The Kernel Based Multiple Instances Learning Algorithm for Object Tracking
To realize real time object tracking in complex environments, a kernel based MIL (KMIL) algorithm is proposed. The KMIL employs the Gaussian kernel function to deal with the inner product used in the weighted MIL (WMIL) algorithm. The method avoids computing the pos-likely-hood and neg-likely-hood m...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2018-06-01
|
| Series: | Electronics |
| Subjects: | |
| Online Access: | http://www.mdpi.com/2079-9292/7/6/97 |
| _version_ | 1798034180901175296 |
|---|---|
| author | Tiwen Han Lijia Wang Binbin Wen |
| author_facet | Tiwen Han Lijia Wang Binbin Wen |
| author_sort | Tiwen Han |
| collection | DOAJ |
| description | To realize real time object tracking in complex environments, a kernel based MIL (KMIL) algorithm is proposed. The KMIL employs the Gaussian kernel function to deal with the inner product used in the weighted MIL (WMIL) algorithm. The method avoids computing the pos-likely-hood and neg-likely-hood many times, which results in a much faster tracker. To track an object with different motion, the searching areas for cropping the instances are varied according to the object’s size. Furthermore, an adaptive classifier updating strategy is presented to handle with the occlusion, pose variations and illumination changes. A similar score range is defined with respect to two given thresholds and a similar score from the second frame. Then, the learning rate will be set to be a small value when a similar score is out of the range. In contrast, a big learning rate is used. Finally, we compare its performance with that of the state-of-art algorithms on several classical videos. The experimental results show that the presented KMIL algorithm is faster and robust to the partial occlusion, pose variations and illumination changes. |
| first_indexed | 2024-04-11T20:40:37Z |
| format | Article |
| id | doaj.art-53b16c1b16d84e14bfe78ec45540af40 |
| institution | Directory Open Access Journal |
| issn | 2079-9292 |
| language | English |
| last_indexed | 2024-04-11T20:40:37Z |
| publishDate | 2018-06-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Electronics |
| spelling | doaj.art-53b16c1b16d84e14bfe78ec45540af402022-12-22T04:04:13ZengMDPI AGElectronics2079-92922018-06-01769710.3390/electronics7060097electronics7060097The Kernel Based Multiple Instances Learning Algorithm for Object TrackingTiwen Han0Lijia Wang1Binbin Wen2Department of Intelligent Manufacture, Hebei College of Industry and Technology, Shijiazhuang 050091, ChinaDepartment of Intelligent Manufacture, Hebei College of Industry and Technology, Shijiazhuang 050091, ChinaDepartment of Intelligent Manufacture, Hebei College of Industry and Technology, Shijiazhuang 050091, ChinaTo realize real time object tracking in complex environments, a kernel based MIL (KMIL) algorithm is proposed. The KMIL employs the Gaussian kernel function to deal with the inner product used in the weighted MIL (WMIL) algorithm. The method avoids computing the pos-likely-hood and neg-likely-hood many times, which results in a much faster tracker. To track an object with different motion, the searching areas for cropping the instances are varied according to the object’s size. Furthermore, an adaptive classifier updating strategy is presented to handle with the occlusion, pose variations and illumination changes. A similar score range is defined with respect to two given thresholds and a similar score from the second frame. Then, the learning rate will be set to be a small value when a similar score is out of the range. In contrast, a big learning rate is used. Finally, we compare its performance with that of the state-of-art algorithms on several classical videos. The experimental results show that the presented KMIL algorithm is faster and robust to the partial occlusion, pose variations and illumination changes.http://www.mdpi.com/2079-9292/7/6/97object trackingkernel based MIL algorithmGaussian kerneladaptive classifier updating |
| spellingShingle | Tiwen Han Lijia Wang Binbin Wen The Kernel Based Multiple Instances Learning Algorithm for Object Tracking Electronics object tracking kernel based MIL algorithm Gaussian kernel adaptive classifier updating |
| title | The Kernel Based Multiple Instances Learning Algorithm for Object Tracking |
| title_full | The Kernel Based Multiple Instances Learning Algorithm for Object Tracking |
| title_fullStr | The Kernel Based Multiple Instances Learning Algorithm for Object Tracking |
| title_full_unstemmed | The Kernel Based Multiple Instances Learning Algorithm for Object Tracking |
| title_short | The Kernel Based Multiple Instances Learning Algorithm for Object Tracking |
| title_sort | kernel based multiple instances learning algorithm for object tracking |
| topic | object tracking kernel based MIL algorithm Gaussian kernel adaptive classifier updating |
| url | http://www.mdpi.com/2079-9292/7/6/97 |
| work_keys_str_mv | AT tiwenhan thekernelbasedmultipleinstanceslearningalgorithmforobjecttracking AT lijiawang thekernelbasedmultipleinstanceslearningalgorithmforobjecttracking AT binbinwen thekernelbasedmultipleinstanceslearningalgorithmforobjecttracking AT tiwenhan kernelbasedmultipleinstanceslearningalgorithmforobjecttracking AT lijiawang kernelbasedmultipleinstanceslearningalgorithmforobjecttracking AT binbinwen kernelbasedmultipleinstanceslearningalgorithmforobjecttracking |