Noise-resilient detection of moving objects based on spatial-temporal blocks
David D. Pokrajac, Vesna Zeljković, Longin Jan Latecki · 2005
Abstract- In this paper we discuss the resilience of moving objects detection algorithm based on spatiotemporal blocks on various types of additive and multiplicative noise. After a given video is decomposed into the spatiotemporal blocks, the algorithm uses dimensionality reduction technique to obtain a compact vector representation of each block and to suppress the influence of noise. We evaluate the algorithm performance by comparing “ground truth ” (hand-labeled moving objects) to properly defined spatial-windows based evaluation statistics. Our results on a PETS repository video show that detection and tracking of moving objects is substantially improved in presence of Gaussian, speckle, multiplicative and Poisson noise.