基于克隆粒子群算法的图像分割方法

最大熵阈值法是目前图像分割中应用最广泛的方法之一。为了快速准确地自动确定图像分割阈值,把克隆选择算法和粒子群算法相结合,提出克隆粒子群优化算法。利用这种改进方法对最大熵图像分割函数进行全局寻优。克隆选择算法和粒子群算法的结合克服了各自的缺点,克隆选择的多样性补偿了粒子群的多样性差的缺点,粒子群的快速性补偿了克隆选择的收敛速度慢的缺点。克隆粒子群方法克服了传统遗传算法易出现早熟、陷入局部最优等的问题,加快了图像分割函数收敛速度,最后能够快速准确地得到图像分割的最佳阈值。实验表明,改进后的算法分割速度较快,易于收敛到最优解,并且得到的分割阈值更加稳定。

Application of image segmentation based on clone selection and particle swarm algorithm

The optimal threshold is one of the most widely used methods in image segmentation.In order to automatically determine the threshold in image segmentation rapidly and exactly,an improved method combined clone selection algorithm with particle swarm algorithm is proposed in this paper,which is used to optimize the function of the optimal threshold image segmentation.The combination of clone selection algorithm with particle swarm algorithm overcomes both of their own shortcomings,the multiformity of clone selection algorithm compensates for that of the particle swarm algorithm and the fast convergence of particle swarm algorithm compensate for that of the clone selection algorithm.The method gets over the precocity of traditional genetic algorithm,meanwhile speeds up the convergence of the function of image segmentation and finally gets the optimal threshold for image segmentation.The experiment demonstrated that with comparatively rapid segmentation the improved algorithm was easier to converge at the optimal threshold which was more stable.