论文部分内容阅读
图像分割技术在图像分析和图像识别上具有重要意义.传统自适应遗传算法有可能使问题求解陷入局部最优解,而求得错误的图像分割阈值.为了得到最优的图像分割阈值,对交叉率和变异率公式进行了重构,使得交叉率和变异率在任何情况下都不为零.同时,以最大二维熵函数作为适应度函数,采用选择、交叉变异等遗传操作作搜索最优分割阈值.仿真实验表明,该方法可以有效地提高图像分割精度和计算速度.
Image segmentation technology is of great importance in image analysis and image recognition.Traditional adaptive genetic algorithm may cause the problem solving to fall into the local optimal solution and get the wrong image segmentation threshold.In order to obtain the optimal image segmentation threshold, Rate and mutation rate formula are reconstructed so that the crossover rate and mutation rate are not zero in any case.At the same time, using the maximum two-dimensional entropy function as the fitness function, genetic algorithm such as selection and crossover mutation is used to search the optimal Segmentation threshold.The simulation results show that this method can effectively improve the image segmentation accuracy and computational speed.