Eureka AIR delivers breakthrough ideas for toughest innovation challenges, trusted by R&D personnel around the world.

Particle filtering face tracking method based on multi-feature fusion

A multi-feature fusion and particle filter technology, applied in the field of computer vision, can solve problems such as lack of spatial information, ineffectiveness, difficulty in overcoming background confusion, etc., and achieve the effect of improving robustness, improving accuracy, and good tracking effect

Pending Publication Date: 2022-01-11
WUHAN NARI LIABILITY OF STATE GRID ELECTRIC POWER RES INST +6
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, when the tracked target rotates or changes pose, the overall color histogram is ineffective for tracking the target due to the lack of spatial information
On the other hand, it is difficult to overcome the influence of background confusion and illumination changes on target tracking by using color features alone.

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
  • Particle filtering face tracking method based on multi-feature fusion
  • Particle filtering face tracking method based on multi-feature fusion
  • Particle filtering face tracking method based on multi-feature fusion

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0042]The present invention will be further described below in conjunction with accompanying drawing:

[0043] A particle filter face tracking method based on multi-feature fusion, comprising the following steps:

[0044] Step 1: Build an adaptive selection color histogram model by seeking the similarity degree of each sub-area of ​​the target area (ie, the area to be tracked);

[0045] The specific method of constructing the adaptive selection color histogram model is to calculate the color histogram of each sub-region and calculate the degree of similarity between them. The colors are very similar, then choose the overall color histogram to model the target color; otherwise, choose a multi-block color histogram.

[0046] Step 2: Based on the particle filter method for solving nonlinear and non-Gaussian problems, construct a visual tracking state model;

[0047] The specific method is: define x=[x, y, θ, Δx, Δy, Δθ] as the state variable, where (x, y) is the center position...

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 particle filtering face tracking method based on multi-feature fusion. The method comprises the following steps: constructing an adaptive selection color histogram model and a visual tracking state model; constructing a visual tracking observation model based on color and edge direction features, and obtaining a dynamic model equation from the ith particle state at the k-1 moment to the ith particle state at the k moment; constructing a multi-source information fusion model to obtain an observation equation of a system state vector Xk total at a k moment; calculating a color histogram and a texture direction histogram of the target area; initializing the particle filter, and carrying out resampling operation; performing further prediction on the particle group through a dynamic model equation to obtain a new particle group, performing weight updating according to an observation value, calculating the weight of the particles by using the obtained observation equation, and performing normalization to obtain mean estimation of a particle weight output target state. The invention has the advantages of high robustness and capability of overcoming the influence of rotation or attitude change, background confusion and illumination change on target tracking.

Description

technical field [0001] The present invention relates to the field of computer vision, in particular to a particle filter face tracking method based on multi-feature fusion. Background technique [0002] Object tracking is a popular research topic in the field of computer vision. Its main task is to detect, recognize and track moving objects from image sequences, and even understand and describe the behavior of target objects. Object tracking is widely used in human motion recognition, video surveillance, video retrieval, virtual reality and human-computer interaction. The description or representation of object features is a key part of object tracking. Commonly used features include color, edge, optical flow, and texture. [0003] Among them, color features are widely used to track moving objects due to their advantages of partial occlusion, strong rotation robustness, and scale invariance. In the process of using color features for object tracking, the color histogram ...

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): G06T7/246G06T7/44G06T7/90
CPCG06T7/251G06T7/44G06T7/90G06T2207/10016G06T2207/20024G06T2207/30201G06T2207/30241
Inventor 史会轩陈少兵贾勇勇李永飞陶家贵王韬孔凡胜陈敏方璟孙宏志佘帆陈新波曹祖亮
Owner WUHAN NARI LIABILITY OF STATE GRID ELECTRIC POWER RES INST
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Eureka Blog
Learn More
PatSnap group products