MCA

内含张正友，tsai的标定方法，可以求解摄像机内参数u轴尺度因子，v轴尺度因子，摄像机外参数：两摄像机坐标系原点之间的旋转矩阵和评议矩阵，可以得出摄像机像素参数，焦距，畸变系数，并且带有优化结果。(Containing Zhang Zhengyou, tsai calibration method that can solve the camera parameters u-axis scale factor v The axis scale factor, the camera extrinsic parameters: the rotation matrix between the two camera coordinate origin, and appraisal matrix can be drawn from the camera pixel parameters, focal length distortion coefficient, and with optimal results. )

本代码主要应用于图像压缩，代码包括采用二维离散余弦变换（DCT）进行图像压缩、采用小波变换进行图像压缩以及基于小波变换与混沌排序图像加密压缩、解密重构 hundungen 是ex3的子程序：产生混沌序列(This code is mainly used in image compression, including the use of code dimensional discrete cosine transform (DCT) for image compression, image compression using wavelet transform and wavelet transform compression and chaotic image encryption based sorting, decryption reconstruction)

本算法为经典的criminisi算法，直接可以用于图像修复，资料中含有测试图片，测试效果明显，可以直接运行。(It is the classic criminisi algorithm and can be directly used for image inpainting, the data contains test pictures, the test results are obvious, the code can be run directly.)

织物疵点检测和识别系统.采用2中方法，有gui界面，支持实际纺织物的检测(Fabric defect detection and identification systems. Using 2 methods, gui interface, support for the detection of the actual textile)

基于最大类间方差准则的阈值分割算法。其基本思想是对像素进行划分，通过划分得到的各类之间的距离达到最大，来确定合适的门限。(Based on the maximum between-cluster variance criteria threshold segmentation algorithm. The basic idea is to divide pixels, obtained through the division of the distance between the various types of maximum, to determine the appropriate threshold.)

图像检测的算法，Faster R-CNN算法，先对整张图像进行卷积计算，然后通过感兴趣区域池化层(RoI Pooling Layer)将选择性搜索算法推荐出来的候选区域和卷积网络计算出的特征映射图进行融合，得到候选区域对应的特征矢量，这种共享卷积计算的操作极大地减少了卷积计算的次数。而且这些特征矢量的维度统一，方便后续的分类工作。通过感兴趣区域池化层处理卷积特征，并将得到的特征送往两个并行计算任务进行训练，分类和定位回归。通过这些方法和改进的框架，Fast R-CNN 用更短的训练和测试时长，取得了比 R-CNN 更好的效果(Faster R-CNN algorithm first convolutes the whole image, then fuses the candidate regions recommended by the selective search algorithm and the feature mapping maps calculated by the convolution network through the RoI Pooling Layer to get the corresponding feature vectors of the candidate regions, which greatly reduces the number of convolution calculations. Moreover, the dimension of these feature vectors is unified, which facilitates the subsequent classification work. The convolution feature is processed by the pooling layer of the region of interest, and the obtained feature is sent to two parallel computing tasks for training, classification and positioning regression. Through these methods and improved framework, Fast R-CNN uses shorter training and testing time and achieves better results than R-CNN.)

一种新型的图像去噪方法，相较于已有的算法，优势在于较小的计算量以及相对更好的还原效果。(A new image denoising method, compared to the existing algorithms, the advantages and the relatively small amount of computation better reproduction.)

[SIAM]Nonlocal Active Contours该文件包含论文及代码，主要内容是在水平集图像分割中引入non-local的性质，构造能量泛函，进行图像分割(this file include both paper and code, the main idea is utilize the non-local peperty in level set method for image segmentation )

经典的Tarel去雾算法，可用于图像去雾的指标性能对比。(It is the classical dehazing algorithm, and it is can be used comparison with other algorithms.)

最经典之做，保证别人没有上传过 实现文章 “GrabCut" Interactive Foreground Extraction using Iterated Graph Cuts 用graphcut实现图像分割，效果非常好(The most classic to do to ensure that other people do not realize uploaded article GrabCut Interactive Foreground Extraction using Iterated Graph Cuts with graphcut realize image segmentation, has very good results)