Design on Intelligent Video Surveillance System Based on Target Identification Algorithm

摘要

This paper makes study on the adjacent frame difference and algorithm realization of SOM i8dentification after improvement, of which it includes motion detection, target identification, the realized video surveillance module makes up the intelligent video surveillance that can reconstruct platform. Motion detection module adopts algorithm of adjacent frame difference after improvement, which can correctly mark the motion object. Target identification module adopts self-mapping nerve net after improvement, it is easier for hardware realization, and meanwhile the accuracy rate of identification is equal to classical algorithm.