Pedestrain Detection Method Based on Partition Ensemble
LUO Hui-Lan1, PENG Kai1, KONG Fan-Sheng2
1.School of Information Engineering, Jiangxi University of Science and Technology, Ganzhou 341000 2.College of Computer Science and Technology, Zhejiang University, Hangzhou 310027
Abstract To improve the accuracy of pedestrain detection, an ensemble approach for pedestrian detection in still images is proposed. Firstly, a partition ensemble method is used to evenly split the entire training window to get small regions, and features of small regions are extracted. Then, the AdaBoost classifiers are trained on different regions to get part classifiers. A global classifier is formed by weighted summing of these part classifiers. More global classifiers are obtained by using different partitioning methods to repeat the process. To improve detection results and achieve better performance, two global classifiers are built by using histograms of oriented gradient, and multi-level version of HOG descriptor features respectively for each partitioning method. The classifier ensemble is used to detect new images and the weighted voting method is used to decide the final results. Experimental results show that the proposed method achieves better performance than the whole window detector on INRIA dataset.
[1] Schwartz W R, Kembhavi A, Harwood D, et al. Human Detection Using Partial Least Squares Analysis // Proc of the 12th IEEE International Conference on Computer Vision. Kyoto, Japan, 2009: 24-31 [2] Wang X Y, Han T X, Yan S C. An HOG-LBP Human Detector with Partial Occlusion Handling // Proc of the 12th IEEE International Conference on Computer Vision. Kyoto, Japan, 2009: 32-39 [3] Lin Z, Davis L S. A Pose-Invariant Descriptor for Human Detection and Segmentation // Proc of the 10th European Conference on Computer Vision. Marseille, France, 2008: 423-436
[4] Walk S, Majer N, Schindler K, et al. New Features and Insights for Pedestrian Detection // Proc of the IEEE Conference on Computer Vision and Pattern Recognition. San Francisco, USA, 2010: 1030-1037 [5] Lampert C H, Blaschko M B, Hofmann T. Beyond Sliding Windows: Object Localization by Efficient Subwindow Search // Proc of the IEEE Conference on Computer Vision and Pattern Recognition. Anchorage, USA, 2008. DOI: 10.1109/CVPR.2008.4587586 [6] Dalal N, Triggs B. Histograms of Oriented Gradients for Human Detection // Proc of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition. San Diego, USA, 2005, I: 886-893 [7] Enzweiler M, Eigenstetter A, Schiele B, et al. Multi-cue Pedestrian Classification with Partial Occlusion Handling // Proc of the IEEE Conference on Computer Vision and Pattern Recognition. San Francisco, USA, 2010: 990-997 [8] Girshick R B, Felzenszwalb P F, McAllester D. Object Detection with Grammar Models[EB/OL]. [2014-12-20]. http://www.cs.berkeley.edu/~rbg/papers/grammar-nips11.pdf [9] Benenson R, Mathias M, Timofte R, et al. Pedestrian Detection at 100 Frames per Second // Proc of the IEEE Conference on Compu-ter Vision and Pattern Recognition. Providence, USA, 2012: 2903-2910 [10] Lowe D G. Distinctive Image Features from Scale-Invariant Keypoint. International Journal of Computer Vision, 2004, 60(2): 91-110 [11] Sabzmeydani P, Mori G. Detecting Pedestrians by Learning Shapelet Features // Proc of the IEEE Conference on Computer Vision and Pattern Recognition. Minneapolis, USA, 2007. DOI: 10.1109/CVPR.2007.383134 [12] Mu Y D, Yan S C, Liu Y, et al. Discriminative Local Binary Pa-tterns for Human Detection in Personal Album // Proc of the IEEE Conference on Computer Vision and Pattern Recognition. Ancho-rage, USA, 2008. DOI: 10.1109/CVPR.2008.4587800 [13] Ren X F, Ramanan D. Histograms of Sparse Codes for Object Detection // Proc of the 26th IEEE Conference on Computer Vision and Pattern Recognition. Portland, USA, 2013: 3246-3253 [14] Ouyang Y, Zhang S Y, Zhang Y. Multi-scale Human Detection Based on Window Gradient Potential Energy with Partial Occlusion Handling. Journal of Electronics & Information Technology, 2012, 34(4): 858-864 (in Chinese) (欧阳毅,张三元,张 引.基于窗口边缘梯度势能的人体遮挡多尺度检测算法.电子与信息学报, 2012, 34(4): 858-864) [15] Zeng B B, Wang G J, Lin X G. Color Self-similarity Feature Based Real-Time Pedestrian Detection. Journal of Tsinghua University: Science and Technology, 2012, 52(4): 571-574 (in Chinese) (曾波波,王贵锦,林行刚.基于颜色自相似度特征的实时行人检测.清华大学学报:自然科学版, 2012, 52(4): 571-574) [16] Lin Z, Hua G, Davis L S. Multiple Instance Feature for Robust Part-Based Object Detection // Proc of the IEEE Conference on Computer Vision and Pattern Recognition. Miami, USA, 2009: 405-412 [17] Ding J H, Geng W D, Wang Y G. Human Detection Method Based on Multi-part Detector and Multi-instance Learning. Pattern Recognition and Artificial Intelligence, 2012, 25(5): 803-809 (in Chinese) (丁建浩,耿卫东,王毅刚.基于多部位多示例学习的人体检测.模式识别与人工智能, 2012, 25(5): 803-809) [18] Felzenszwalb P F, Girshick R B, McAllester D, et al. Object Detection with Discriminatively Trained Part-Based Models. IEEE Trans on Pattern Analysis and Machine Intelligence, 2010, 32(9): 1627-1645 [19] Ye Q X, Jiao J B, Jiang S Q. Fast and Robust Pedestrian Detection Algorithm with Multi-scale Orientation Features. Journal of Software, 2011, 22(12): 3004-3014 (in Chinese) (叶齐祥,焦建彬,蒋树强.基于多尺度方向特征的快速鲁棒人体检测算法.软件学报, 2011, 22(12): 3004-3014) [20] Wojek C, Walk S, Schiele B. Multi-cue Onboard Pedestrian Detection // Proc of the IEEE Conference on Computer Vision and Pattern Recognition. Miami, USA, 2009: 794-801 [21] Maji S, Berg A C, Malik J. Classification Using Intersection Kernel Support Vector Machines is Efficient // Proc of the IEEE Conference on Computer Vision and Pattern Recognition. Anchorage, USA, 2008. DOI: 10.1109/CVPR.2008.4587630 [22] Freund Y, Schapire R E. Decision-Theoretic Generalization of On-line Learning and an Application to Boosting. Journal of Computer and System Sciences, 1997, 55(1): 119-139 [23] Friedman J, Hastie T, Tibshirani R. Additive Logistic Regression: A Statistical View of Boosting. The Annals of Statistics, 2000, 28(2): 337-407
2015, Vol. 28 
