基于蛙眼视觉特性的运动目标模糊化区域理解跟踪方法

A Fuzzy Region Understanding Tactic for Object Tracking Based on Frog's Vision Characteristic

动态场景下的运动目标检测与跟踪是计算机视觉研究的前沿方向,对场景的背景突变和目目标的外观突变的鲁棒性是当前研究的难点所在,针对这种情形,本文提出一种基于蛙眼视觉特性的鲁棒跟踪方法,该方法利用蛙眼视觉认知的生理特性和外部特性,设计了一种与之相应的模糊化区域理解的运动目标跟踪方法,针对实验室环境下的动态序列的实验结果验证了方法的有效性;并进一步将该方法与传统的Canny算子理解结果及经典的Mean shift算法理解结果进行对比,显示了方法的优越性。

The detection and tracking of moving objects under dynamic scenes is one of the hot topics in computer vision research. The problem is difficult when local scenes or object＇s appearance vary saliently. In this paper, we develop a robust intelligent tracking tactic based on the intrinsic and extrinsic features of frog＇s visual system. It is achieved through a so called ＂fuzzy region understanding＂. Experimental results on real lab-video and comparison with two traditional algorithms are reported to demonstrate the validity and robustness of our algorithm.