Reinspection method used during tracking failure of monocular long-time visual tracking method

A technology for visual tracking and tracking failure, applied in the field of tracking and rechecking, it can solve the problems of high feature dependence, slow training speed, and large memory footprint, and achieve the effect of low memory footprint, fast training speed, and improved tracking ability.

Inactive Publication Date: 2017-12-19
HANGZHOU DIANZI UNIV
View PDF2 Cites 10 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

This type of model has a fast detection speed, but it is slow in training speed and highly dependent on features. The most important thing is that it often needs to use more decision trees for the effect, which makes it take up a lot of memory.

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
  • Reinspection method used during tracking failure of monocular long-time visual tracking method
  • Reinspection method used during tracking failure of monocular long-time visual tracking method
  • Reinspection method used during tracking failure of monocular long-time visual tracking method

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0033] Such as figure 1 As shown, a method of re-examination used when the tracking of a monocular long-term visual tracking method fails, the method specifically includes the following steps:

[0034] Step 1. When a new t frame comes, the main tracker processes the image through the KCF algorithm and the DSST algorithm scale update strategy, and generates an output g(x,y) of the current detection area, where g(x,y ) is a two-dimensional variable, and the point with the largest value is the target position;

[0035] Step 2. Use the peak sidelobe ratio PSR re-detection method to judge g(x, y), where g max , μ s1 , σ s1 are the peak value, sidelobe mean and sidelobe standard deviation of the correlation response g(x,y), respectively;

[0036] To get the current trace result, use the following formula:

[0037]

[0038] Among them, N is a fixed constant; f psr (x frame ) is the distribution function of the peak sidelobe ratio PSR; where i is an arbitrary constant, x fr...

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 reinspection method used during the tracking failure of a monocular long-time visual tracking method. A weighting matrix is used for updating an STC (Spatio-Temporal Context) model, and the characteristics of the STC model for saptio-temporal context modeling are taken as reinspection equipment. When a kernel-related filter response can not meet inspection requirements, the STC model is used for carrying out reinspection to finish position inspection under a covering situation. A method that an HSV (Hue, Saturation and Value) space assists in determining a position is used for improving a boundary effect brought to a WSTC (Weighted Spatio-Temporal Context) algorithm by a Hamming window under a situation that operation speed is guaranteed, and the tracking ability of the WSTC algorithm for a quick movement object is improved.

Description

technical field [0001] The invention relates to a tracking and rechecking method, in particular to a rechecking method used when tracking fails in a monocular long-term visual tracking method. Background technique [0002] Visual tracking is an important computer vision task. Due to partial occlusion, deformation, motion blur, fast motion, illumination changes, background clutter, scale changes, etc., it still faces many challenges in practical applications, so it is still the most active in computer vision. one of the research fields. Among them, the discriminative classifier is a core component in modern tracking methods, and a binary classifier is learned online to distinguish the target from the background in each frame. Correlation filters have been successfully applied to object detection and recognition. Recently, as a discriminative tracking method, correlation filters have been applied to the field of visual tracking and achieved good results. [0003] Most of the...

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): G06T7/231G06T7/269
CPCG06T7/231G06T7/269G06T2207/10016G06T2207/20081
Inventor 徐云杰余世平吴涛
Owner HANGZHOU DIANZI UNIV
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