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

Correlation filtering tracking method and device based on depth information

A technology of correlation filtering and depth information, applied in the field of visual target tracking, can solve problems such as ineffective use of depth information, affect tracking effect, lack of fusion, etc. The effect of the tracking effect

Active Publication Date: 2018-03-09
HARBIN INST OF TECH SHENZHEN GRADUATE SCHOOL
View PDF2 Cites 24 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

At present, the tracking algorithms based on depth information are divided into two categories. Among them, the two-dimensional tracking method cannot effectively use the depth information, and does not deeply integrate the depth information with the existing tracking algorithms.
However, due to the lack of mature 3D feature extraction technology in the 3D tracking method, the 3D appearance model of the target is not robust, which affects its tracking effect.
However, when the target is occluded, the scale of the target changes, or it is in a complex background, especially when the target is long-term or severely occluded, how to accurately track the target is still a difficult problem for researchers.

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 tracking method and device based on depth information
  • Correlation filtering tracking method and device based on depth information
  • Correlation filtering tracking method and device based on depth information

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0036] It should be understood that the specific embodiments described here are only used to explain the present invention, not to limit the present invention.

[0037] Specifically, please refer to figure 1 , figure 1 It is a schematic flow chart of a preferred embodiment of the correlation filtering tracking method based on depth information proposed by the present invention.

[0038] Such as figure 1 As shown, the first embodiment of the present invention proposes a correlation filtering tracking method based on depth information, including:

[0039] Step S1, based on the image segmentation technology of the depth map, adaptively quantizes the depth information, and obtains the result of the depth image segmentation;

[0040] Specifically, first use the statistical results of the target depth value to initialize the relevant parameters of the K-Means clustering, and estimate the relevant parameters, the relevant parameters include the cluster center and the number of clu...

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 tracking method and device based on depth information. The method comprises the following steps: quantifying the depth information adaptively to obtaina depth image segmentation result based on a depth image segmentation technology; constructing layered structures of three-dimensional space models corresponding to different scenes by means of the depth image segmentation result; and processing target scale change and detecting shielding by means of the layered structures and in combination with a kernel correlation filtering tracking algorithm.According to the invention, foreground and background information is filtered to reduce interference factors of tracking on the one hand, and a sophisticated image character extraction technology is combined; on the other hand, with the using method of simplifying the depth information by means of the layered structures, processing of target scale change and detection of shielding are easier. By combining the kernel correlation filtering tracking algorithm, a tracking method of a two-dimensional appearance model in a space structure is employed, so that shielding and target scale change can becoped with effectively, and the visual tracking effect is improved.

Description

technical field [0001] The present invention relates to the technical field of visual target tracking, in particular to a depth information-based correlation filter tracking method and device. Background technique [0002] Visual target tracking belongs to video analysis. As an important branch of computer vision, its basic task is to predict the position, area and trajectory of the target in the video sequence based on the position information of the given target in the initial frame. Video analysis supports many applications, such as detecting the motion of objects, classifying objects, understanding the behavior of objects, etc. System decisions provide both semantic and non-semantic information. In recent years, with the continuous innovation of various tracking algorithms, the rapid development of deep learning and the improvement of computer processing speed, real-time and even high-speed target tracking algorithms have emerged, which has effectively promoted the deve...

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): G06T7/215G06K9/62G06T7/207G06T7/50
CPCG06T7/207G06T7/215G06T7/50G06T2207/10016G06F18/23213
Inventor 王轩刘新卉漆舒汉蒋琳廖清姚霖李晔关键刘泽超吴宇琳李化乐贾丰玮
Owner HARBIN INST OF TECH SHENZHEN GRADUATE SCHOOL
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