算法步骤：
（1）　　计算图像直方图，灰度级为Ns，图像的平均灰度值为Tc；
（2）　　分别计算第一个灰度级和其余灰度级的信息熵并求和，前两个灰度级和其余灰度级的信息熵并求和，以此类推计算Ns个和；
（3）　　找到H在Tc附近最大值的位置Tenf即为阈值；
（4）　　图像中大于阈值的像素置1，其余置0而实现二值分割。

算法的核心：信息熵越大越有不确定性，分割出的图片越好。(Algorithm steps:
(1) the histogram of the image is calculated, the gray level is Ns, and the average gray value of the image is Tc.
(2) calculate the information entropy of the first gray level and the rest of grayscale, and sum up the information entropy of the first two grayscale and the rest gray level, and calculate Ns sum by analogy.
(3) to find the location of the maximum value of H near Tc, Tenf is the threshold;
(4) 1 pixels larger than the threshold in the image, and the rest 0 to achieve the two value segmentation.

The core of the algorithm: information entropy is more uncertainty, the segmented image.)