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

On-line target tracking method based on random fern cluster and random projection

A random fern and random projection technology, applied in the field of visual tracking, can solve problems such as consumption, large memory, and information loss, and achieve the effects of improving tracking efficiency, reducing memory resources, and saving computing resources

Inactive Publication Date: 2015-06-03
XIDIAN UNIV
View PDF3 Cites 6 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Second, the comparison of each pixel pair produces two possible outputs 0 or 1, causing other information loss, so more pixel pairs are needed to make up the loss
These two reasons cause this method to consume 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
  • On-line target tracking method based on random fern cluster and random projection
  • On-line target tracking method based on random fern cluster and random projection
  • On-line target tracking method based on random fern cluster and random projection

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0029] The technical solutions of the present invention will be further described in detail below in conjunction with the accompanying drawings.

[0030] refer to figure 1 , the online target tracking method based on random ferns and random projection that the present invention proposes, carries out as follows:

[0031] Step 1: Use real features to represent all pixel pairs of the image block in the image to be detected.

[0032] (1a) Use feature set F={f for the image block whose size is w×h pixels in the image to be detected 1 ,f 2 ,...,f q ,...,f n}, where q=1,2,...,n; divide the feature set F into S subsets F i ={f 1 ,f 2 ,...,f j ,...,f N}, where j=1,2,...,N,N=n / S, each subset corresponds to a fern, the correlation of data is maintained inside the fern, and the ferns are independent of each other; according to the division of ferns Express the image block feature as F={F 1 , F 2 ,...,F i ,...,F S},i=1,2,...,S;

[0033] (1b) Randomly select a pixel pair in t...

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 an on-line target tracking method based on a random fern cluster and random projection, and mainly aims to solve the problems that an on-line tracking method is large in calculation amount and inaccurate in tracking result. The on-line target tracking method comprises the following steps: firstly, detecting the position of a target in an image by using a random fern cluster detector; secondly, tracking the general position of the target by using a mid-value stream tracker; mutually fusing the results of the detector and the tracker to obtain a final target position; finally, updating the detector by using an IIR filter on line, so as to detect and classify a next frame of image to be detected. The on-line target tracking method has the advantages of being high in tracking efficiency, accurate in tracking result and small in memory resource and calculation resource consumption, and can be applied to view tracking of an embedded system.

Description

technical field [0001] The invention belongs to the field of computer vision, in particular to a visual tracking method, which can be used in the visual tracking of embedded systems with high precision and requirements. Background technique [0002] In visual tracking, Avidan led to the increased use of detection tracking, since then the tracking problem is more concerned with the classification problem of distinguishing the foreground from the background. Classifiers used in detection tracking are repeatedly trained offline, while many recent methods train classifiers with upcoming frames, enabling tracking of arbitrary objects prior to tracking and without the need for target training samples. [0003] For the classifier online training method, Grabner et al. modified the Boosting method so that it can adapt to the online method and proposed an online Boosting tracker. The main idea is that in the first frame of the image, use the Boosting framework to use the image patch...

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): G06K9/66
Inventor 刘凯张劲张勉程飞
Owner XIDIAN UNIV