Unlock instant, AI-driven research and patent intelligence for your innovation.

KCF target tracking method based on cubic spline interpolation and Markov chain

A Markov chain, target tracking technology, applied in the field of occluded target detection and positioning tracking, which can solve the problem of completely occluded targets being powerless and so on.

Pending Publication Date: 2021-06-18
北京京航计算通讯研究所
View PDF11 Cites 1 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Although it has achieved certain effects, it is powerless for completely occluded targets.

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
  • KCF target tracking method based on cubic spline interpolation and Markov chain
  • KCF target tracking method based on cubic spline interpolation and Markov chain
  • KCF target tracking method based on cubic spline interpolation and Markov chain

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0218] In this embodiment, some video sequences in the LaSot dataset (see https: / / arxiv.org / abs / 1809.07845) are selected, and the first 500 frames of each video sequence are selected for tracking performance testing.

[0219] In the present invention, cubic spline interpolation and Markov chain are used to solve the problem of target prediction after occlusion, and at the same time, the search range is expanded so that the target can be found when the target reappears. For the convenience of subsequent description, this algorithm is recorded as MS-KCF algorithm.

[0220] In order to illustrate the usefulness of the target tracking algorithm of the present invention, the traditional KCF algorithm is used as the first comparison algorithm, and the classic KCF algorithm based on Kalman filter is used as the second comparison algorithm, which is denoted as K-KCF algorithm. In the K-KCF algorithm, in the process of target tracking, two-dimensional Kalman filter is added to correct ...

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 provides a KCF target tracking method based on cubic spline interpolation and a Markov chain for solving the shielding problem in the target tracking process. The specific thought of the method is characterized by the following steps: when a shielding problem occurs in the target tracking process, updating of a KCF filtering template being stopped, and according to the motion characteristics of a target before shielding, a plurality of cubic spline polynomials of different weights trained offline being used for simulating the motion trail of the target after shielding; and based on the hypothesis that the current state in the Markov chain is only related to the state at the previous moment, considering that the weight of the current image in the tracking process is only related to the weight of the previous frame image, thereby realizing prediction tracking of the motion position of the shielded target. The method is not limited to features in the image, focuses on the motion trail of the target, solves the problem of shielding in the target tracking process through analysis and simulation of the motion trail of the object, and has the characteristics of high tracking accuracy and high calculation speed.

Description

technical field [0001] The invention relates to the field of occlusion target detection and location tracking, in particular to a KCF target tracking method based on cubic spline interpolation and Markov chain. Background technique [0002] Target tracking refers to identifying the moving target in the video, and finding the target in the next frame of image, and so on to complete the tracking. Target tracking algorithms can be divided into generative model methods and discriminative model methods according to different methods and ideas. The generative method uses a feature model to describe the appearance characteristics of the target, and then minimizes the reconstruction error between the tracking target and the candidate target to confirm the target. The generative method focuses on the feature extraction of the target itself, ignoring the background information of the target. Target drift or target loss is prone to occur when drastic changes or occlusions occur; discr...

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): G06T7/246G06T3/40G06K9/46G06N3/04G06N3/08
CPCG06T7/246G06T3/4007G06N3/04G06N3/084G06T2207/30241G06T2207/10016G06T2207/20081G06T2207/10024G06V10/44
Inventor 王磊丁祥涛梁锦威张依漪李娜王佳佳宋明艳滕岩梅铁林
Owner 北京京航计算通讯研究所