Position correction long-range tracking method based on continuous correlation filter

A correlation filter, long-range technology, applied in the field of visual tracking algorithms, which can solve problems such as tracking offset, tracking failure, and tracker offset error accumulation.

Active Publication Date: 2021-10-26
HARBIN INST OF TECH AT WEIHAI
View PDF5 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Although target tracking is widely used, long-range tracking in complex scenes is still a challenging issue, because the tracker often causes tracking deviation due to various interference factors, including target occlusion, fast moving and scale changes etc.
Moreover, the offset error of the tracker is easy to accumulate, and it is easy to cause tracking failure in the case of long-distance tracking

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
  • Position correction long-range tracking method based on continuous correlation filter
  • Position correction long-range tracking method based on continuous correlation filter
  • Position correction long-range tracking method based on continuous correlation filter

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0054] The present invention will be further described below in conjunction with the accompanying drawings.

[0055] The goal of the present invention is to first build a correlation filter in the continuous domain, thereby obtaining higher tracker performance. Suppose the training sample x of the tracker contains D feature channels x 1 ,...,x D , each feature channel x d can generate a corresponding correlation filter weight w d . Then the traditional correlation filter can be obtained by solving the following loss function:

[0056]

[0057] where λ is the regularization parameter (λ≥0), and y is the desired output of the correlation filter, which is the shape of a two-dimensional Gaussian function. ||f(k)|| 2 represents the binorm of the sequence f(k), and

[0058] In the proposed algorithm, all parameters involved in the loss function are defined in the continuous domain. But in fact, the feature channel x of the training sample d comes from the extracted ima...

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 present invention relates to the technical field of visual tracking algorithms, in particular to a continuous correlation filter-based positioning correction long-range tracking algorithm with better tracking performance and tracking stability, including: realizing target positioning in the continuous domain, selecting target features , by separately training the three templates R t , R s and R d implementation, where the target localization template R t It is to encode changes in the appearance of the target in real time, and at the same time need to maintain robustness to some tracking disturbances such as fast movement and lighting changes. Choose VGG‑19 convolutional neural network and use conv3‑4, conv4‑4 and conv5‑4 The output features of the layer are used to train the target localization template R t , while for R s and R d The template, considering the problem of computational efficiency, uses HOG features; realizes target relocation.

Description

Technical field: [0001] The invention relates to the technical field of visual tracking algorithms, in particular to a continuous correlation filter-based location correction long-distance tracking method with better tracking performance and tracking stability. Background technique: [0002] Image or video-based target tracking technology is an important research topic in the field of computer vision, and is widely used in video surveillance, intelligent navigation / assisted driving, intelligent robots, motion behavior analysis and other fields. The basic framework of target tracking is to initialize a rectangular box containing the target of interest in the first frame of the video sequence, and the tracker calibrates the position, area and shape of the target in the next video frame. Although target tracking is widely used, long-range tracking in complex scenes is still a challenging issue, because the tracker often causes tracking deviation due to various interference fact...

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): G06T7/246H04N7/18
CPCG06T2207/20024G06T7/246H04N7/18
Inventor 康文静李欣尤刘功亮
Owner HARBIN INST OF TECH AT WEIHAI
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