Unlock instant, AI-driven research and patent intelligence for your innovation.

Correlation filtering tracking method based on response graph confidence region self-adaptive feature fusion

A technology of feature fusion and confidence region, applied in image analysis, image data processing, instruments, etc., can solve problems such as difficulty in uniformly adapting to different video sequences, and difficulty in uniformly adapting to different video frames.

Active Publication Date: 2019-06-25
YUNNAN UNIV
View PDF8 Cites 1 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0008] Therefore, in the target tracking system for actual complex scenes, since the target and background are constantly changing, the fixed weight fusion method is not only difficult to uniformly adapt to different video sequences, but also difficult to uniformly adapt to different video frames of the same video sequence.

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 more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Correlation filtering tracking method based on response graph confidence region self-adaptive feature fusion
  • Correlation filtering tracking method based on response graph confidence region self-adaptive feature fusion
  • Correlation filtering tracking method based on response graph confidence region self-adaptive feature fusion

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0067] Embodiment 1: as Figure 1-3 as shown,

[0068] According to the technical solution of the present invention, a correlation filter tracking method based on adaptive feature fusion of response map confidence regions selects Shaking video sequences for tracking, which has five functions: illumination change, scale change, background confusion, out-of-plane rotation, and in-plane rotation. a challenging attribute. Proceed as follows:

[0069] Step 1. Input the first frame

[0070] Select the Shaking video, where the video frame width and height are Width=624 and Height=352. Select the rectangular area (225, 135, 61, 71) of the target to be tracked in the first frame, that is, the selected tracking target such as figure 2 Indicated by the blue rectangle. Among them, (225,135) is the coordinates of the upper left corner of the rectangular area, and (61,71) is the width and height of the rectangular area.

[0071] Step 2. Initialize the target template

[0072] 2.1 Ca...

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

No PUM Login to View More

Abstract

The invention discloses a correlation filtering tracking method based on response graph confidence region self-adaptive feature fusion. The method is characterized in that on the basis of a correlation filtering tracking framework, two complementary features of a HOG and a color histogram are adopted for feature extraction, and fusion parameters of the two features are adaptively set according toa confidence region of a response graph in a specific scene of each video frame. A multi-feature fusion parameter is adaptively set according to a response map confidence region of each video frame topromote the stability of a tracking system.

Description

technical field [0001] The present invention relates to the field of video filter tracking methods, in particular to a correlation filter tracking method based on adaptive feature fusion of response map confidence regions. Background technique [0002] Video object tracking is one of the research hotspots in the field of computer vision, and its purpose is to estimate the position of the object in the video image sequence. This technology plays an important role in many applications such as video surveillance, human-computer interaction, robotics and driverless cars. Real-time and stability are the two goals of the target tracking system. The convolution principle shows that the time-consuming convolution operation can be converted into an element-wise dot product operation in the Fourier transform domain. The correlation filter technology based on the convolution principle is introduced into the target tracking, and its extremely high processing speed meets the real-time ...

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
no application Login to View More
Patent Type & Authority Applications(China)
IPC IPC(8): G06T7/277G06T7/246G06K9/46G06K9/62
Inventor 高赟赵江珊张学杰
Owner YUNNAN UNIV