Initialization method of UAV target tracking frame based on computer vision
A technology of computer vision and initialization method, which is applied in the field of computer vision, and can solve problems such as useless background information of the initial frame interfering with useful information, insufficiency, and affecting accuracy
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Publication Date
- 2020-03-31
Smart Images

Figure 1 
Figure 2 
Figure 3
Abstract
Description
Technical field
[0001] The invention relates to the technical field of computer vision, in particular to a method for initializing a UAV target tracking frame based on computer vision. Background technique
[0002] Target tracking is an important technology in the field of computer vision, and has important applications in military, medical, surveillance, and human-computer interaction. It combines comprehensive application technologies of many different technologies such as image processing, pattern recognition, artificial intelligence and automatic control, and is widely used in various fields. For target tracking, there are many tracking algorithms, and most of them use the tracking frame method for tracking.
[0003] In the process of tracking the target by the tracking frame, the tracking frame must be updated in each frame. And the initial tracking frame is often used as a matching template in the algorithm, so the initial tracking frame has an important influence on the su...
Examples
Embodiment Construction
[0043] The present invention aims to propose a method for initializing a UAV target tracking frame based on computer vision, which solves the problem that the traditional technology adopts the method of artificially selecting the initial tracking frame. The initial frame may contain too much useless background information and cause interference or contain insufficient useful information. The question of accuracy.
[0044] The core idea is: after the user manually selects the initial frame, the initial frame is adaptively enlarged to generate the filled frame, and the outline response graph of the filled frame is generated, and then the initial frame is slided in multiple sizes within the filled frame to produce different sizes Finally, after filtering these windows, the edge response graph of the digital image is used to analyze the contour density in different windows, and the score is sorted to select the best target tracking frame.
[0045] The solution of the present invention ...