Abstract:

This paper presents a SURF based target tracking and online object model update algorithm in complex scenarios and target shade, especially in the variation of target outline and scale. The algorithm initially extracts SURF points of the tracked target and RGB histograms of them and their adjacent areas with the robustness of SURF to image zoom, illumination variation and rotation. It then searches optimal matching SURF points in the current frame based on such characteristic as higher target similarity among the sequential frames. It calculates the exact position and size of the target according to the target model and its features, and then updates the target model based on the varied object status. Experimental results show that the algorithm can accurately localize a target.

Key words: complex scenarios, SURF, shade, target tracking, zoom