结合HOG-LBP特征及多实例在线学习的随机蕨目标跟踪方法
PDF下载 (274)吉培培,陈 恳 *,刘 哲,吴 盈.结合HOG-LBP特征及多实例在线学习的随机蕨目标跟踪方法[J].宁波大学学报(理工版),2015,28(04):42-47.DOI:
JI Pei-pei,CHEN Ken *,LIU Zhe,WU Ying.Random Ferns with HOG-LBP Feature and Online Multi-instance Learning for Complex Tracking[J].Journal of Ningbo University(Natural Science & Engineering Edition),2015,28(04):42-47.DOI:
| Title: | Random Ferns with HOG-LBP Feature and Online Multi-instance Learning for Complex Tracking |
| 作者: | 吉培培, 陈 恳 *, 刘 哲, 吴 盈 |
| Author(s): | JI Pei-pei, CHEN Ken *, LIU Zhe, WU Ying |
| 关键词: | 视觉跟踪; 目标模型; 随机蕨; HOG-LBP特征; 多示例学习 |
| Keywords: | visual tracking; target modeling; random ferns; HOG-LBP feature; multi-instance learning |
| 分类号: | TN919.8 |
| 文献标识码: | A |
| 摘要: | 提出了通过联合随机蕨与HOG-LBP特征建立目标模型的方法. 首先利用选定目标图像块中的HOG-LBP特征向量完成初始化及在其后的过程中产生新的随机蕨, 该随机蕨用于对兴趣目标的检测和跟踪. 目标模型的更新通过多实例在线学习与更新蕨池实现. 提出的方法在选定6个标准视频序列进行测试, 测试结果与现在较流行的OnlineBoostingTracker、MILTracker等跟踪算法进行了比较和分析. 结果表明, 在各种复杂环境下, 本研究提出的方法具备良好的跟踪鲁棒性, 在抗局部遮挡和光照变化方面具有一定的优势; 同时算法具备一定的抗尺度变化能力; 在抗旋转方面, 该算法仍有一定的可改善空间. |
| Abstract: | In this paper, a target modeling approach using random ferns with HOG-LBP feature and online multi-instance learning is presented. The proposed HOG-LBP feature vector with a target image block is utilized to initialize and later form new random ferns which ultimately serve as the target detector and tracker. The target model updating is attained through online multi-instance learning with a fern pool. The proposed approach is put to test on the given number of standard test sequences, and compared with other recently better known tracking algorithms, manifesting its more satisfying performance and objective strength in tracking in complex scenes. |
| 参考文献 /References: | [1] Saffari A, Leistner C, Santner J, et al. Online random forests[C]//Proceeding of the 12th International Conference on Computer Vision: Kyoto Japan, 2009: 1393-1400. [2] Ozuysal M, Calonder M, Lepetit V, et al. Fast keypoint recognition using random ferns[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2010, 32(3): 448-461. [3] Ozuysal M, Fua P, Lepetit V. Fast keypoint recognition in ten lines of code[C]//IEEE Conference on Computer Vision and Pattern Recognition, Minneapolis Minnesota USA, 2007:1-8. [4] Villamizar M, Moreno-Noguer F, Andrade-Cetto J, et al. Efficient rotation invariant object detection using boosted random ferns[C]//IEEE Conference on Computer Vision and Pattern Recognition: San Francisco USA, 2010: 1038-1045. [5] Grabner H, Grabner M, Bischof H. Real-time tracking via on-line boosting[C]//IEEE Conference on Computer Vision and Pattern Recognition: New York USA, 2006: 260-267. [6] Grabner H, Leistner C, Bischof H. Semi-supervised on-line boosting for robust tracking[C]//European Conference on Computer Vision: Marseille France, 2008: 234-247. [7] Stalder S, Grabner H, Gool L, Beyond semi-supervised tracking: Tracking should be as simple as detection, but not simpler than recognition[C]//International Conference on Computer Vision: Kyoto Japan, 2009:1409-1416. [8] Babenko B, Yang M H, Belongie S. Robust object tracking with online multiple instance learning[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2011, 33(8):1619-1632. [9] Babenko B, Yang M H, Belongie S. Visual tracking with online multiple instance learning[C]//IEEE Conference on Computer Vision and Pattern Recognition: Miami, Florida USA, 2009:983-990. [10] Hong Seunghoon, Han Bohyung. Visual tracking by sampling tree-structured graphical models[C]//European Conference on Computer Vision, Florence Zurich Switzerland, 2014:1-16. [11] Antic B, Ommer B. Learning latent constituents recognition of group activities in video[C]//European Conference on Computer Vision, Florence Zurich Switzerland, 2014:33-47. [12] Dalal N, Triggs B. Histograms of oriented gradients for human detection[C]//IEEE Conference on Computer Vision and Pattern Recognition, San Diego USA, 2005: 886-893. [13] Ahmed E, Shakhnarovich G, Maji S. Knowing a good HOG filter when you see it: Efficient selection of filters for detection[C]//European Conference on Computer Vision, Florence, Zurich, Switzerland, 2014:80-94. [14] Zhang Kaihua, Zhang Lei, Yang M H. Real-time Compressive Tracking[C]//European Conference on Computer Vision, Florence Italy, 2012:866-879. |
| 备注/Memo: | Received date: 2014-10-31. JOURNAL OF NINGBO UNIVERSITY ( NSEE ): http://journallg.nbu.edu.cn/ Foundation items: Supported by National Natural Science Foundation (2011ZX03002-004-02); Supported by Natural Science Foundation of Ningbo (2014A610065); Supported by Teaching and Researching Project of Ningbo University (XKXL1308). The first author: JI Pei-pei (1989?), female, Luoyang Henan, M.S. candidate, research area: video tracking. E-mail: helloji_130@163.com *Corresponding author: CHEN Ken (1962?), male, Chongqing, associate professor, research area: image and video processing, multimedia communication, intelligent control. E-mail: chenken@nbu.edu.cn 宁波大学学报（理工版）网址：http://journallg.nbu.edu.cn/ |
