融合量子克隆进化与二维Tsallis熵的医学图像分割算法

Medical Image Segmentation Algorithm Based on Quantum Clonal Evolution and Two-Dimensional Tsallis Entropy

针对进化算法用于图像分割时收敛速度慢、易早熟的缺点,提出一种改进的量子克隆进化算法.首先利用量子空间的多样性丰富种群信息,在量子变异中根据适应度的不同对个体施以不同的混沌扰动,以克服量子门旋转方向单一、大小固定的缺陷,避免种群陷入局部早熟;然后利用克隆算子将最优个体信息扩充至下一代,以提高其局部寻优能力,加快收敛速度;最后将此算法用于寻找二维Tsallis熵的最佳阈值,实现了对图像的分割.实验结果表明,该算法有效地解决了进化算法收敛速度慢和容易陷入局部极值的问题,而且在分割速度和精度上得到了较大提高,分割效果良好,可以满足医学图像三维重建要求.

The paper proposes an improved clone quantum evolutionary algorithm in view of the shortcomings existed in the image segmentation by evolutionary algorithm such as slow convergence and easy prematurity.The improved algorithm is able to overcome the weaknesses like the singe variation of quantum gate,the fault of fixed-size,and avoid the prematurity in small parts by using diverse population information in quantum space and imposing different chaotic perturbation on each unit which depends its fitness in quantum variation.The algorithm also can partially improve optimization ability,increase the converging speed and transmit the information of optimal unit to next generation by using clonal operator.The algorithm will be applied to seek the two dimension optimal Tsallis entropy,divide the picture eventually.It is showed that the algorithm can not only effectively solve the problems of slow converging and liability to local extremum,but considerably increase both dividing speed and precision,it meets the request of rebuilding in three dimension.