基于改进量子遗传算法的图像边缘检测研究

摘要

针对量子遗传算法在图像边缘检测存在的不足,提出一种改进的措施.首先对量子比特实施变换来控制量子态的演化和传递,对量子门旋转角大小增设系数来改变控制算法的收敛速度,旋转角方向调整系数避免部分个体因种群规模较大而使得角度方向改变值过小；然后进行量子比特幅编码,令有基因位的概率幅相同,使在整个解空间中所有可能解的取值概率相同；最后给出了图像边缘评价函数.实验仿真通过不同算法对比,本文算法对图像边缘检测效果好.%Aiming at the Quantum Genetic Algorithm's shortcomings in the target edge detection in an image,an improved measure is proposed. At first,the quantum bits' conversion is implemented to control the transformation and delivery of the evolution of quantum states. And additional factors are added to the rotation angle of quantum gates to control the convergence speed. And adjustment factor is introduced to the rotation direction to avoid too large population and too small angle change of direction;then quantum bit rate is coded to make the probability of gene locus to be same. Finally,the edge detection is processed. Simulation results show that the algorithm is more effective for image edge detection.