Perceptual Organization Based Computational Model for Robust Segmentation of Moving Objects

Abstract

The role of perceptual organization in motion analysis has heretofore been minimal. In this work we present a simple but powerful computational model and associated algorithms based on the use of perceptual organizational principles, such as temporal coherence (or common fate) and spatial proximity, for motion segmentation. The computational model does not use the traditional frame by frame motion analysis; rather it treats an image sequence as a single 3D spatio-temporal volume. It endeavors to find organizations in this volume of data over three levels-signal, primitive, and structural. The signal level is concerned with detecting individual image pixels that are probably part of a moving object. The primitive level groups these individual pixels into planar patches, which we call the temporal envelopes. Compositions of these temporal envelopes describe the spatio-temporal surfaces that result from object motion. At the structural level, we detect these compositions of temporal envelopes by utilizing the structure and organization among them. The algorithms employed to realize the computational model include 3D edge detection, Hough transformation, and graph based methods to group the temporal envelopes based on Gestalt principles. The significance of the Gestalt relationships between any two temporal envelopes is expressed in probabilistic terms. One of the attractive features of the adopted algorithm is that it does not require the detection of special 2D features or the tracking of these features across frames. We demonstrate that even with simple grouping strategies, we can easily handle drastic illumination changes, occlusion events, and multiple moving objects, without the use of training and specific object or illumination models. We present results on a large variety of motion sequences to demonstrate this robustness. c 2002 Elsevier Science (USA)

Space-time A Contrario Clustering for Detecting Coherent Motions

“…However, their work remains formal with no computational theory. More applied works are [7], [8] in which some motion structures are sought in spatio-temporal slices. Still more recently, Gryn, Wildes and Tsotsos [9] have specified even more precise motion templates, driven by the application, in other words trading generality for better computational efficiency.…”

Parallelizing motion segmentation by perceptual organization of XYT

“…This is not meant to be an exhaustive review, but is included for clarity. Those wishing more detail are referred to [7]. POM is a method of motion segmentation based on spatio-temporal volume analysis using perceptual organization principles.…”

“…Recently, Sarkar, Majchrzak, and Kormilli [7] presented a method for motion segmentation based on 3D spatiotemporal volume analysis and perceptual organization principles, referred to as the POM method in this paper. This method was shown to be robust with respect to image noise, drastic illumination changes, and occlusion events.…”

Video shot interpretation using principles of perceptual prominence and perceptual grouping in spatio-temporal domain

“…This involves use of Gestalt's principles like similarity, proximity, adjacency, parallelism etc over the low level visual patterns in the 2D space, and the principle of common fate that influences the strengthening of the groups in the temporal domain. There has been prior work [4], [5] on perceptual grouping in spatio-temporal domain using motion cues. [5] have formulated the problem as that of analyzing a 3-D volume of features where one of the axes is time.…”

“…There has been prior work [4], [5] on perceptual grouping in spatio-temporal domain using motion cues. [5] have formulated the problem as that of analyzing a 3-D volume of features where one of the axes is time. The feature trajectories are approximated as piecewise planar surfaces that are then grouped using properties of parallelism, proximity etc.…”