Anti-shielding and target recapturing method of correlation filtering target tracking algorithm

A target tracking and correlation filtering technology, applied in the field of target tracking, can solve problems such as poor matching effect, high algorithm complexity, weak detection ability, etc., achieve real-time frame rate, low algorithm complexity, and avoid interference

Pending Publication Date: 2021-09-24
CHINA KEY SYST & INTEGRATED CIRCUIT
View PDF0 Cites 2 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] The above method can solve some occlusion problems, but it needs to manually set the threshold or learn the weight, and the algorithm is complex, and there are still problems such as weak detection ability and poor matching effect

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
  • Anti-shielding and target recapturing method of correlation filtering target tracking algorithm
  • Anti-shielding and target recapturing method of correlation filtering target tracking algorithm
  • Anti-shielding and target recapturing method of correlation filtering target tracking algorithm

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0065] The present invention adopts the method of speed prediction to give the target search area, invokes the correlation filter target tracking algorithm to give the temporary target position, and uses the characteristic distance of the gradient direction histogram to describe the similarity between the target and the historical target template. Use the gradient direction histogram feature distance of the target and the 8 environmental regions around the target to describe the similarity between the target and the environment, calculate the occlusion judgment threshold, and then compare the distance between the target and the template and the relative size of the occlusion judgment threshold, and classify the occlusion status of the target. There are three states: unoccluded, semi-occluded and fully occluded: (1) For the unoccluded state, update the filter template, target speed, position, occlusion judgment threshold and historical target template pool; (2) For the semi-occlu...

Embodiment 2

[0109] The following is the tracking effect display on the OTB100 with occluded target tracking test data.

[0110] The relevant parameters of anti-occlusion and target recapture methods are set as follows: occlusion judgment constant K 1 and K 2 They are respectively set to 0.4 and 0.9, the length of the historical target template pool is set to 5, the number of stored historical velocities is set to 5, and the learning rate υ of the occlusion judgment threshold is set to 0.015. The correlation filter target tracking algorithm chooses the classic KCF algorithm, and the filter template reference size is set to 128 pixels.

[0111] 1. Select the occlusion tracking test data in OTB100 as the test set, input the first frame of the tracking data and the initial position information of the target into the KCF target tracking algorithm for initialization, and the KCF algorithm calculates the initial filter template. The gradient direction histogram feature of the target area is ex...

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 an anti-shielding and target recapturing method of a correlation filtering target tracking algorithm, and belongs to the field of target tracking. A speed prediction mode is adopted, a target search area is given, a correlation filtering target tracking algorithm is called to give a temporary target position, and a gradient direction histogram feature distance is adopted to describe the similarity between a target and a historical target template. The method comprises describing the similarity degree of a target and an environment by using the feature distance of a gradient direction histogram of the target and eight environment areas around the target, calculating a shielding judgment threshold value, then comparing the distance between the target and a template with the relative size of the shielding judgment threshold value, and dividing the shielding state of the target into three states of non-shielding, semi-shielding and full shielding; for a non-shielding state, updating a filter template, a target speed, a target position, a shielding judgment threshold value and a historical target template pool; updating only the speed and the position of the target in the semi-shielding state; for a full-shielding state, updating no parameter, and enabling a target position to adopt a speed prediction position.

Description

technical field [0001] The invention relates to the technical field of target tracking, in particular to an anti-occlusion and target recapture method of a correlation filtering target tracking algorithm. Background technique [0002] Object tracking has a wide range of applications in surveillance, motion analysis, medical imaging, behavior recognition, monitoring, and human-computer interaction. As one of the current mainstream target tracking algorithms, correlation filter target tracking technology has the advantages of simplicity, high precision and fast speed, and has attracted the attention and research of many experts and scholars. Such as Kernel Correlation Filter (KCF) tracker using gradient orientation histogram feature, and visual tracker using adaptive color feature (CN) based on KCF. In common situations, the correlation filter target tracking algorithm does have excellent tracking effect, but when the target is occluded, the algorithm itself cannot discrimina...

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): G06T7/246G06T7/269G06T7/66G06T5/40G06T5/00
CPCG06T7/66G06T5/40G06T7/246G06T7/269G06T5/70Y02T10/40
Inventor 杨赟辉朱鸿泰程虎周弦杨希
Owner CHINA KEY SYST & INTEGRATED CIRCUIT
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