Moving target tracking method based on kcf and human brain memory mechanism

A moving target and memory technology, applied in the field of computer vision, can solve the problems of KCF algorithm tracking accuracy decline, lack of adaptive model update strategy, and inability to deal with complete target occlusion, achieve fast tracking speed, avoid the process of matrix inversion, Strong anti-occlusion ability

Active Publication Date: 2021-06-01
BEIJING INSTITUTE OF TECHNOLOGYGY
View PDF4 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] On the other hand, although the KCF algorithm has advantages, it also has certain limitations: first, it cannot deal with the problem of complete occlusion of the target; second, it lacks an adaptable model update strategy
Therefore, when the moving target has a sudden change in attitude, the target reappears after a short disappearance, or the target is occluded, the tracking accuracy of the KCF algorithm will drop significantly, or even fail to track.

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
  • Moving target tracking method based on kcf and human brain memory mechanism
  • Moving target tracking method based on kcf and human brain memory mechanism
  • Moving target tracking method based on kcf and human brain memory mechanism

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0053]The target tracking method based on KCF and human brain memory mechanism is disclosed in this example, and its overall process is attached.figure 1 As shown, the specifically includes the steps of:

[0054]Step 1: Initialize the memory space and tracking window.

[0055]Initialize the memory space, first, build three memory spaces to save the target matching template, respectively Q.t, Parameter α of the classifiertAnd classifier target template XtAnd each memory space includes short-time memory space and long memory space, respectively. Then, one instantaneous space is established to save the current frame target data, that is, the estimated template.

[0056]Enter the first frame of the video, determine the initial target tracking window, determine the initial target can be selected or determined according to the initial position data of the target.

[0057]Step 2: Calculate the feature of the tracking window.

[0058]The HOG feature of the current frame tracking window is calculated. Firs...

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 moving target tracking method based on KCF and human brain memory mechanism. The visual information processing cognitive model based on the human brain memory mechanism is introduced into the target template update process of the KCF algorithm, and the memory-based Template Updating (MTU) method forms a new model update strategy, so that each Each template has to be transmitted and processed through the three spaces of short-term memory, short-term memory and long-term memory. During the target tracking process, the target template is updated according to different update strategies according to the matching degree between the current frame target template and the target template in the memory space. By effectively memorizing the previous scene, the method can continue to track robustly when the current target has a sudden change in attitude, reappears after a short disappearance, or occlusion occurs.

Description

Technical field[0001]The present invention relates to a tracking method of moving target in an image sequence, specifically, a moving target tracking method based on a nuclear correlation filtering algorithm (KCF, Kernelized Correlation Filter) and a human brain memory mechanism, belonging to computer vision technology.Background technique[0002]Sports Target Tracking is an important research direction in the field of computer visual, widely used in intelligent security, visual monitoring, human-machine interaction. In recent years, although the target tracking technique has made great progress, it is still difficult to achieve accurate target tracking under complex conditions such as light changes, target geometric deformation, target occlusion, rapid movement.[0003]At present, the target tracking method based on the discrimination has become mainstream, including: Structure Output Tracking With Kernel, Struck Tracking Method, Tracking - Learning - Detection, TLD) Tracking Method, V...

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 Patents(China)
IPC IPC(8): G06K9/00G06K9/46G06K9/62G06T7/246
CPCG06T7/251G06V20/42G06V10/443G06F18/2453
Inventor 宋勇赵尚男赵宇飞李云李旭
Owner BEIJING INSTITUTE OF TECHNOLOGYGY
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
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap