performing traffic behavior recognition by using the trained target detection model and behavior recognition model. According to the invention, traffic behaviors of red light running of non-motor vehicles and pedestrians can be effectively identified.

The application provides a method for detecting and processing red light running based on an electric bicycle helmet and an electric bicycle. Real-time state information of the electric bicycle is obtained. When the electric bicycle is in a riding state, a camera loaded on the front of the helmet of the electric bicycle is controlled to enter a snapshot mode, and riding environment image information obtained by the camera is obtained at a first preset time interval. When two conditions are met in the riding environment image information, i.e., a preset red light image and an included angle between a shooting direction of the camera and a current driving direction of the electric bicycle on a horizontal plane is less than a preset angle, a first preset alarm instruction is executed. The camera on the helmet can detect the red light image, and timely early warning and reminding are performed to avoid red light running of the user as much as possible during riding, improve travel safety, and have the advantages of high timeliness and high detection efficiency.

The invention relates to the technical field of intelligent driving, in particular to an automatic intelligent safe driving assistance system. An automatic intelligent safe driving assistance system is characterized by comprising an automobile data recorder, a cloud background, a fixed camera and a data interaction module, the automobile data recorder is adaptive to various automobile types, has the functions of network connection, video acquisition and automobile positioning, and is used for acquiring image data and positioning data in real time and uploading the data to the cloud background; the cloud background is used for receiving multi-source data, constructing a shared neural network, calculating a blind area and dead angle coverage rate and analyzing road condition information at the same time; and the fixed camera is erected in a road area where a blind area is easy to form. The beneficial effects of the invention are that the system breaks through the limitation of single-vehicle sensing, constructs a global shared sensing network through integrating the multi-source data of social vehicles and fixed cameras, eliminates the blind areas and dead angles of traditional single-vehicle sensing, and especially can cope with the complex scenes of bridge pier shielding, ghost probe, red light running at intersections and the like.

The invention belongs to the technical field of intelligent traffic and automatic driving, and discloses a red light running event classification early warning method under a vehicle-road cloud integrated system, which comprises the following steps of: S1, constructing an iteration triggering mechanism based on a distance gradient, and setting a judgment process as multi-time iteration prediction advancing along with the distance by utilizing a dynamic physical boundary; s2, multi-dimensional features are extracted, and combined working conditions are recognized; s3, constructing a working condition adaptive neural network prediction model, and adaptively adjusting attention weights of different feature dimensions in the model according to the combined working condition by using an attention mechanism; and S4, executing stepped iterative operation based on total element fusion, executing multi-target conflict resolution calculation at close-range nodes, and finally outputting probability distribution of three types of intentions of'intentional red light running ', 'unintentional red light running' and'forced red light running '. Through multi-dimensional feature deep fusion and sequential reasoning, accurate prediction of the red light running event in a complex interaction scene is realized, and the traffic efficiency and safety of the signalized intersection are improved.

The utility model discloses an attached intersection steerable non-motor vehicle red light running monitoring device which comprises a monitoring equipment body, the monitoring equipment body is detachably connected with a universal spherical holder, a shaft sleeve is arranged on the lower side of the universal spherical holder, the shaft sleeve is connected with a support jackscrew through a fastening screw, and a support is fixedly connected with the right side face of a connecting plate. The connecting plate is in bolted connection with the two double-bolt bow-shaped hoops through the first bolts, the number of the first bolts is four, the double-bolt bow-shaped hoops are in bolted connection with the double-bolt C-shaped hoops through the fastening bolts, and the number of the fastening bolts is four. The non-motor vehicle red light running monitoring device is specially designed for non-motor vehicle red light running monitoring, is of an attached type, is low in cost, adopts the universal spherical holder, realizes accurate angle locking of 360 degrees in the horizontal direction and + / -90 degrees in the pitching direction, has the error smaller than or equal to 1 degree, is comprehensive in monitoring angle, and meets the multi-direction monitoring requirement of an intersection.

The application discloses an Internet of Things data fusion and cooperative control management system based on an edge computing gateway, relates to the technical field of edge computing, and sets a data collection module to monitor illegal behaviors of red light running and overspeeding in road traffic in real time; and saves the monitoring picture in an edge data storage module of each region; sets an edge data processing module to identify human feature data and vehicle feature data in illegal behaviors; sets a regional data storage module to pre-save identity information and vehicle information of permanent urban residents; and sets a regional data query module to query real identities of personnel corresponding to the human feature data and the vehicle feature data in the city or across cities; and the problems of slow processing speed caused by large processing amount of traffic illegal data are solved.