Visual Object Tracking in RGB-D Data via Genetic Feature Learning

Visual object tracking is a fundamental component in many computer vision applications. Extracting robust features of object is one of the most important steps in tracking. As trackers, only formulated on RGB data, are usually affected by occlusions, appearance, or illumination variations, we propos...

Full description

Saved in:
Bibliographic Details
Main Authors: Ming-xin Jiang, Xian-xian Luo, Tao Hai, Hai-yan Wang, Song Yang, Ahmed N. Abdalla
Format: Article
Language:English
Published: Wiley 2019-01-01
Series:Complexity
Online Access:http://dx.doi.org/10.1155/2019/4539410
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849397451788648448
author Ming-xin Jiang
Xian-xian Luo
Tao Hai
Hai-yan Wang
Song Yang
Ahmed N. Abdalla
author_facet Ming-xin Jiang
Xian-xian Luo
Tao Hai
Hai-yan Wang
Song Yang
Ahmed N. Abdalla
author_sort Ming-xin Jiang
collection DOAJ
description Visual object tracking is a fundamental component in many computer vision applications. Extracting robust features of object is one of the most important steps in tracking. As trackers, only formulated on RGB data, are usually affected by occlusions, appearance, or illumination variations, we propose a novel RGB-D tracking method based on genetic feature learning in this paper. Our approach addresses feature learning as an optimization problem. As owning the advantage of parallel computing, genetic algorithm (GA) has fast speed of convergence and excellent global optimization performance. At the same time, unlike handcrafted feature and deep learning methods, GA can be employed to solve the problem of feature representation without prior knowledge, and it has no use for a large number of parameters to be learned. The candidate solution in RGB or depth modality is represented as an encoding of an image in GA, and genetic feature is learned through population initialization, fitness evaluation, selection, crossover, and mutation. The proposed RGB-D tracker is evaluated on popular benchmark dataset, and experimental results indicate that our method achieves higher accuracy and faster tracking speed.
format Article
id doaj-art-aebfedba454f46e6beb93d47a3b6f493
institution Kabale University
issn 1076-2787
1099-0526
language English
publishDate 2019-01-01
publisher Wiley
record_format Article
series Complexity
spelling doaj-art-aebfedba454f46e6beb93d47a3b6f4932025-08-20T03:38:59ZengWileyComplexity1076-27871099-05262019-01-01201910.1155/2019/45394104539410Visual Object Tracking in RGB-D Data via Genetic Feature LearningMing-xin Jiang0Xian-xian Luo1Tao Hai2Hai-yan Wang3Song Yang4Ahmed N. Abdalla5Jiangsu Laboratory of Lake Environment Remote Sensing Technologies, Huaiyin Institute of Technology, Huaian 223003, ChinaFaculty of Mathematics and Computer Science, Quanzhou Normal University, Quanzhou 362000, ChinaComputer Science Department, Baoji University of Arts and Sciences, Shanxi 721031, ChinaJiangsu Laboratory of Lake Environment Remote Sensing Technologies, Huaiyin Institute of Technology, Huaian 223003, ChinaJiangsu Laboratory of Lake Environment Remote Sensing Technologies, Huaiyin Institute of Technology, Huaian 223003, ChinaJiangsu Laboratory of Lake Environment Remote Sensing Technologies, Huaiyin Institute of Technology, Huaian 223003, ChinaVisual object tracking is a fundamental component in many computer vision applications. Extracting robust features of object is one of the most important steps in tracking. As trackers, only formulated on RGB data, are usually affected by occlusions, appearance, or illumination variations, we propose a novel RGB-D tracking method based on genetic feature learning in this paper. Our approach addresses feature learning as an optimization problem. As owning the advantage of parallel computing, genetic algorithm (GA) has fast speed of convergence and excellent global optimization performance. At the same time, unlike handcrafted feature and deep learning methods, GA can be employed to solve the problem of feature representation without prior knowledge, and it has no use for a large number of parameters to be learned. The candidate solution in RGB or depth modality is represented as an encoding of an image in GA, and genetic feature is learned through population initialization, fitness evaluation, selection, crossover, and mutation. The proposed RGB-D tracker is evaluated on popular benchmark dataset, and experimental results indicate that our method achieves higher accuracy and faster tracking speed.http://dx.doi.org/10.1155/2019/4539410
spellingShingle Ming-xin Jiang
Xian-xian Luo
Tao Hai
Hai-yan Wang
Song Yang
Ahmed N. Abdalla
Visual Object Tracking in RGB-D Data via Genetic Feature Learning
Complexity
title Visual Object Tracking in RGB-D Data via Genetic Feature Learning
title_full Visual Object Tracking in RGB-D Data via Genetic Feature Learning
title_fullStr Visual Object Tracking in RGB-D Data via Genetic Feature Learning
title_full_unstemmed Visual Object Tracking in RGB-D Data via Genetic Feature Learning
title_short Visual Object Tracking in RGB-D Data via Genetic Feature Learning
title_sort visual object tracking in rgb d data via genetic feature learning
url http://dx.doi.org/10.1155/2019/4539410
work_keys_str_mv AT mingxinjiang visualobjecttrackinginrgbddataviageneticfeaturelearning
AT xianxianluo visualobjecttrackinginrgbddataviageneticfeaturelearning
AT taohai visualobjecttrackinginrgbddataviageneticfeaturelearning
AT haiyanwang visualobjecttrackinginrgbddataviageneticfeaturelearning
AT songyang visualobjecttrackinginrgbddataviageneticfeaturelearning
AT ahmednabdalla visualobjecttrackinginrgbddataviageneticfeaturelearning