【摘要】 将贪婪思想引入到遗传算法中设计了贪婪遗传算法,以最大类间方差函数作为遗传算法中适应度的评价函数,利用贪婪算法局部搜索能力强和遗传算法全局搜索能力强的特点,实现图像阈值的自动选取。与基本遗传算法相比,本文算法具有更强的寻优性能,从而加快了收敛速度。实验表明基于贪婪遗传算法的阈值自动选取方法具有计算耗时少的特点,更适用于实时图像处理系统。更多还原

【Abstract】 In this paper, the greedy idea was introduced into the genetic algorithm. With the Otsu method being used as the fitness evaluation function, the greedy genetic algorithm was designed, which combines the local convergent ability of the greedy algorithm with the global exploring ability of the genetic algorithm. Compared with the simple genetic algorithm, the greedy genetic algorithm has stronger optimal searching ability and faster convergence speed. Experiment results indicate that the automatic choosing of image threshold based on the greedy genetic algorithm is fit for the real-time image dealing system.更多还原