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

Human body target tracking method based on deep learning and correlation filtering

A technology of correlation filtering and human target, applied in character and pattern recognition, instrument, biological neural network model, etc., can solve the problem of low accuracy rate, achieve the effect of improving efficiency, enhancing accuracy rate, and ensuring real-time performance

Inactive Publication Date: 2018-10-30
SHANGHAI JIAO TONG UNIV
View PDF11 Cites 5 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] One of the purposes of the present invention is to solve the problem of low accuracy of existing tracking technology in long-term target tracking
[0006] The second purpose of the present invention is to solve the problem that the existing tracking technology is difficult to achieve real-time calculation while ensuring accuracy

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
  • Human body target tracking method based on deep learning and correlation filtering
  • Human body target tracking method based on deep learning and correlation filtering
  • Human body target tracking method based on deep learning and correlation filtering

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0027] The present invention will be described in detail below with reference to the drawings and specific embodiments. This embodiment is implemented on the premise of the technical solution of the present invention, and provides detailed implementation and specific operation procedures, but the protection scope of the present invention is not limited to the following embodiments.

[0028] Such as figure 1 As shown, this embodiment implements a human target tracking method based on deep learning and related filtering. The method is composed of a human detector, a lightweight human discriminator, and a related filter tracker to complete the human target tracking task. It includes the following steps:

[0029] 1) Read the current frame of the video to be tracked, and use the human body detector to detect the position of the human body;

[0030] 2) Read the next frame of the video to be tracked, and determine whether the end of the video has been reached, if it is, then it ends, if no...

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 relates to a human body target tracking method based on deep learning and correlation filtering. The method comprises the steps that (1) a current frame of a to-be-tracked video is read,and a human body position is detected through a human body detector; (2) the next frame of the to-be-tracked video is read, whether the end of the video is reached is judged, if yes, the process is ended, or otherwise the step (3) is executed; (3) the human body position obtained through detection in the previous step is utilized to initialize a correlation filtering tracker to perform human bodytracking, and tracking lasts for set time; and (4) a lightweight human body discriminator is utilized to determine whether a current tracking target is a human body, if yes, the human body position is recorded, and the step (2) is returned to, or otherwise the step (1) is returned to. Compared with the prior art, the method has the advantages of being high in accuracy, good in instantaneity and the like.

Description

Technical field [0001] The present invention relates to a target tracking technology, in particular to a human target tracking method based on deep learning and correlation filtering. Background technique [0002] Human target tracking in video is a research topic involving pattern recognition, computer vision, artificial intelligence and other fields. Because of its wide application value in intelligent video surveillance, security, education and other fields, it has always been a hot research topic. . However, in real scenes, due to factors such as human posture changes, video jitter, and occlusion, it is difficult for tracking algorithms to balance real-time calculations and accuracy of results at the same time. This problem is particularly obvious in long-term tracking tasks. Therefore, how to achieve an accurate and real-time human target tracking method is still a difficult point in research. [0003] The document "High-speed tracking with kernelized correlation filters" (H...

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): G06K9/00G06K9/62G06N3/04
CPCG06V40/103G06N3/045G06F18/214G06F18/24
Inventor 张君鹏申瑞民姜飞
Owner SHANGHAI JIAO TONG 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