|
|
Object counting based on regularized risk minimization |
WU Peng-zhou, YU Hui-min, ZENG Xiong |
Department of Information Science and Electronic Engineering, Zhejiang University, Hangzhou 310027, China |
|
|
Abstract Current studies of object counting commonly ignore the spatial information of objects and encounter difficulties when dealing with high density object groups. An object counting approach was presented which estimated a density function for every input image, whose integral over any image region gives the count within that region. A parametric model of density function was built by mathematical derivation. The conditions that feature vectors should satisfy and the effects of image segmentation were analyzed. The parameters in the model of density function were estimated by the principle of regularized risk minimization, and the density function empirical risk minimization can be boiled down to a linear program. Experimental results show that the method can accurately estimate the object counts for testing images with only a few training images. For high density object groups, the approach also gives counts, not only density levels.
|
Published: 04 August 2014
|
|
基于正则化风险最小化的目标计数
针对现有研究对目标空间信息的普遍忽视及其对高密度群体精确计数的困难,提出对输入图像估计一个密度函数.通过该函数在任意图像区域上的积分得出该区域中的目标个数.经过数学推导,得到密度函数的参数化模型,分析特征向量需要满足的条件以及加入图像分割对结果的影响. 由正则化风险最小化原理求取密度函数模型的参数,将密度函数的经验风险最小化问题简化为一个线性规划问题. 实验表明,该方法只需少量图像进行训练, 就可以准确地估计测试图像的目标数目. 对于高密度群体,该方法能够给出目标计数, 而不仅是密度等级估计.
|
|
1] LIN S F, CHEN J Y, CHAO H X. Estimation of number of people in crowded scenes using perspective transformation [J]. IEEE Transactions on Systems, Man and Cybernetics, Part A: Systems and Humans, 2001, 31(6): 645-654.
[2] DESCOMBES X, MINLOS R, ZHIZHINA E. Object extraction using a stochastic birth-and-death dynamics in continuum [J]. Journal of Mathematical Imaging and Vision, 2009, 33(3): 347-359.
[3] DAVIES A C, YIN J H, VELASTIN S A. Crowd monitoring using image processing [J]. Electronics and Communication Engineering Journal, 1995, 7(1):37-47.
[4] PARAGIOS N, RAMESH V. A MRF-based approach for real-time subway monitoring [C]∥Proceedings of IEEE Conference on Computer Vision and Pattern Recognition. [S.l.]: IEEE, 2001: I-1034-I-1040.
[5] MA R, LI L, HUANG W, et al. On pixel count based crowd density estimation for visual surveillance [C]∥Proceedings of IEEE Conference on Cybernetics and Intelligent System. [S.l.]: IEEE, 2004: 170-173.
[6] VELASTIN S A, YIN J H, DAVIES A C, et al. Automated measurement of crowd density and motion using image processing [C]∥Proceedings of IET International Conference on Road Traffic Monitoring and Control. [S.l.]: IET, 1994: 127-132.
[7] REGAZZONI C S, TESEI A. Distributed data fusion for real-time crowding estimation [J]. Signal Processing, 1996, 53(1): 47-63.
[8] CHO S Y, CHOW T W S, LEUNG C T. A neural-based crowd estimation by hybrid global learning algorithm [J]. IEEE Transactions on Systems, Man, and Cybernetics, Part B: Cybernetics, 1999, 29(4): 535-541.
[9] MARANA A, DA COSTA L, LOTUFO R, et al. On the efficacy of texture analysis for crowd monitoring [C]∥Proceedings of IEEE International Symposium on Computer Graphics, Image Processing, and Vision. [S.l.]: IEEE, 1998: 354-361.
[10] LI W, WU X, MATSUMOTO K, et al. Crowd density estimation: an improved approach [C]∥Proceedings of IEEE International Conference on Signal Processing. [S.l.]: IEEE, 2010: 1213-1216.
[11] VERONA V V, MARANA A N. Wavelet packet analysis for crowd density estimation [C]∥Proceedings of the IASTED International Symposia on Applied Informatics. Innsbruck: [s.n.], 2001.
[12] RAHMALAN H, NIXON M, CARTER J. On crowd density estimation for surveillance [C]∥Proceedings of The Institution of Engineering and Technology Conference on Crime and Security. [S.l.]: IET, 2006: 540-545.
[13] WU X Y, LIANG G Y, LEE K K, et al. Crowd density estimation using texture analysis and learning [C]∥ Proceedings of IEEE International Conference on Robotics and Biomimetics. [S.l.]: IEEE, 2006: 214-219.
[14] CHAN A B, LIANG Z, VASCONCELOS N. Privacy preserving crowd monitoring: counting people without people models or tracking [C]∥Proceedings of IEEE Conference on Computer Vision and Pattern Recognition. [S.l.]: IEEE, 2008: 17.
[15] CHAN A B, VASCONCELOS N. Bayesian Poisson regression for crowd counting [C]∥Proceedings of IEEE International Conference on Computer Vision. [S.l.]: IEEE, 2009: 545-551.
[16] RYAN D, DENMAN S, FOOKES C, et al. Crowd counting using multiple local features [C]∥Proceedings of IEEE Conference on Digital Image Computing: Techniques and Applications. [S.l.]: IEEE, 2009: 81-88.
[17] SCHREIBER D, RAUTER M. A CPU-GPU hybrid people counting system for real-world airport scenarios using arbitrary oblique view cameras [C]∥Proceedings of IEEE Conference on Computer Vision and Pattern Recognition Workshops. [S.l.]: IEEE, 2012: 83-88.
[18] MURPHY K P. Machine learning: a probabilistic perspective [M]. [S.l.]: MIT, 2012: 204-207.
[19] LEMPITSKY V, ZISSERMAN A. Learning to count objects in images [C]∥Proceedings of Neural Information Processing Systems (NIPS). Vancouver: Curran Associates Inc,2010.
[20] BENTLEY J L. Programming pearls: perspective on performance [J]. Communications of the ACM, 1984, 27(11): 1087-1092.
[21] BENTLEY J L. Programming pearls: algorithm design techniques [J]. Communications of the ACM, 1984, 27(9): 865-871.
[22] SRA S, NOWOZIN S, WRIGHT S J. Optimization for machine learning [M]. [S.l.]: MIT, 2012: 185-196.
[23] LOWE D G. Distinctive image features from scale-invariant keypoints [J]. International Journal of Computer Vision, 2004, 60(2): 91-110. |
|
Viewed |
|
|
|
Full text
|
|
|
|
|
Abstract
|
|
|
|
|
Cited |
|
|
|
|
|
Shared |
|
|
|
|
|
Discussed |
|
|
|
|