Improved dual‐mode compressive tracking integrating balanced colour and texture features
Discriminative tracking methods can achieve state‐of‐the‐art performance by considering tracking as a classification problem tackled with both object and background information. As a high efficient discriminative tracker, compressive tracking (CT) has attracted much attention recently. However, it m...
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Wiley
2018-12-01
|
| Series: | IET Computer Vision |
| Subjects: | |
| Online Access: | https://doi.org/10.1049/iet-cvi.2018.5198 |
| _version_ | 1797684190026661888 |
|---|---|
| author | Shuifa Sun Shichao Liu Shiwei Kang Chong Xia Zhiping Dan Bangjun Lei Yirong Wu |
| author_facet | Shuifa Sun Shichao Liu Shiwei Kang Chong Xia Zhiping Dan Bangjun Lei Yirong Wu |
| author_sort | Shuifa Sun |
| collection | DOAJ |
| description | Discriminative tracking methods can achieve state‐of‐the‐art performance by considering tracking as a classification problem tackled with both object and background information. As a high efficient discriminative tracker, compressive tracking (CT) has attracted much attention recently. However, it may easily fail when the object suffers from long‐term occlusions, and severe appearance and illumination changes. To address these issues, the authors develop a robust tracking framework based on CT by considering balanced feature representation as well as dual‐mode classifier construction. First, the original measurement matrix of CT works as a dominated texture feature extractor. To obtain a balanced feature representation, they propose to induce a complementary measurement matrix by considering both texture and colour features. Then, they develop two classifiers (dual mode) by using previous and current sample sets, respectively, and subsequently combine them into one ensemble classifier to track the target, which can help to avoid tracking failure suffering from severe appearance changes and long term occlusion. Moreover, they propose a classifier updating schema to prevent the inclusion of unsatisfied positive samples by predicting the occlusions with their ensemble classifier. The extensive experiments demonstrate the superior performance of their tracking framework under various situations. |
| first_indexed | 2024-03-12T00:26:01Z |
| format | Article |
| id | doaj.art-30a524f65c654171a2dbe0e545afb783 |
| institution | Directory Open Access Journal |
| issn | 1751-9632 1751-9640 |
| language | English |
| last_indexed | 2024-03-12T00:26:01Z |
| publishDate | 2018-12-01 |
| publisher | Wiley |
| record_format | Article |
| series | IET Computer Vision |
| spelling | doaj.art-30a524f65c654171a2dbe0e545afb7832023-09-15T10:32:11ZengWileyIET Computer Vision1751-96321751-96402018-12-011281200120610.1049/iet-cvi.2018.5198Improved dual‐mode compressive tracking integrating balanced colour and texture featuresShuifa Sun0Shichao Liu1Shiwei Kang2Chong Xia3Zhiping Dan4Bangjun Lei5Yirong Wu6Hubei Key Laboratory of Intelligent Vision Based Monitoring for Hydroelectric Engineering/College of Computer and Information TechnologyChina Three Gorges UniversityYichang443002People's Republic of ChinaHubei Key Laboratory of Intelligent Vision Based Monitoring for Hydroelectric Engineering/College of Computer and Information TechnologyChina Three Gorges UniversityYichang443002People's Republic of ChinaHubei Key Laboratory of Intelligent Vision Based Monitoring for Hydroelectric Engineering/College of Computer and Information TechnologyChina Three Gorges UniversityYichang443002People's Republic of ChinaHubei Key Laboratory of Intelligent Vision Based Monitoring for Hydroelectric Engineering/College of Computer and Information TechnologyChina Three Gorges UniversityYichang443002People's Republic of ChinaHubei Key Laboratory of Intelligent Vision Based Monitoring for Hydroelectric Engineering/College of Computer and Information TechnologyChina Three Gorges UniversityYichang443002People's Republic of ChinaHubei Key Laboratory of Intelligent Vision Based Monitoring for Hydroelectric Engineering/College of Computer and Information TechnologyChina Three Gorges UniversityYichang443002People's Republic of ChinaDepartment of RadiologyUniversity of Wisconsin‐Madison School of Medicine and Public Health600 Highland AvenueMadisonWI53792USADiscriminative tracking methods can achieve state‐of‐the‐art performance by considering tracking as a classification problem tackled with both object and background information. As a high efficient discriminative tracker, compressive tracking (CT) has attracted much attention recently. However, it may easily fail when the object suffers from long‐term occlusions, and severe appearance and illumination changes. To address these issues, the authors develop a robust tracking framework based on CT by considering balanced feature representation as well as dual‐mode classifier construction. First, the original measurement matrix of CT works as a dominated texture feature extractor. To obtain a balanced feature representation, they propose to induce a complementary measurement matrix by considering both texture and colour features. Then, they develop two classifiers (dual mode) by using previous and current sample sets, respectively, and subsequently combine them into one ensemble classifier to track the target, which can help to avoid tracking failure suffering from severe appearance changes and long term occlusion. Moreover, they propose a classifier updating schema to prevent the inclusion of unsatisfied positive samples by predicting the occlusions with their ensemble classifier. The extensive experiments demonstrate the superior performance of their tracking framework under various situations.https://doi.org/10.1049/iet-cvi.2018.5198improved dual-mode compressive trackingbalanced colour featuretexture featurediscriminative tracking methodsclassification problemobject information |
| spellingShingle | Shuifa Sun Shichao Liu Shiwei Kang Chong Xia Zhiping Dan Bangjun Lei Yirong Wu Improved dual‐mode compressive tracking integrating balanced colour and texture features IET Computer Vision improved dual-mode compressive tracking balanced colour feature texture feature discriminative tracking methods classification problem object information |
| title | Improved dual‐mode compressive tracking integrating balanced colour and texture features |
| title_full | Improved dual‐mode compressive tracking integrating balanced colour and texture features |
| title_fullStr | Improved dual‐mode compressive tracking integrating balanced colour and texture features |
| title_full_unstemmed | Improved dual‐mode compressive tracking integrating balanced colour and texture features |
| title_short | Improved dual‐mode compressive tracking integrating balanced colour and texture features |
| title_sort | improved dual mode compressive tracking integrating balanced colour and texture features |
| topic | improved dual-mode compressive tracking balanced colour feature texture feature discriminative tracking methods classification problem object information |
| url | https://doi.org/10.1049/iet-cvi.2018.5198 |
| work_keys_str_mv | AT shuifasun improveddualmodecompressivetrackingintegratingbalancedcolourandtexturefeatures AT shichaoliu improveddualmodecompressivetrackingintegratingbalancedcolourandtexturefeatures AT shiweikang improveddualmodecompressivetrackingintegratingbalancedcolourandtexturefeatures AT chongxia improveddualmodecompressivetrackingintegratingbalancedcolourandtexturefeatures AT zhipingdan improveddualmodecompressivetrackingintegratingbalancedcolourandtexturefeatures AT bangjunlei improveddualmodecompressivetrackingintegratingbalancedcolourandtexturefeatures AT yirongwu improveddualmodecompressivetrackingintegratingbalancedcolourandtexturefeatures |