说明：  算法通过将二维直方图分为四部分：背哥、目标．受噪声干扰的背景和受噪声干扰的目标，以选取这四部分的信息熵的和最大作为闽值的选取准则。访方法有三个优点：尽可能包括背罱部分和目标部分；可以有效地提高对噪声敷据的鲁棒性；不过度地引入噪声和边缘信息。实验结果表明，谊方法具有较好的图像分割效果。
(Through two-dimensional histogram algorithm will be divided into four parts: back elder brother, target. By noise interference background and the noise interference goal, in order to select the four parts of the information entropy and the maximum as fujian value selection criteria. Visit method has three advantages: as far as possible including back Lan part and target parts Can effectively improve the noise apply according to robustness Not overly into noise and edge information. Experimental results show that the method has good friendship image segmentation effect. )