In order to overcome the problems of poor universality and sensitivity of light change in the existing shadow detection methods based on optical properties, a moving shadow detection method based on space location feature is introduced. A calibrated binocular stereo camera is used to capture stereo background images at the same time. The position relationships between matching points from the pair of background images is calculated off-line. The cast shadows adhere to the background surface, and the matching points of that shadow have the same position relationships with that of the background stereo images, so the cast shadows can be removed by comparing the level of similarities of pixel values between the left and right images on the position of matching points in the background. Prior knowledge about neither the environment characteristics nor the lighting conditions is required in the proposed approach, and all kinds of complicated lighting conditions will not affect the detection precision. Experimental results demonstrate that this approach can effectively detect moving shadows.
[1] Cheung S C S, Kamath C. Robust background subtraction with foreground validation for urban traffic video[J]. Eurasip Journal on Applied Signal Processing, 2005, 2005(14): 2330-2340. [2] Rao B, Zheng G, Chen T M, et al. An efficient hierarchical method for image shadow detection[C]//2nd International Workshop on Knowledge Discovery and Data Mining. Piscataway, NJ, USA: IEEE, 2009: 622-627.[3] Prati A, Mikic I, Trivedi M M, et al. Detecting moving shadows: Algorithms and evaluation[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2003, 25(7): 918-923. [4] Martel-Brisson N, Zaccarin A. Learning and removing cast shadows through a multidistribution approach[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2007, 29(7): 1133-1146. [5] Tatsuya T, Atsushi S, Daisaku A, et al. Non-parametric background and shadow modeling for object detection[M]//Lecture Notes in Computer Science, vol.4843. Heidelberg, Germany: Springer-Verlag, 2007: 159-168.[6] Joshi A J, Papanikolopoulos N P. Learning to detect moving shadows in dynamic environments[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2008, 30(11): 2055-2063. [7] Leone A, Distante C. Shadow detection for moving objects based on texture analysis[J]. Pattern Recognition, 2007, 40(4): 1222-1233. [8] Cucchiara R, Grana C, Piccardi M, et al. Improving shadow suppression in moving object detection with HSV color information[C]//IEEE Conference on Intelligent Transportation Systems. Piscataway, NJ, USA: IEEE, 2001: 334-339.[9] Zhang W, Fang X Z, Yang X K, et al. Moving cast shadows detection using ratio edge[J]. IEEE Transactions on Multimedia, 2007, 9(6): 1202-1213. [10] 曹健,陈红倩,张凯,等.结合区域颜色和纹理的运动阴影检测方法[J].机器人,2011,33(5): 628-633. Cao J, Chen H Q, Zhang K, et al. Moving cast shadow detection based on region color and texture[J]. Robot, 2011, 33(5): 628-633.[11] 马颂德,张正友.计算机视觉[M].北京:科学出版社,1998: 72-73. Ma S D, Zhang Z Y. Computer vision[M]. Beijing: Science Press, 1998: 72-73.[12] 刘勇.基于图像的空间三维数据获取及建模[D].武汉:武汉大学,2004. Liu Y. 3D scene acquiring and modeling based on images[D]. Wuhan: Wuhan University, 2004.[13] Mei X, Sun X, Zhou M C. On building an accurate stereo matching system on graphics hardware[C]//IEEE International Conference on Computer Vision. Piscataway, NJ, USA: IEEE, 2011: 467-474.[14] Wang Z F, Zheng Z G. A region based stereo matching algorithm using cooperative optimization[C]//IEEE Conference on Computer Vision and Pattern Recognition. Piscataway, NJ, USA: IEEE, 2008: 1-8.[15] 仲思东,高智.序列图像编码的自动识别[J].计算机应用研究,2005(5): 148-150. Zhong S D, Gao Z. Automatic recognition of sequence image coding[J]. Application Research of Computers, 2005(5): 148-150.[16] Zhang X, Yang J. A novel algorithm to segment foreground from a similarly colored background[J]. International Journal of Electronics and Communications, 2009, 63(10): 831-840.