说明：  迭代最佳阈值分割——初始阈值选取为图像的平均灰度T0，然后用T0将图像的象素点分作两部分，计算两部分各自的平均灰度，小于T0的部分为TA，大于T0的部分为TB 计算,将T1 作为新的全局阈值代替T0，重复以上过程，如此迭代，直至TK 收敛，即TK＋1 ＝TK
(Iterative optimal threshold segmentation- the initial threshold selection for image average gray T0, and then use the T0 image pixel point into two parts, calculation of the two parts of the respective average gray, less than some of T0 for the TA, the part is greater than T0 for the TB calculation, will be T1 as a new threshold in place of the overall situation T0, repeat the above process, so iteration until convergence TK, that is, TK 1 = TK)