The Research And Realization Of Target Tracking Algorithms On Embedded Platform

Tracking of moving targets on the embedded platform is a comprehensive study technique, which it is a combination of automatic control theory, signal processing, computer vision, video image processing and other multiple disciplines. It has a wider application in military, civil, industrial, medical and other fields.This paper combines embedded systems and target tracking algorithm, presented study of the target tracking algorithm on embedded platform. The main content is the brief introduction to the overall functional design and embedded platforms ICETEK-DM6446-EVMS. It is the analysis and comparison of several image matching algorithms, and ultimately select cross-correlation algorithm, and divide the whole process into the target acquisition and target tracking, the acquisition section is used for confirmation of the initial target position, the tracking part uses the auto-adaptive wave gate and auto-adaptive template to accurately locate the target tracking.The experimental results show that the proposed tracking algorithm has a more stable tracking effect in target partial occlusion, instantaneous loss and other interference case.