A region-level motion-based graph representation and labeling for tracking a spatial image partition

作者:

Highlights:

摘要

This paper addresses two image sequence analysis issues under a common framework. These tasks are, first, motion-based segmentation and second, updating and tracking over time of a spatial partition of an image. By spatial partition, we mean that constituent regions display an intensity, color or texture-based homogeneity criterion. Several issues in dynamic scene analysis or in image sequence coding can motivate this kind of development. A general-purpose methodology involving a region-level motion-based graph representation of the partition is presented. This graph is built from the topology of the spatial segmentation map. A statistical motion-based labeling of its nodes is carried out and formalized within a Markovian approach. Groups of spatial regions with consistent motion are identified using this labeling framework, leading to a motion-based segmentation that is both useful in itself and for propagating the spatial partition over time. Results on synthetic and real-world image sequences are shown, and provide a validation of the proposed approach.

论文关键词:Image sequence analysis,Motion-based segmentation,Partition tracking,Markov random fields

论文评审过程:Received 15 March 1999, Available online 7 June 2001.

论文官网地址:https://doi.org/10.1016/S0031-3203(99)00083-7