SAR

使用EM算法完成图像的分割，使用混合高斯模型，可以修改分类数。运行EM.m(Expectation Algorithm)

高光谱图像的分类，适合各种光谱数据或者RGB图像。只需要调整输出路径就好了。(Hyperspectral image classification, for a variety of spectral data or RGB images. Only you need to adjust the output path just fine.)

C++ code implementing the estimation of errors-in-variables models under point dependent noise. It includes examples for linear, ellipse, fundamental matrix and trifocal tensor estimation. The theory is described in A general method for errors-in-variables problems in computer vision(err)

k均值聚类，聚为2类，结果显示散点图，红色十字表示聚类中心(K-means clustering, the class number is 2, and the results showed that the clustering center)

图像检索。通过计算所要检索的图像与图像库中的距离dis，dis最小者与检索图像同类。(image retrieval. Calculated to be indexed images and pictures of the dis the distance, the smallest dis with similar image retrieval.)

图像增强效果评价指标：1.欧拉范数误差指标 2.等效视数 3.边缘保持指数 4.对比度增强指数(Evaluating results of image enhancement: 1. Oula Fan error index number 2. Equivalent number of 3. Edge-preserving index 4. Contrast enhancement index)

可以实现图像的重采样，通过设置参数来减小或增大图像尺寸(Resampling of the image can be achieved by setting parameters to decrease or increase the image size)

说明：  用c写的jpg编解码，包括jpeg2bmp和bmp2jpeg，可在vc6下调试。(With c written jpg codec, including jpeg2bmp and bmp2jpeg, may be under vc6 debugger.)

针对分水岭算法存在的过分割问题以及VCH-F1切片图像的特点，提出一种能够有效消除局部极小值和噪声干扰的自动分割方法。首先比较彩色分量梯度图，选择分量图像的梯度信息，达到有效提取图像边缘信息的目的；然后提出基于多阈值分割的方法消除无效梯度信息；最后介绍了算法的步骤及结果。实验结果证明，通过该方法处理的梯度图像再进行分水岭算法处理，即使不进行区域合并也能达到很好的效果。(Watershed algorithm for over-segmentation problem of the existence of VCH-F1 as well as the characteristics of the image slice, a can effectively eliminate the local minimum value and noise automatic segmentation method. Comparing the first color component gradient map, select the quantity of image gradient information, achieve an effective extraction of image edge information purposes then made based on multi-threshold segmentation method to eliminate invalid gradient information Finally introduce the steps of the algorithm and results. Experimental results show that the adoption of the method of treatment of gradient watershed algorithm for image re-treatment, even if there is no regional merger will also achieve good results.)

何凯明暗通道先验法进行单幅图像去雾的实现。 严格按照每一步编程。步骤清晰，代码易懂，更改图片路径后便可执行。(Make a single image to achieve HE dark fog channel priori method. In strict accordance with each step programming. Step clear, easy to understand code, change the image path after the executable.)