hough变换 速度慢 不能滤除无效直线

说明：  svm 图像分割 用kmeans得到两类分割的图，在显示的图像中用鼠标取点得到2*num个坐标位置的二维向量，前num个为正样本，后num各为负样本(image segmentation)

对模糊图像进行复原（代码中样例）后，进行ISNR值的计算。由于复原后图像和原图像可能存在对齐问题，而该对齐问题不影响复原结果的主观效果，因此该代码首次考虑这一问题并给出相应的图像对齐及ISNR计算方法。(Restoration of blurred images (sample code), the calculation ISNR values. Due to possible alignment problems and the restored image after the original image, and the alignment problem does not affect the results of the subjective effects of recovery, so the code for the first time to consider this issue and the corresponding image alignment and ISNR calculation.)

说明：  基于matlab 的图像除雾算法。可以很大程度去除雾霾对图像的影响。(Image Defogging Algorithm Based on matlab. The influence of smog on images can be removed to a great extent.)

surf特征点提取与匹配，该算法能达到实时性的要求，不仅能用于两幅图的匹配，还能用于单幅图的复制粘贴取证(surf feature point extraction and matching, this algorithm can achieve real-time requirements, not only for the two graph matching, but also for single figure copy and paste forensics)

求解一个图像的曲率，程序完整，正确，可以用于图像处理(Solving an image of the curvature, the program complete, correct, can be used for image processing)

在IDL中运行，可以将多光谱、高光谱等多波段影像拆分为为单一波段影像(It works in IDL and it can split multibands into single band)

单尺度SSR(Single Scale Retinex)(A Retinex Function)

用kinect获取3D图像，并从3D图像中用OpenCV方法从深度图中获取梯度边缘，从而得到更多信息，做后续处理(Get with kinect 3D images, and get the edge gradient from the depth map from a 3D image using OpenCV method to obtain more information, do follow-up treatment)

基于Contourlet变换和SVM的SAR图像目标识别 (Contourlet transform and SVM based on the SAR image target recognition)

简单的帧间差分法，不过可以和其他方法结合来改进算法(A simple inter-frame difference method, but other methods can be combined to improve the algorithm for)