Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Correlation filtering target tracking method based on self-adaptive weight combined learning

A technology of correlation filtering and self-adaptive weight, applied in image data processing, instrumentation, computing, etc., can solve problems such as drift and failure of tracking results, and achieve the effect of enhancing robustness and improving capabilities

Active Publication Date: 2018-11-02
JIANGNAN UNIV
View PDF5 Cites 19 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, under complex conditions such as illumination changes, background clutter, and target deformation, traditional tracking algorithms based on correla

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 target tracking method based on self-adaptive weight combined learning
  • Correlation filtering target tracking method based on self-adaptive weight combined learning
  • Correlation filtering target tracking method based on self-adaptive weight combined learning

Examples

Experimental program
Comparison scheme
Effect test

Example Embodiment

[0031] The technical solution of the present invention will be further described below in conjunction with specific embodiments.

[0032] A related filtering target tracking method based on adaptive weight joint learning, the flowchart is as follows figure 1 As shown, the steps are as follows:

[0033] Step 1: Read in the first image Image 1 And tracking target initial rectangle information;

[0034] Step 2: Train the position-related filter

[0035] (2-1) For the position filter, in the framework of correlation filtering, around the center point of the first frame of image target, cyclically sample candidate samples in the candidate area according to the initial rectangular scale, and extract the directional gradient for each candidate sample Histogram of Oriented Gradient (HOG) feature to get the training sample set X 1 , Where each training sample is d is the characteristic dimension;

[0036] (2-2) Construct an objective function for the training sample x(m, n), and add a spatia...

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 target tracking method based on self-adaptive weight combined learning and belongs to the field of machine vision. According to the method, a correlation filtering model and a color model based on a color histogram are combined, the discrimination characteristics of the correlation filtering model are fully utilized to effectively distinguish a target and a background, and meanwhile histogram scores are acquired through the color model to better cope with occlusion, blockage, deformation and other complicated environments. In order to fully utilize the advantages of the two models, confidence weights are proposed to self-adaptively combine the two models. Meanwhile, when a correlation filter is trained, background information is fully utilized to construct spatial regularization items, interference of the background information is effectively suppressed, and the robustness of an algorithm in the tracking process is further improved. For the purpose of making a target model better cope with continuous changes of a target scale in the tracking process, a separate quick scale detection model is constructed.

Description

Technical field [0001] The present invention belongs to the field of machine vision, and in particular relates to a related filtering target tracking method based on adaptive weight joint learning. Background technique [0002] Machine vision is currently one of the subjects with the most development potential. It is trying to establish an artificial intelligence system that obtains information from multidimensional data such as voice, image, and video. Target tracking is an important research direction of machine vision. Its main task is to determine the position of the target of interest in the continuous video sequence, and at the same time to obtain the target's motion parameters, so as to carry out deeper processing and analysis. It has a wide range of applications in the fields of automatic monitoring systems, intelligent transportation systems, human-computer interaction, precise military reconnaissance, robot visual navigation, and intelligent medical diagnosis. In recen...

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
IPC IPC(8): G06T7/246
CPCG06T7/248G06T7/251G06T2207/10016G06T2207/20081
Inventor 孔军王本璇蒋敏丁毅涛
Owner JIANGNAN UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

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

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com