17 results about "Traffic camera" patented technology

The invention discloses a non-motor vehicle passenger-vehicle identification method and system in a crowded scene, and belongs to the technical field of intelligent traffic management. In order to solve the problem of accurate recognition of non-motor vehicles, persons and vehicles in a crowded scene, image data of a traffic intersection is collected through a traffic camera, target detection processing is performed on the image data by using a target detection model, and a person target and a non-motor vehicle target are detected to obtain a person target frame and a non-motor vehicle target frame; judging whether the scene category to which the image data of the traffic intersection belongs is a crowded scene or not to obtain crowded scene image data; personnel key points and non-motor vehicle key points in the crowded scene image data are extracted; associating the personnel key points with the non-motor vehicle key points to determine personnel key points associated with the non-motor vehicle key points to obtain a personnel-vehicle association target; and identifying whether the person-vehicle associated target has traffic violation behaviors or not through an image tracking technology, and if so, identifying the identity information of the person target for traffic violation early warning and processing.

The invention discloses a road construction area intelligent sealing control and scheduling method fused with edge calculation, and relates to the technical field of road construction traffic management, and the method comprises the steps: S1, data collection and fusion: deploying edge calculation nodes at the periphery of a road construction area, and enabling the edge calculation nodes to access and fuse data collected by a multi-source sensor in real time, the data comprises traffic flow data acquired by a traffic camera, traffic flow speed data measured by a millimeter wave radar, global positioning system (GPS) positioning data of a construction vehicle, and visibility and precipitation data acquired by a weather sensor. By fusing edge calculation and multi-source sensor data, the traffic flow change can be responded in real time, the sealing control range is shrunk in the peak to reduce congestion, and road resources are reasonably adjusted and prevented from being idle in the flat peak; the device can adapt to weather condition changes, a safe buffer area is expanded in low-visibility environments such as rainstorm and heavy fog, and the traffic capacity is recovered in time when the weather turns good.

The utility model relates to a device facilitating disassembly, maintenance and cleaning of traffic cameras and an information acquisition system. The traffic camera dismounting, maintaining and cleaning device comprises a base, a transmission shell is fixedly installed at the top end of the base, a stand column is fixedly installed in the middle of the top end of the transmission shell, and a transverse rod is transversely arranged on one side of the top end of the stand column; the hanging bracket is fixedly mounted at the bottom end of the cross rod, a camera body is movably mounted at the bottom end of the hanging bracket, and a positioning mechanism is arranged between the camera body and the hanging bracket; a winding mechanism is arranged between the stand column and the camera body, a spraying mechanism is arranged on the left side of the middle of the stand column, and a driving mechanism is arranged in the transmission shell. According to the utility model, the camera body is convenient to disassemble or maintain, the position of the lens is adjusted through the system of the camera body, so that the lens faces the nozzle and is convenient to clean.

A rainfall inversion algorithm based on transfer learning includes the following steps: dataset preparation, dataset classification, dataset preprocessing, and construction of a multivariate Gaussian mixture model featuring grayscale changes and saturation, further filtering candidate raindrops within the ROI region. The raindrops obtained after the above filtering steps are used to invert rainfall amounts using a Gamma distribution model based on the definition of raindrop spectra. This invention, due to the adoption of the above technical solution, has the following advantages: it can effectively invert rainfall amounts using videos captured by traffic cameras without relying on other professional rainfall monitoring equipment; it has stronger anti-interference performance in complex environments; it reduces the probability of misjudging rainfall in sunny environments; and it has greater practical significance and application value.

The invention provides a traffic violation behavior identification method based on multi-source data fusion. The method comprises the following steps: acquiring multi-source data in a vehicle driving process through a plurality of sensors; fusing the multi-source data according to a preset rule, and outputting the fused multi-source data; and based on the fused multi-source data, identifying the traffic violation behavior of the vehicle. According to the invention, the data of the recorder, the traffic camera and the V2X communication equipment are fused, so that the problems of limited identification precision, especially high false alarm rate in a complex traffic environment and the like due to the fact that most of the prior art depends on a single data source are solved.

This application relates to the field of information technology and discloses a multi-source heterogeneous data fusion and analysis platform for smart cities. It includes a data acquisition end error calibration module, a transmission link integrity assurance module, a multi-source data fusion and cleaning module, and a dynamic timeliness management module. Each module achieves data flow through a unified data interaction protocol. The acquisition end is used to collect data from traffic cameras and environmental monitoring sensors. The data acquisition end error calibration module is used to calibrate errors generated by the data collected by the acquisition end. The acquisition end error calibration module also includes an environmental adaptive sensing unit and a device health assessment unit. Through the synergistic effect of the environmental adaptive sensing unit and the device health assessment unit, data reliability is improved from the source. The environmental adaptive sensing unit uses a pure hardware electrical signal transmission link to achieve real-time response to changes in visibility and illumination.

The invention discloses a method and a system for predicting red light running behaviors of pedestrians in a sparse scene, and belongs to the technical field of intelligent traffic control. In order to solve the problem that pedestrian red light running behaviors in a sparse scene are judged in advance, a first area and a second area are arranged in a traffic camera acquisition range of a traffic intersection; performing real-time target detection processing on the first area, performing static behavior detection on the pedestrian by using a behavior detection model to determine a target pedestrian, and obtaining an image sequence of the target pedestrian in the first area and an image sequence of the target pedestrian in the second area; performing trajectory extraction to obtain trajectory data of the target pedestrian, and performing behavior extraction to obtain a behavior sequence of the target pedestrian; according to the speed trend of the trajectory data of the target pedestrian, calculating a behavior probability value of the target pedestrian to judge whether the target pedestrian is a red light running intention, and constructing a positive sample and a negative sample of red light running; and constructing a red light running prediction model to obtain a trained red light running prediction model for predicting the red light running behavior of the pedestrian in the sparse scene.

The invention discloses a traffic violation detection and insurance acceptance automation method, and relates to the technical field of intelligent traffic systems, and the method comprises the steps: synchronously collecting multi-source video stream data in a vehicle driving process through a vehicle-mounted camera and a traffic camera, the time-space consistency alignment of the video streams is realized based on the timestamp and the space coordinate mapping; transmitting the aligned multi-source video stream to a cloud server, and carrying out preprocessing, target detection and behavior analysis on the video stream by using a multi-modal analysis model based on deep learning to judge whether a traffic violation behavior exists or not; generating a violation record containing violation time, place, type and vehicle information according to the violation judgment result, and uploading the violation record to a cloud server for storage through a data compression and encryption technology; and based on the violation record, an insurance acceptance process of an insurance company system is automatically triggered through a standardized API interface, insurance auditing and compensation processing are completed, and a processing result is fed back to a user side and a traffic management department.

The invention provides a traffic monitoring video processing method and system, and the method comprises the steps: starting timing when a traffic camera located on a road is detected to start to work in real time, and carrying out the segmentation processing of a traffic monitoring video generated in real time at preset time intervals, so as to generate a plurality of corresponding sub-videos; the sub-videos are analyzed to detect a starting moment, an intermediate moment and a termination moment corresponding to the sub-videos in real time, and the starting moment, the intermediate moment and the termination moment all comprise specific numerical values; generating an encryption key corresponding to the sub-video in real time according to the starting moment, the intermediate moment and the ending moment based on a preset rule, and encrypting the sub-video through the encryption key to generate a corresponding encrypted video in real time; and generating a corresponding target folder in real time based on the encrypted video, and correspondingly storing the encrypted video into the target folder. According to the invention, the secrecy performance of the video can be improved, and the working efficiency is correspondingly improved.

The utility model discloses a traffic monitoring camera convenient to overhaul, which relates to the technical field of traffic monitoring systems and comprises a transverse mounting rod inserted into a longitudinal upright post. The camera assembly is installed on the transverse installation rod, the transverse installation rod is located on a road conveniently, traffic cameras are arranged and installed in the direction perpendicular to the driving direction of the road, road information can be monitored comprehensively and accurately, use is convenient and stable, and the longitudinal assembly can control the transverse assembly to rotate and ascend and descend. Therefore, subsequent overhaul, maintenance, replacement and other operations of the camera assembly are facilitated, climbing operation and road closing are not needed, and the problems that an existing monitoring camera is not provided with any lifting structure, climbing operation is needed during subsequent overhaul and maintenance and is troublesome, and the arrangement direction of the monitoring camera is generally perpendicular to the driving direction of the road, so that the safety of the monitoring camera is influenced are solved. Therefore, the road needs to be blocked during maintenance, and the traffic rate of the road is reduced.

The invention discloses a traffic signal control method, device and equipment based on a large language model, and belongs to the technical field of traffic control. Information data of information media are obtained, first road information of a target area road is collected, and the first road information comprises image data collected by a traffic camera and vehicle sensing data collected by a vehicle-mounted sensor; determining traffic fusion data according to the information data, the first road information and a preset large language model; predicting virtual traffic flow data of a target area road according to the traffic fusion data; road control data are generated according to the virtual traffic flow data, the road control data are data for controlling traffic signal equipment of the target area, and the situation of reducing road congestion is achieved.

A UAV data processing and management system is provided including an encoder broadcaster and data manager which provides end users with a single interface for searching and sharing any video or imagery source, from any device with no client software required. Data inputs to the resource can include UAV video, photos, traffic cameras, fixed surveillance systems, iOS and Android device pictures, and other data (e.g., texts, emails) and inputs from all forms of social media, such as Twitter, Instagram, Facebook, etc. The cloud-based manager is built with collaboration and sharing in mind while at the same time maintaining data privacy, security protection, chain-of-custody control and audit trail maintenance. Analytic tools are integrated accordingly as “plug-ins” or as a store of available app resources which are easily removed, added and customized based on user needs and system requirements and cost constraints.

This invention relates to a method and system for predicting pedestrian red-light running behavior in sparse scenarios, belonging to the field of intelligent traffic control technology. To address the problem of predicting pedestrian red-light running behavior in sparse scenarios, the invention establishes a first region and a second region within the acquisition range of traffic cameras at intersections. Real-time target detection processing is performed in the first region, and a behavior detection model is used to perform static behavior detection to identify target pedestrians, resulting in image sequences of the target pedestrians in the first region and the second region. Trajectory extraction is performed to obtain the trajectory data of the target pedestrians, and behavior extraction is performed to obtain the behavior sequences of the target pedestrians. The speed trend of the target pedestrians' trajectory data and the probability value of their behavior are calculated to determine whether there is an intention to run a red light, constructing positive and negative samples for red-light running. A red-light running prediction model is constructed, resulting in a trained red-light running prediction model used to predict pedestrian red-light running behavior in sparse scenarios.

The application provides a vehicle detection tracking method, device and equipment for quasi-real-time digital twinborn display, which comprises the following steps: collecting a video stream from a traffic camera; using a detection classification neural network to detect the detection frame of a vehicle in each frame of the video stream in real time, and obtaining the color classification of the vehicle type; matching the vehicle positions between different video frames based on the detection frame information of the vehicle, tracking the path sequence of the vehicle, and recording the vehicle identification serial number; projecting and converting the coordinates of the obtained path sequence to a virtual twinborn space based on the internal and external parameters of the camera, matching the projected trajectory, and setting a unified vehicle identification serial number; performing smoothing filtering and completion on the trajectory in the virtual twinborn space, and outputting the three-dimensional trajectory, vehicle type, color and vehicle serial number of the vehicle with constant delay for the vehicle animation display of the twinborn scene. The application provides tracking data to a digital twinborn system with sub-second delay, and the obtained vehicle driving trajectory is continuous and stable.

This invention discloses a spherical camera for traffic management, relating to the field of traffic camera technology. It includes a mounting base, a protective housing, and a detection controller. The mounting base is symmetrically fixedly connected to the outer wall of the protective housing. The detection controller is fixedly installed at the bottom of the protective housing. A moisture-proof cleaning mechanism is provided on the top of the protective housing. Through a transparent glass shell adapted to the size of the camera, and via the meshing of a beveled toothed plate and a helical gear, the transparent glass shell can be rotated, causing contaminated areas to enter the protective housing for cleaning. The cleaned areas then simultaneously rotate back to the shooting position, ensuring the camera's monitoring field of view is not obstructed throughout the process, guaranteeing continuous and stable video. The cleaning brush and dryer only act on the non-monitoring areas of the transparent glass shell that are rotated into the protective housing. Combined with precise transmission control of the drive wheel and conveyor belt, interference with the real-time monitoring image is avoided, ensuring the integrity of the video data required for traffic management.

An example method includes obtaining information about motion of a vehicle using video or image frames of the vehicle passing through a traffic intersection during a time window obtained by a traffic camera; obtaining road definition information indicative of a traffic lane at the traffic intersection, wherein the road definition information corresponds to a motion of the vehicle indicated in a video frame or image frame of the vehicle as obtained by the traffic camera; obtaining traffic signal light information of traffic signal lights at the traffic intersection, the traffic signal light information corresponding to a time window of the vehicle passing through the traffic intersection, the traffic signal light information indicating timing of the traffic signal lights; and associating the traffic lane with the traffic light by determining that the traffic signal is associated with the lane using the information about the movement of the vehicle, the road definition information, and the traffic light information.

The embodiment of the invention relates to a processing method and system for calibrating external parameters of traffic cameras. The method comprises the following steps: acquiring own vehicle coordinate sequence fragments and video image sequences corresponding to each camera through data acquisition; carrying out target detection based on the two-dimensional / three-dimensional visual target detection model to obtain a two-dimensional / three-dimensional detection sequence; carrying out two-dimensional target vehicle detection frame positioning on each two-dimensional detection sequence by using an image feature extraction model, and carrying out three-dimensional target vehicle detection frame positioning based on the two-dimensional target vehicle detection frame; recognizing two-dimensional image coordinates of the target vehicle based on the three-dimensional target vehicle detection frame; and performing external parameter calibration according to the internal parameter matrix, the depth scale factor, the image coordinate sequence and the self-vehicle coordinate sequence fragment corresponding to each camera. According to the invention, the labor cost can be reduced, the calibration efficiency is improved, the calibration real-time performance is improved, and the external parameter calibration precision and calibration quality can be stably improved.