212 results about "Trajectory analysis" patented technology

The invention relates to a full automatic target detecting and tracking system in the computer vision field, wherein, an input module is responsible for collecting digital images shot by a binocular camera to be taken as system input, the obtained digital images are input into a feature extraction module and feature analysis is carried out to one image to obtain a plurality of characteristic points to be taken as the subsequently processed images. By matching the characteristic points of two images, the parallax of the two images is calculated, and by combining the pre-informed external and internal parameters of the camera, the lower coordinate of a camera coordinate system of the characteristic points can be calculated; furthermore, by the relationship between a world coordinate system and the camera coordinate system, the coordinate of the world coordinate system of the characteristic points can be known. A clustering module clusters the characteristic points into an aggregation for expressing target position, while a trajectory analysis module estimates the target position on a time sequence to obtain the motion trajectory of the target. The invention can effectively and steadily detect the targets in a designated area, track the targets and calculate the motion trajectories of the targets.

The invention relates to the technical field of movement behavioral analysis and prediction in a sense network environment, and specifically to a movement space-time trajectory analysis method in the sense network environment. The movement space-time trajectory analysis method in the sense network environment comprises data reception of receiving trajectory movement position data generated by a positioning device and resolving the data format into a data format applicable to data treatment; semantic treatment of performing clustering operation on the semantic trajectory data; space-time correlation of performing characteristic analysis and statistics on clustered semantic trajectory data in a time domain and a space domain, and performing time-space correlation analysis in combination with the time domain and the space domain; correlation similarity analysis of calculating space-time correlation similarity of the semantic trajectory and performing analysis and calculation on the correlation among different space domains and different movement objects; outputting a result. The movement space-time trajectory analysis method in the sense network environment solves the problem of continuous treatment and mutual correlation of time and space dimensions in a traditional transactional database, and meets the need from a sense network application service to real-time analysis of trajectory movement data.

The present invention relates to a technical field of traffic monitoring, and more particularly to a method for detecting traffic violation. The present invention includes firstly localizing vehicle salient parts through salient features including vehicle license numbers and vehicle rear lights, and representing a vehicle with the vehicle salient parts, then tracking the vehicle with a Kalman filter based on the vehicle salient parts, and finally detecting vehicle violation through moving trajectory analysis and setting violating detecting areas. The present invention solves vehicle violation detection problems in complex engineering application conditions such as illumination change and detection noise, and is suitable for city traffic management under complex conditions.

The invention discloses a carpooling method and a carpooling system based on trajectory analysis. People going the same way are found based on analysis of historical driving trajectories, so as to realize high-efficiency carpooling in daily life. The technical scheme is as follows: a server receives a carpooling request initiated by a terminal; the server looks for people going the same way through trajectory analysis based on stored historical driving trajectories of all users and carpooling parameters of the carpooling request; and the server confirms whether a person going the same way agrees to share a car, and carries out information interaction between a user initiating the carpooling request and a confirmed person going the same way.

The invention discloses a user trajectory analysis method and system. The method comprises the steps that the signaling data of a user are acquired, wherein the signaling data include time information, space information and the user ID; user trajectory point information corresponding to each user ID is obtained according to the signaling data, wherein the user trajectory point information includes trajectory point acquisition time, the located cell ID, the cell longitude and the cell latitude; the abnormal data in the user trajectory point information are eliminated; and mobile point pairs in the user trajectory points after elimination of the abnormal data are determined so that mobile trajectory sequences are obtained. The accuracy of user trajectory analysis can be enhanced by the user trajectory analysis method and system.

The invention relates to video processing, mode identification and the like. In order to provide a video abnormal event detection method by which a multi-motion parameter variation rule of the same event in a plurality of visual angles of a cross camera can be represented and of which the accuracy is high, the technical scheme of the invention is that: a method for detecting a monitored video abnormal event based on trace analysis comprises the following steps of: 1, classifying abnormal events based on the trace analysis; and 2, modeling the abnormal events based on a dynamic Bayesian network, wherein the modeling is finished by the following three steps of: a, constructing the dynamic Bayesian network which consists of three levels from bottom to top, namely a characteristic layer, an element layer and an event layer; b, learning the dynamic Bayesian network in the following three situations: firstly, the labels and the number of elements are known; secondly, the number of the elements is known and the labels are unknown; and thirdly, the number of the elements and the labels are unknown; and c, selecting a characteristic. The method is mainly applied to video processing, mode identification and the like.

Embodiments of the invention provide an abnormal event detection and early warning system based on analysis of a target abnormal behavior trajectory. The system includes a target behavior trajectory acquisition device, a PTZ camera and a server. The target behavior trajectory acquisition device acquires a behavior trajectory of a target and then transmits real-time data to the server, the server performs comparison on the real-time data by a trained abnormal event alarm model, an alarm message is sent when it detects that a target behavior or trajectory is abnormal, the PTZ camera collects animage of an area where an abnormal event occurs to confirm the alarm message, a real-time picture of a road collected by the PTZ camera is stored in the server, the server sends the verified alarm message to a monitoring center, and the monitoring center confirms the alarm message sent by the server and displays the alarm message to give an alarm to the public or a third-party system. The system solves the problems that the detection precision of an existing abnormal event detection system is low, a part of event types cannot be detected, and early warning cannot be performed in advance.

The invention discloses a target detection recognition algorithm based on 3D (Three-dimensional) trajectory analysis. In the method, a binocular camera installed above the rear door of a bus is used for collecting a video image to carry out target detection recognition, the motion positions of the same passenger at different moments are taken as movement targets, the motion targets are subjected to preliminary recognition through the depth map of a video screen image, then, a pseudo-motion target is removed, and finally, the accurate detection recognition of the motion target is realized. Through the 3D trajectory analysis, target recognition detection is carried out, trajectory tracking failure caused by illumination in a trajectory tracking process in the prior art is solved, and the accuracy of target recognition detection is improved. The target detection recognition algorithm is free from the restriction of a hardware environment, can carry out real-time detection and is high in detection accuracy. The two-dimensional trajectory of the image of the motion target is converted into a 3D trajectory under a world coordinate system, the most common problem in traditional target recognition that the target can not be detected since the target is covered is solved, and the accuracy of the target recognition detection is improved.

The invention discloses a central support system and method for a quick service response. The system comprises an intelligent service appeal determination module, a service trajectory analysis module, and a marketing-distribution cooperation application data sharing module. In addition, the method includes: historical user information is invoked from an electricity information acquisition system based on user request information, a correlation analysis is carried out on the user based on the historical user information, and a service corresponding to a current user is generated according to an analysis result; service trajectory, user behavior, service channel and hotspot task analyses are carried out on the service successively to obtain at least one to-be-processed work order corresponding to the current user; and the work order is classified and send the classified work order is sent to a distribution network automation system, a production management system or a marketing system according to the type of work order. Therefore, the service determination system support is enhanced, so that service determination in the power system becomes intelligent and accurate; and a service determination result is obtained intelligently, so that a basis is provided for formulating a differentiated service strategy.

The invention provides a pedestrian event detection method based on shape features and trajectory analysis. A foreground target is obtained through target segmentation by means of background differencing, a connected domain of the same target is marked through a block-based method, meanwhile an enclosing rectangle of the connected domain is recorded and geometric shape characteristics of the enclosing rectangle are extracted, and target recognition is finished. When a target similar to a pedestrian is recognized, corner points of the target are extracted, and the corner points are tracked and matched by means of corner point position information. The preceding processes are repeated, and the motion track of the target is obtained. Segmented inflection points of the track are obtained, and linear analysis is carried out in each segment where the inflection points are formed, so that the speed of the target is obtained. Therefore, pedestrian and event state information is analyzed, and traffic safety warning is finished. The detection method is suitable for complex and changeable traffic scenes, can precisely recognize, track and early warn the pedestrian appearing in a monitoring video range, is high in practical value, and has wide application prospects.

The invention provides an automatic highway parking detection method which comprises the following steps: 1) range setting; 2) target detection; 3) target tracking; 4) trajectory analysis; and 5) monitoring scene background update. By analyzing the track of a vehicle target, three kinds of events, which are abnormal parking event, abnormal parking relieving event and normal driving event, are obtained through classification, and thus behaviors of a vehicle can be tracked and analyzed better. Through trajectory analysis of the vehicle target, different background model update strategies and update rates are adopted for happening regions of different kinds of events, thereby improving accuracy of target detection, improving accuracy of abnormal events and reducing false alarm rate.

The invention relates to a personnel location trajectory analysis system and analysis method based on a WeChat public platform. The system comprises a background management center, a WeChat public account, a plurality of two-dimension codes with geographical location information and a plurality of handheld intelligent mobile terminals loaded with WeChat client sides. Common users are in data interaction with a communication mechanism in a WeChat public platform server and the background management center under WiFi or 3G/4G wireless network environment through the handheld intelligent mobile terminals and the WeChat client sides. Based on existing mobile phone terminal equipment, WeChat platform software, the existing two-dimension code scanning function, the communication function and the geographical location information recording and uploading function, the mode of limiting the two-dimension codes is utilized, the characteristic of multi-point trajectory recording is kept, the possibility of inputting a large amount of invalid information is eliminated, system work efficiency is improved, the real geographical location information of specific personnel at the real moment is effectively recorded, the whole system is low in cost, and the important role in the field of electronic patrolling and guarding is achieved.

The invention discloses a vehicle retrograde motion intelligent detection method based on tracking track analysis. The method comprises the following steps: detecting a target vehicle in a video imageby utilizing a primary classifier and a secondary classifier; extracting a target vehicle area, distributing a nuclear correlation filtering tracker for the target vehicle; wherein each target vehicle is matched with one nuclear correlation filtering tracker, the nuclear correlation filtering tracker is used for obtaining a tracking area of the target vehicle, then the movement track growth direction of the target vehicle is obtained according to the tracking area and compared with the initially marked retrograde movement direction, and if the movement track growth direction is the same as the initially marked retrograde movement direction, the target vehicle retrograde movement occurs; And if not, the target vehicle does not run reversely. By means of the mode, real-time detection of vehicle retrograde running is achieved, and the problems existing in manual detection are solved; And meanwhile, the reliability of vehicle detection and recognition is improved by using the cascade classifier, and simultaneous tracking of multiple vehicle targets is realized.

The invention discloses a large data intelligent analysis platform oriented to the field of social security, comprises a relational network analysis module, which is used for constructing one or morethematic databases according to multi-source data, constructing a network topology structure based on the thematic databases, obtaining labels of each node in the network topology structure, and mining key nodes and/or community structures in the network topology structure according to a set scene task; The trajectory behavior analysis module is used for acquiring different types of trajectory data, generating different types of trajectory models according to the trajectory data, and performing trajectory analysis based on various trajectory models; The video intelligent analysis module is used for obtaining the multi-scale features of pedestrians according to the set pedestrian detection and tracking algorithm, extracting the semantic features, gait features and visual features of pedestrians according to the multi-scale features, so as to identify and search the pedestrians. The invention can effectively improve the pre-warning capability and the pre-warning level, and provides powerful technical support for the emergency response.

The invention discloses a gait feature extraction method based on human body gravity center track analysis. The method comprises the steps of: S1, gait detection and tracking: original video images are converted into single channel gray scale images, and the images are smoothed through Gauss filtering, and the images are further converted into binary images; S2, positions of human body gravity centers are calculated, and gravity center coordinates of the image of each frame in a moving object in motion are connected in a same coordinate system to obtain a gravity center track of a pedestrian; and S3, gait waveforms after being subjected to denoising are input into a computer, and a harmonic wave amplitude corresponding to a K subharmonic frequency is calculated through the utilization of a formula, and a corresponding frequency spectrogram is drawn and output. According to the invention, problems encountered in present feature extraction can be solved, and a one-dimensional gait characteristic vector can be extracted from the complex gait images, so that further analysis, training and identification are convenient.

The invention provides a pedestrian monitoring method and device based on distance image head and shoulder features. According to the method, a distance image is acquired, calculation is performed to obtain a height image, the background is filtered out by adopting a height threshold value, a pedestrian detection result is obtained through the head and shoulder height feature template matching and the color image textural feature template matching, and the passenger flow parameter extraction and event discrimination are performed through the pedestrian tracking and trajectory analysis. The device for performing pedestrian detection by adopting the algorithm comprises an integrated probe which is used to acquire the distance image and the color image at the same time and perform synchronous correction, a video image processing and central control unit, a cache module, a memory module, an encryption chip, a network transport module, a USB interface and a power module.

The invention relates to a fatigue driving safety monitoring method based on human eye state detection. According to the invention, various complex factors in the driving process are fully considered. Firstly, the situation that human eye data caused by accidental factors are not reliable is eliminated. Therefore, human eyes can be accurately positioned. After that, the current state of the human eyes can be judged from the visual level of the human eyes. In this way, the complex eyeball analysis technique and the trajectory analysis technique are not required.

The invention relates to a traffic monitoring method and a traffic monitoring system. The method comprises the following steps: S1, collecting video information of a road through a camera; S2, performing multi-target detection and identification on the video information acquired in the step S1 to detect vehicles and pedestrians in the video; S3, performing tracking and trajectory analysis on the multiple targets detected and identified in the step S2; and S4, according to the multi-target tracking and trajectory analysis result in the step S3, obtaining the speed and acceleration of the target, judging whether a traffic collision accident occurs or not, and if the traffic collision accident is judged to occur, automatically carrying out alarm processing. Video collection is standardized, and video sources can be effectively managed; automatic analysis of video structuralization is established, and post-event retrieval depending on manpower is liberated; and compared with an existing traditional target detection algorithm, the traffic monitoring method and the traffic monitoring system are high in detection precision and good in effect; and multi-stream video structural analysis canbe input at the same time, and results can be output.

The invention discloses a traffic video comprehensive analysis cloud platform system. The system comprises a host, a server, a monitoring device and a cloud platform system; the cloud platform systemis connected with the host, the server and the monitoring device; the cloud platform system comprises a system management system, a data storage system, an intelligent analysis system, a network transmission and communication system, a scheduling and monitoring system and an application platform system. According to the traffic video comprehensive analysis cloud platform system of the invention, existing video resources of existing traffic monitoring points, electronic polices, public security monitoring and the like can be fully utilized so as to realize automatic alarm, automatic analysis and snapping functions for traffic road conditions and motor vehicle rule violations; and on the basis of the application platform, vehicle density analysis, vehicle trajectory analysis, intersection rule violation statistics and the like can be performed on traffic road condition information and various rule violations, and therefore, sufficient basis and rich research data can be provided for traffic police in law enforcement and planning management.

The invention discloses a high-altitude parabolic detection method and device based on trajectory analysis, wherein the method comprises the following steps: acquiring multiple frames of images of a current monitoring video, moving object detection is carried out on the multiple frames of images; determining a moving object on each frame of image; and processing each frame of image of the determined moving object, determining a central coordinate of the moving object on each frame of image, performing linear fitting according to the central coordinate of the moving object on each frame of image to obtain a motion trail equation of the moving object, and verifying the motion trail equation of the moving object to determine the object of the high-altitude parabolic object. Motion detection is carried out on each frame of image. According to the method, the moving object on each frame of image is obtained, then the motion trail equation is verified according to the motion trail equation of the coordinates of the moving object on each frame of image, so that the object of high-altitude parabolic motion is determined; the high-altitude parabola can be actively detected; the problem thatdeep learning is difficult to identify is solved; and the detection accuracy is improved.

The invention relates to the technical field of on-board equipment, in particular to a vehicle trajectory analysis method and device. The method comprises the following steps of acquiring multiple vehicle positions reported by a vehicle to be analyzed at different moments; merging the multiple vehicle positions, and merging two vehicle positions with the distance being smaller than a first preset value into one vehicle position; generating a vehicle trajectory according to each vehicle position obtained through merging; displaying the generated vehicle trajectory. Therefore, repetitive point drawing can be avoided, and users can more clearly know the vehicle trajectory of the vehicle to be analyzed.

The invention discloses a driver lane-changing behavior analysis method based on an aerial photography video vehicle trajectory. Based on the image tracking technique, the movement video data of a vehicle, collected by the aerial photography of an unmanned aerial vehicle, are extracted and processed. After that, through the data fitting, the mapping and the conversion of coordinate systems, a corresponding movement track graph, a displacement space-time diagram in the direction of lanes and a lane deviation fluctuation diagram perpendicular to the direction of lanes are obtained. According tothe invention, a vehicle trajectory processing and conversion method is provided by the invention. By the method, vehicle trajectories in a video can be efficiently, conveniently and accurately extracted and processed. The method has important significance on the research and the development of trajectory analysis-based relevant traffic phenomena and driver characteristics.

The invention provides a multi-dimensional trajectory analysis method and device. The multi-dimensional trajectory analysis method comprises the steps: structuring multi-dimensional data and then storing the structured multi-dimensional data in a data warehouse; extracting a target trajectory and each associated trajectory from a data warehouse; dividing the target trajectory and the associated trajectories into trajectory segments; determining a space-time characteristic value of each track section; for an associated trajectory similar to the target trajectory, counting the number of trajectory segments with the same spatial-temporal characteristic values as the associated trajectory and the target trajectory, and taking the number as an associated value of the associated trajectory; forassociated trajectories different from the target trajectories, converting the target trajectories into feature trajectories similar to the associated trajectories, and counting the number of trajectory segments with the same spatial-temporal feature values as the associated trajectories to serve as associated values of the associated trajectories; sorting the associated trajectories according tothe associated values; and determining the associated track ranked at the top preset position as the track of the target associated object. By applying the embodiment of the invention, the efficiencyand accuracy of multi-dimensional trajectory analysis are improved.