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

Convolutional neural network-based self-adaptive feature selection target tracking method

A convolutional neural network and feature selection technology, applied in image data processing, instruments, character and pattern recognition, etc., can solve the problems of feature dimension disaster, feature redundancy, and low robustness

Active Publication Date: 2018-07-17
ZHEJIANG UNIV OF TECH
View PDF4 Cites 13 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] Aiming at the problems of feature dimension disaster, feature redundancy, and low robustness caused by the existing convolutional neural network features, the present invention proposes a feature based on the distance between convolutional neural network feature maps. Selection method; this method incorporates multiple correlation filter weights to determine the target position at the same time; uses information propagation strategy to adaptively select features

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
  • Convolutional neural network-based self-adaptive feature selection target tracking method
  • Convolutional neural network-based self-adaptive feature selection target tracking method
  • Convolutional neural network-based self-adaptive feature selection target tracking method

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0051] The present invention will be further described below in conjunction with the accompanying drawings.

[0052] refer to Figure 1 ~ Figure 3 , an adaptive feature selection video tracking method for convolutional neural networks, including the following steps:

[0053] 1) Multi-layer CNN feature extraction, the process is as follows:

[0054] The target position p at a given (t) moment video frame and (t-1) moment t-1 First determine the target search area R(p t -1 ), its scale is M*N, which is generally defined as about 2 times the target scale, and then according to the needs of VGG-Net, the image scale of the search area is adjusted to 224*224 by the image interpolation method, and the output of different layers of the network is used as the extraction The multi-layer convolution feature is obtained, and the extracted feature map is multiplied by a cosine window (cosine) to eliminate the discontinuity of the feature map caused by the edge effect of the image;

[...

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 convolutional neural network-oriented self-adaptive feature selection video target tracking method. The method comprises the following steps of: 1) extracting multi-layer CNNfeatures; 2) training related filters; 3) selecting features; and 4) carrying out target tracking: carrying out observation by taking convolution features of conv3-2, conv3-4, conv3-8, conv3-12 and conv3-16 as target features, respectively training five related filters wk after feature selection, and finally deciding a target position (x*, y*) through weighted voting, wherein wk belongs to [1, 5]and a weight parameter wk is online updated according to a learning rate <Rho>. The method is capable of remarkable reducing the feature dimensionality and reducing the operation amount without losing the feature distinguishing ability.

Description

technical field [0001] The invention belongs to the technical field of video target tracking, in particular to an adaptive feature selection target tracking method. Background technique [0002] Target tracking is the process of feature extraction, appearance modeling, motion analysis and target association of the target of interest in the image sequence. It is widely used in intelligent transportation, visual servoing, human-computer interaction and other fields. Although great progress has been made in the field of target tracking in recent years, it still faces great challenges, such as target occlusion, illumination changes, attitude changes, complex background, low resolution and fast movement of targets, etc. will cause target drift or even tracking failure. [0003] With the great success of convolutional neural network in target detection and object recognition, more and more researchers have begun to focus on applying convolutional neural network features to the fie...

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): G06T7/246G06K9/62
CPCG06T7/246G06T2207/10016G06T2207/20081G06T2207/20084G06T2207/20056G06F18/23G06F18/22G06F18/211
Inventor 周小龙李军伟陈胜勇邵展鹏产思贤
Owner ZHEJIANG UNIV OF TECH