Abstract: According to the advantages and disadvantages of the genetic algorithm and the minimum error thresholding method, a partial minimum error thresholding method of cavity image combined with genetic algorithm is developed by improving and combining genetic algorithm with minimum error thresholding method. This method uses the partial information of the cavity image to confirm the best threshold range, and uses simulated annealing algorithm to adjust the individual fitness degree, thus the premature convergence is avoided and the operation speed is improved. The results show that this method can segment the cavity image fast and accurately, and it is an effective method of cavity image segmentation.

Key words: threshold segmentation, minimum error thresholding method, genetic algorithm, local threshold, cavity image

摘要： 针对遗传算法和最小误差分割法各自的优缺点，将最小误差分割法与遗传算法进行改进并且相互结合，提出了一种结合遗传算法的局部最小误差孔穴图像分割法。该方法利用局部图像信息确定最佳阈值范围，并根据模拟退火思想对个体适应度进行自适应的调整，从而避免了早熟现象，提高了运算速度。实验结果表明：该方法不但能够准确地分割出孔穴图像，而且运算速度较快，是一种有效的孔穴图像分割方法。

关键词: 阈值分割, 最小误差法, 遗传算法, 局部阈值, 孔穴图像