A video object mutual occlusion processing method that integrates object appearance model and game theory
An appearance model and occlusion processing technology, which is applied in image data processing, image analysis, image enhancement, etc., can solve problems such as similar target occlusion, occlusion target interference, etc.
Active Publication Date: 2019-07-05
ZHEJIANG UNIV OF TECH
View PDF4 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
[0004] In order to overcome the problem that the existing mutual occlusion processing method based on the target appearance model cannot solve the problem of similar target occlusion and interference of other occluded targets, and to ensure that all targets in the occluded area can be tracked accurately, the present invention proposes a more robust target Appearance model, this model considers the spatial constraints and anti-interference factors of the target appearance at the same time, and combines the game theory method to solve the optimal search and matching problem of all targets in the occluded area, thereby improving the accuracy and precision of video target tracking
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
[0051] The present invention will be further described below in conjunction with the accompanying drawings.
[0052] refer to Figure 1 to Figure 5 , a video target mutual occlusion processing method that integrates the target appearance model and game theory, comprising the following steps:
[0053] 1) Occlusion area prediction, the process is as follows:
[0054] like figure 1 As shown in (a), the predicted position of the target at time t-1 (or ) as the center of the circle, construct the circumcircle C of the target bounding box i (or C j ), with a radius of (or in, (or )and (or are the predicted target states (or ), and ||·|| is the Euclidean norm (the same below). if and only if When , as shown in formula (1), the areas where the targets i and j are located are predicted as candidate occlusion areas, otherwise, no occlusion is considered to occur:
[0055]
[0056] 2) The occlusion area is determined, the process is as follows:
[0057] In...
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 discloses a method for processing video target mutual occlusion by combining a target appearance model with a game theory. The method comprises the following steps: 1) occlusion area prediction; 2) occlusion area determination; 3) target appearance model establishment; and 4) game theory based occlusion target tracking: 4.1) occlusion area target function establishment, 4.2) multi-person non-cooperative game establishment, and 4.3) seeking Nash equilibrium. According to the method disclosed by the present invention, a more robust target appearance model is proposed, and the model takes both the spatial constraints of the target appearance and the anti-interference factors into account, and is combined with the game theory method to solve the optimal search matching problem of all targets in the occlusion area, so that video target tracking accuracy and precision are improved.
Description
technical field [0001] The invention belongs to the technical field of video object tracking, and in particular relates to a multi-video object tracking method and an occlusion processing method. Background technique [0002] Multi-moving target video tracking is one of the key technologies of intelligent video surveillance system. For each frame in the video sequence, the system should be able to automatically and accurately obtain the position, speed and size of each moving target in the image. At present, researchers have proposed many effective multi-target video tracking methods. However, how to effectively solve the data association problem between the target and the measurement value and the mutual occlusion problem between the targets is still the focus and difficulty of current research. [0003] Compared with the traditional target tracking method, the method based on GM-PHD filter can effectively avoid the complex data association process. However, the existing...
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): G06T7/292
CPCG06T2207/10016G06T2207/30232
Inventor 周小龙产思贤邵展鹏林家宁陈胜勇
Owner ZHEJIANG UNIV OF TECH