Cmp_vis_canny

基于matlab的JPEG编解码程序。里面有2个程序，一个是程序是编解码过程，一个纯编码过程。含解释。(Program based on Matlab JPEG codec. There are two programs, one program is the process of encoding and decoding, a pure coding process. Containing explained.)

本程序实现了对影像特征点自动提取，利用Morevac、Forstner、Harris3个经典算子。在此基础上利用相关系数法实现影像自动匹配，并且引入最小二乘平差，使匹配点精度有所提高。 在搜索点过程中，利用了核线影像特性，对二维影像搜索使用了爬山法启发式搜索。对大数据量影像采用影像金字塔结构处理。 1、使用GDAL库读取影片，支持TIFF、PNG、JPEG、JPG、BMP、GIF、IMG格式读取。使用GDI绘图。 2、防止大数据量绘图视图闪烁，图片显示采用双缓存技术。 3、保存视图数据为图片文件，支持TIFF、PNG、JPEG、JPG、BMP、GIF格式保存。 4、TreeCtrl控件、ListCtrl控件的基本操作。 5、MFC单文档程序视图通讯、更换视图、视图分割。(This program of image features automatic extraction point, use Morevac, Forstner, Harris3 a classic operator. Based on the correlation coefficient method using the image matching, realization and introducing least squares adjustment, the matching point to improve precision. In the search process, the use of the image characteristics of nuclear line, 2d image search using the heuristic method of climbing. For large quantity of data using image processing image pyramid structure. 1, using film, read GDAL repository support JPEG, TIFF, PNG, GIF, BMP, JPG, read IMG format. Use GDI drawing. 2 and prevent large amount of data, drawing view pictures showed adopt double caching techniques. 3, save for the image file, views data support JPEG, TIFF, JPG, PNG, GIF, BMP format. 4 and TreeCtrl controls, the basic operation ListCtrl controls. 5 and MFC single document procedure view communication, replace view, view.)

图像的纹理分析应用实例,主要可用于模式识别当中(image texture analysis application, which can be used for pattern recognition)

说明：  criminisi算法。传统的criminisi图像修复算法，可对破损图像进行纹理修复(Criminisi algorithm. The traditional criminisi image restoration algorithm can repair the damaged image texture)

基于纹理和边缘方向特征融合的目标跟踪程序代码，是基于mean shift框架的(This paper proposes a powerful and robust local descriptor,called textureorientation descriptor(TOD). TOD consists of two components: texture and orientation. Considering that the human shape shows more short, fragmentary and vertical edges, the proposed TOD descriptor encodes the spatial information of pedestrian image. We combine the proposed TOD representation model with mean shift framework to perform pedestrian tracking. Results from experiments show the improved performance of the proposed algorithm, in the scenarios of heavy occlusions, illumination variations and similar objects interference, when compared with the other descriptors based trackers.)

实现直角坐标系至极坐标系下的图形图像变换。本方法为最新的快速变换方法，可应用于相位相关下的带旋转平移、缩放等变换。(the method can transform pic to radan.it can be used by PC to get the roation.)

基于亚像素运动估计算法的块匹配运动，有源码跟参考文献，初学的小伙伴可以看看(Block-matching motion estimation algorithm based sub-pixel motion, there is source code with references, beginner can look at a small partner)

说明：  简单的Canny边缘检测程序，另外包含霍夫变换检测道路直线。(Simple Canny edge detection procedures, including Hough transform detection of road lines.)

张正友标定算法，包含自己的测试图片以及实验分析。(Zhang Zhengyou calibration algorithm, contains its own test images and experimental analysis.)

说明：  实现车道线检测，运用Hough变换的原理(For the process of lane line detection, this paper first searches for the edge points of the lane line, mainly including image graying, image filtering, image edge enhancement and image binarization. Then the lane line edge line is fitted. This paper uses the Hough transform to perform a straight line fitting.)