Abstract Due to the large sample and multiple characteristics of video sequence in the field of human action recognition, a method of kernel Fisher nonlinear discriminant (KFLD) - scale invariant feature transform (SIFT) and relevance vector machine (RVM) fuzzy integral fusion recognition based on tensor is proposed. Firstly, video sequence is pre-processed into binary video sequence, and then it is described as third-order tensor. Furthermore, as for large sample characteristics, a local feature extraction method of KFLD-SIFT is proposed to reduce the dimension around the key points under different initial scales. Meanwhile, RVM fuzzy integral fusion algorithm for behavior classification is presented. Finally, the proposed method and other relevant methods are compared through four kinds of evolution indexes and average recognition rates. The video sequence of KTH human action database and triple-cross verification method are used to test the recognition methods. Experimental results show that the proposed method achieves good recognition effect, and its average recognition rate rises by at least 2.3% compared to other mainstream methods for human action recognition.
[1] Poppe R.A Survey on Vision-Based Human Action Recognition.Image and Vision Computing, 2010, 28(6): 976-990 [2] Aggarwal J K, Ryoo M S. Human Activity Analysis: A Review. ACM Computing Surveys, 2011.DOI: 10.1145/1922649.1922653 [3] Danafar S, Gheissari N. Action Recognition for Surveillance Applications Using Optic Flow and SVM // Proc of the Asian Conference on Computer Vision. Tokyo, Japan, 2007: 457-466 [4] Wang L, Hu W M, Tan T N. A Survey of Visual Analysis of Human Motion. Chinese Journal of Computers, 2002, 25(3): 225-237 (in Chinese) (王 亮,胡卫明,谭铁牛.人运动的视觉分析综述.计算机学报, 2002, 25(3): 225-237) [5] Du Y T, Chen F, Xu W L, et al.A Survey on the Vision-Based Human Motion Recognition. Acta Electronica Sinica, 2007, 35(1): 84-90 (in Chinese) (杜友田,陈 峰,徐文立,等.基于视觉的人的运动识别综述.电子学报, 2007, 35(1): 84-90) [6] Wu Q X, Deng F Q, Kang W X. Human Action Recognition in Complex Scenes Based on Fuzzy Integral Fusion. Journal of South China University of Technology: Natural Science Edition, 2012, 40 (1): 146-151 (in Chinese) (吴秋霞,邓飞其,康文雄.基于模糊积分融合的复杂场景下人体行为识别.华南理工大学学报:自然科学版, 2012, 40(1): 146-151) [7] Zhang Y. Objects Recognition and Unusual Events Modelling & Analysing in Intelligent Video Surveillance. Ph.D Dissertation. Shanghai, China: Shanghai Jiao Tong University, 2009 (in Chinese) (张 一.智能视频监控中的目标识别与异常行为建模与分析.博士学位论文.上海:上海交通大学, 2009) [8] Zhao M M, Cao J Q. Image Registration Algorithm of SIFT Based on Edge and Corner Point. Journal of Chongqing Jiaotong University: Natural Sciences, 2013, 32(4): 721-724 (in Chinese) (赵萌萌,曹建秋.基于边缘角点的SIFT图像配准算法.重庆交通大学学报:自然科学版, 2013, 32(4): 721-724) [9] Yi J K, Bian G H, Jiang D G. Optimization of Scale-Invariant Feature Transform (SIFT) Feature Extraction. Journal of Beijing University of Chemical Technology: Natural Science, 2013, 40(1): 115-119 (in Chinese) (易军凯,边锆辉,姜大光.尺度不变特征转换特征提取优化算法研究.北京化工大学学报:自然科学版, 2013, 40(1): 115-119) [10] Liu Z, Xing B C, Chen Y Y. An Efficient Parallel PCA-SIFT Algorithm for Multi-core Processor. Journal of National University of Defense Technology, 2012, 34(4): 103-107 (in Chinese) (刘 仲,邢彬朝,陈跃跃.一种面向多核处理器的高效并行PCA-SIFT算法.国防科技大学学报, 2012, 34(4): 103-107) [11]Wang X, Fan J L. Multi-kernel Fisher Discriminant Analysis Based on Diversity Sample. Modern Electronics Technique, 2012, 35(11): 73-76 (in Chinese) (王 昕,范九伦.基于多样本的多核Fisher判别分析研究.现代电子技术, 2012, 35(11): 73-76) [12] Shen S H, Liu Y C. Efficient Multiple Faces Tracking Based on Relevance Vector Machine and Boosting Learning. Journal of Visual Communication and Image Representation, 2008, 19(6): 382-391 [13] Ko Y C, Fujita H, Tzeng G H. A Fuzzy Integral Fusion Approach in Analyzing Competitiveness Patterns from WCY 2010. Knowledge-Based Systems, 2013, 49: 1-9 [14] Ling Z G,Liang Y, Pan Q, et al. Human Action Recognition Based on Tensor Subspace Learning. Journal of Image and Gra-phics, 2009, 14(3): 394-400 (in Chinese) (凌志刚,梁 彦,潘 泉,等.基于张量子空间学习的人行为识别方法.中国图象图形学报, 2009, 14(3): 394-400) [15] Yan S C, Xu D, Yang Q, et al. Discriminant Analysis with Tensor Representation // Proc of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition. San Diego, USA, 2005, I: 526-532 [16] Tipping M E. Sparse Bayesian Learning and the Relevance Vector Machine. Journal of Machine Learning Research, 2001, 1: 211-244 [17] Temko A, Macho D, Nadeu C. Fuzzy Integral Based Information Fusion for Classification of Highly Confusable Non-speech Sounds. Pattern Recognition, 2008, 41(5): 1814-1823 [18] Zhai J H, Wang X Z, Zhang S F. Integration of Multiple Fuzzy Decision Trees Based on Fuzzy Integral. Journal of Computer Research and Development, 2009, 46(3): 470-477 (in Chinese) (翟俊海,王熙照,张素芳.基于模糊积分的多模糊决策树融合.计算机研究与发展, 2009, 46(3): 470-477) [19] Xue Z J, Dong M, Song W, et al. Infrared Gait Recognition Based on Wavelet Transform and Support Vector Machine. Pattern Recognition, 2010, 43(8): 2904-2910 [20] Wang X, Mu X, Song S L, et al. Human Action Recognition Based on Local Binary Pattern Feature in Video Sequences. Opto-Electronic Engineering, 2013, 40(3): 108-114 (in Chinese) (王 宪,慕 鑫,宋书林,等.视频序列中基于 LBP 特征的人体行为识别.光电工程, 2013, 40(3): 108-114) [21] Huang J, Kong L F, Li H T. Human Behavior Recognition Based on HMM. Computer Simulation, 2011, 28(7): 245-248 (in Chinese) (黄 静,孔令富,李海涛.基于傅里叶-隐马尔科夫模型的人体行为识别.计算机仿真, 2011, 28(7): 245-248) [22] Hong Y G. Human Action Recognition Based on Improved Canny Operator and Neural Network. Computer Engineering and Applications, 2013, 49(8): 156-159, 252 (in Chinese) (洪运国.基于改进Canny算子和神经网络的人体行为识别模型.计算机工程与应用, 2013, 49(8): 156-159, 252)
2014, Vol. 27 
