一种新的机器人视觉图像阈值实时获取方法

A New Realtime Acquisition Approach for the Image Threshold of Robot Vision

为了解决机器人足球比赛中机器人视觉系统的图像阈值实时获取问题,提出了一种基于改进遗传算法的阈值寻优算法。该算法采用自适应交叉和变异的算子来进行图像阈值获取,不仅提高了阈值的选取速度,保证了搜索的稳定,还提高了机器人视觉系统的实时性和鲁棒性。实验结果表明该方法可行。

To solve the problem of real-time acquistion for the image threshold of robot vision in a robot soccer game,this paper proposes a threshold optimization strategy based on an improved genetic algorithm.This algorithm uses adaptive crossover and mutation operators to obtain the threshold for an image,which significantly accelerates the speed of selecting threshold value and ensures the stability for searching and enhances real time and robustness of robot vision system.Experimental result demonstrates that this method is feasible.