Vehicle Tracking Method and System
A tracking system and vehicle technology, applied in the direction of instruments, calculations, character and pattern recognition, etc., can solve the problems of limited memory, difficult to guarantee real-time performance, and difficult to apply smart devices, so as to improve the accuracy rate and reduce the false alarm rate Effect
Active Publication Date: 2019-12-17
开易(深圳)科技有限公司
View PDF4 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Its processing flow is as figure 1 As shown. The disadvantage of this method is that it is difficult to guarantee real-time performance, especially when multiple vehicles need to be tracked, it is difficult to apply on smart devices
[0003] The current ADAS system generally runs on an intelligent hardware platform, and resources such as CPU and memory are relatively limited.
Method used
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment Construction
[0029] The main problem to be solved by this Fang Ming is to ensure that the tracking effect of the vehicle is stable and accurate on the smart device on the basis of meeting the real-time requirements, and greatly reduce the drift of the tracking target.
[0030] The field that the present invention relates to is the advanced driving assistance system (ADAS) in the field of computer vision, which specifically includes front vehicle collision warning, lane departure warning and the like.
[0031] The invention discloses a vehicle tracking method, comprising the following steps:
[0032] Detection step: detect new vehicles that appear in the field of view, and detect vehicles that have been in the field of view for the tracking step to correct the tracking results;
[0033] Tracking step: track each detected vehicle;
[0034] Fusion step: the detection result of the detection step is fused with the tracking result of the tracking step, and the fusion step is run at the current...
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More PUM
Login to View More Abstract
The present invention provides a vehicle tracking method and system. The vehicle tracking method includes: a detection step: detecting a vehicle newly appearing in the field of view, and detecting a vehicle that has been in the field of view for use in the tracking step to correct the tracking result; the tracking step: Track each detected vehicle; fusion step: fuse the detection result of the detection step with the tracking result of the tracking step, and the fusion step runs in the current frame of the detection step; verification step: check the similarity of the tracking of two adjacent frames The verification step is run after the tracking step is run; the regression correction step: perform regression correction on the tracking result of the vehicle; the detection step is run once at a specified time interval, and the tracking step is run once at a specified frame interval. The beneficial effects of the present invention are: the vehicle tracking method and system of the present invention can solve the real-time and accuracy problems that must be solved when the ADAS runs on the smart device, effectively reducing the false alarm rate and improving the accuracy rate.
Description
technical field [0001] The invention relates to the technical field of vehicle tracking processing in front vehicle collision warning, in particular to a vehicle tracking method and system. Background technique [0002] Because vehicle detection algorithms are relatively slow, vehicle tracking technology is indispensable. Traditional object tracking algorithms such as Mean-shift and Cam-shift algorithms are relatively fast, but it is common to track target drift in complex scenes. Excessive tracking drift may cause frequent false alarms in advanced driver assistance systems (ADAS) in the field of computer vision, deteriorating user experience. The existing object tracking algorithm with better performance is a tracking method based on detection, which continuously updates the "significant feature points" of the tracking module and the target model and related parameters of the detection module through an improved online learning mechanism, so that the tracking effect is mor...
Claims
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More Application Information
Patent Timeline
Login to View More Patent Type & Authority Patents(China)
IPC IPC(8): G06K9/62
CPCG06F18/24133G06F18/253
Inventor 刘鹏
Owner 开易(深圳)科技有限公司