| In this thesis, we propose a new algorithm for detecting and tracking multiple moving objects in both outdoor and indoor environments. The proposed method measures the change of a combined color-texture feature vector in each image block to detect moving objects. The texture feature is extracted from DCT frequency domain. An attributed relational graph (ARG) is used to represent each object, in which vertices are associated to an object's sub-regions and edges represent spatial relations among the sub-regions. Multiple cues including color, texture, and spatial position are integrated to describe each object's sub-regions. Object tracking and identification are accomplished by graph matching. The inexact graph matching technique enables us to track partially occluded objects and to cope with object articulation. An ARG adaptation scheme is incorporated into the system to handle the changes in object scale and appearance. The experimental results prove the efficiency of the proposed method. |
