Multi-camera tracking algorithm study based on information fusion

摘要

Intelligent video surveillance technology is a part of pattern recognition used for analyzing, extracting and recognizing behavior characteristics of a moving target basing on computer algorithms. The target tracking algorithm combines particle filter and Mean-shift in overlapping multi-camera environment on the intelligent video surveillance system. Multi-camera tracking is studied using a fusion of SURF characteristics, color characteristics and geometric characteristic in Matlab. Experimental results show that the method of tracking between multiple cameras has good accuracy and stability.