Pfinder: Real-Time Tracking of the Human Body
IEEE Transactions on Pattern Analysis and Machine Intelligence
Motion detection with nonstationary background
Machine Vision and Applications
Segmenting Foreground Objects from a Dynamic Textured Background via a Robust Kalman Filter
ICCV '03 Proceedings of the Ninth IEEE International Conference on Computer Vision - Volume 2
Data Clustering: Theory, Algorithms, and Applications (ASA-SIAM Series on Statistics and Applied Probability)
Online Learning for Matrix Factorization and Sparse Coding
The Journal of Machine Learning Research
Image segmentation in video sequences: a probabilistic approach
UAI'97 Proceedings of the Thirteenth conference on Uncertainty in artificial intelligence
Gaussian-Based codebook model for video background subtraction
ICNC'06 Proceedings of the Second international conference on Advances in Natural Computation - Volume Part II
| Hi-index | 0.00 |
Background detection is an important procedure for visual information processing with static cameras. In real applications, background always co-occurred with disturbance information such as small motions and unusual lighting conditions. In this paper, a robust method is proposed by using the spatial codebook method, where the elementary codebook generation unit is the pixels with their neighborhood ones. In the training procedure. features are first extracted for spatial unit. Second, a clustering technique using k-means method is adopted to generate the preliminary codebook. Then the outlier removal technique is used to obtain more descriptive codebook. In the testing procedure, if the pixel belongs to the background region, it should be represented by one of the generated codebooks. Experimental evaluations show that the proposed method is effective in clustered background.