實(shí)時(shí)超像素跟蹤算法
doi: 10.11999/JEIT150705 cstr: 32379.14.JEIT150705
基金項(xiàng)目:
國家自然科學(xué)基金(61141009)
Real-time Superpixels Based Tracking Method
Funds:
The National Natural Science Foundation of China (61141009)
-
摘要: 建立有效的目標(biāo)表觀模型是視覺跟蹤算法的關(guān)鍵。該文采用中層次視覺線索(超像素)對目標(biāo)表觀進(jìn)行建模，提出一種實(shí)時(shí)超像素跟蹤(RSPT)算法。算法采用K近鄰(KNN)方法從超像素特征集合中學(xué)習(xí)目標(biāo)的判別式表觀模型；在后續(xù)幀中，根據(jù)學(xué)習(xí)到的表觀模型計(jì)算目標(biāo)-背景置信圖，然后巧妙地采用積分圖方法估計(jì)目標(biāo)狀態(tài)，實(shí)現(xiàn)了高速的全局最優(yōu)估計(jì)；最后設(shè)計(jì)了目標(biāo)表觀模型的在線更新策略，引入遮擋因子對遮擋進(jìn)行判斷。在配置i5處理器的電腦中，所提RSPT算法使用未經(jīng)優(yōu)化的Matlab代碼以19幀/s的速度實(shí)時(shí)運(yùn)行。對若干序列的對比實(shí)驗(yàn)表明，所提算法能夠在多種復(fù)雜環(huán)境下穩(wěn)定跟蹤目標(biāo)，具有良好的魯棒性。Abstract: Target appearance model is crucial for tracking. In this paper, a Real-time SuperPixels based Tracking (RSPT) method is proposed in a tracking-by-detection framework, by investigating mid-level vision cue superpixels. Firstly, a discriminative appearance model is constructed relying superpixels feature and K-Nearest Neighbor (KNN) learning method. Then the tracking problem is posed by computing a confidence map, and detecting the best target station by maximizing an object location likelihood function. The integral image data structure is adopted for fast detection, innovatively. Implemented in MATLAB without code optimization, the proposed tracker runs at 19 frames per second on an i5 laptop. Extensive experimental results on challenging sequences show that the proposed algorithm performs favorably against some state-of-the-art methods in terms of accuracy and robustness.
-
WU Yi, LIM Jongwoo, and YANG M H. Online object tracking: A benchmark[C]. IEEE Conference on Computer Vision and Pattern Recognition, Portland, USA, 2013: 1354-1362. 陳思, 蘇松志, 李紹滋, 等. 基于在線半監(jiān)督boosting的協(xié)同訓(xùn)練目標(biāo)跟蹤算法[J]. 電子與信息學(xué)報(bào), 2014, 36(4): 888-895. doi: 10.3724/SP.J.1146.2013.00826. CHEN Si, SU Songzhi, LI Shaozi, et al. A novel co-training object tracking algorithm based on online semi-supervised boosting[J]. Journal of Electronics Information Technology, 2014, 36(4): 888-895. doi: 10.3724/SP.J.1146.2013.00826. AVIDAN S. Ensemble tracking[C]. IEEE Conference on Computer Vision and Pattern Recognition, San Diego, USA, 2005: 494-501. BABENKO B, BELONGIE S, and YANG M H. Visual tracking with online multiple instance learning[C]. IEEE Conference on Computer Vision and Pattern Recognition, Miami, USA, 2009: 1003-1010. ZHANG Kaihua, ZHANG Lei, and YANG M H. Fast compressive tracking[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2014, 36(10): 2002-2015. FEDERICO P and BIMBO A D. Object tracking by oversampling local features[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2014, 36(12): 2538-2551. ROSS D, LIM J, LIN R, et al. Incremental learning for robust visual tracking[J]. International Journal of Compute Vision, 2008, 77(1): 125-141. COMANICIU D, RAMESH V, and MEER P. Kernel-based object tracking[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2003, 25(2): 564-577. KWON J and LEE K M. Visual tracking decomposition[C]. IEEE Conference on Computer Vision and Pattern Recognition, Columbus, USA, 2014: 1269-1276. MEI X and LING H. Robust visual tracking using l1 minimization[C]. 2009 IEEE 12th International Conference on Computer Vision, Kyoto, Japan, 2009: 1436-1443. 薛模根, 朱虹, 袁廣林. 基于在線判別式字典學(xué)習(xí)的魯棒視覺跟蹤[J]. 電子與信息學(xué)報(bào), 2015, 37(7): 1654-1659. doi: 10.11999/JEIT141325. XUE Mogen, ZHU Hong, and YUAN Guanglin. Robust visual tracking based on online discrimination dictionary learning [J]. Journal of Electronics Information Technology, 2015, 37(7): 1654-1659. doi: 10.11999/JEIT141325. ACHANTA R, SHAJI A, SMITH K, et al. Slic superpixels compared to state-of-the-art superpixel methods [J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2012, 34(11): 2274-2282. YANG Fan, LU Huchuan, and YANG M H. Robust superpixel tracking[J]. IEEE Transactions on Image Processing, 2014, 23(4): 1639-1651. REN X and MALIK J. Tracking as repeated figure/ground segmentation[C]. IEEE Conference on Computer Vision and Pattern Recognition, Minneapolis, USA, 2007: 1-8. DANELLJAN M, KHAN F S, FELSBERG M, et al. Adaptive color attributes for real-time visual tracking[C]. IEEE Conference on Computer Vision and Pattern Recognition, Columbus, USA, 2014: 1-8. VIOLA P and JONES M. Rapid object detection using a boosted cascade of simple features[C]. IEEE Conference on Computer Vision and Pattern Recognition, Hawaii, USA, 2001: 511-518. ADAM A, RIVLIN E, and SHIMSHONI I. Robust fragments-based tracking using the integral histogram[C]. IEEE Conference on Computer Vision and Pattern Recognition, New York, USA, 2006: 798-805. GRABNER H, GRABNER M, and BISCHOF H. Real-time tracking via on-line boosting[C]. British Machine Vision Conference, Edinburgh, England, 2006: 47-56. -
計(jì)量
- 文章訪問數(shù): 1512
- HTML全文瀏覽量: 158
- PDF下載量: 750
- 被引次數(shù): 0
下載:
