particlefilter

说明：  MFC结合VTK做图像处理的程序，实现了图像的显示和缩放功能。(MFC combines VTK to do image processing procedures, to achieve an image display and the zoom feature.)

说明：  基于各种算法的运动视频背景消除。使用opencv工具库，通过读取视频文件，处理视频，最后存储。(Background Subtract Algorithm)

几何变换，包括图形的基本变化，以及计算机图形学中的算法(Geometric transformations, including changes in the basic graphics, and computer graphics algorithms)

Graph Cutshi是最近在计算机视觉中研究和应用比较多的优化算法，典型的用于stereo matching, image restortion, texture synthesis等应用，以上代码给出了graph cut，mean shift, 以及相关的ppm/pgm图像借口，本来是用于stereo matching的，已经把相关的matching的实现代码给去掉了。大家可以以此为平台做一些相关的研究和应用。 (Graph Cutshi recently in computer vision research and application of a relatively large number of optimization algorithms, typical for the stereo matching, image restortion, texture synthesis applications, the above code gives the graph cut, mean shift, and the associated ppm/pgm images an excuse, it is used for stereo matching, and has been related to matching the code to realize removed. U.S. can also be used as a platform to do some related research and applications.)

这篇是96年的水印论文，全英文版，中文题目：一种安全强大的扩频水印技术(This is a 96-year watermark papers, all in English, Chinese subjects: a safe and powerful Spread Spectrum Watermarking)

一个很不错的指纹分析算法源代码，可以从采集的指纹图片中进行二次黑白化，平滑图像处理。可以对指纹图像特征信息提取，比对，以轮廓显示以及色彩覆膜特显等技术算法。主要的分析速度很快，值得学习的算法。要在指纹信息处理中比较不错，本站隆重推荐。(A very good fingerprint analysis algorithm source code, from the collection of fingerprint images, the second black and white, smooth image processing. Fingerprint image feature information extraction, compared to the contour display and color coated special significant technical algorithm. Analysis fast, it is worth learning algorithm. Fingerprint information processing site is highly recommended.)

基于样本的纹理合成加速方法，MRF，基于Li-Yi Wei的论文。(Acceleration method based on texture synthesis samples, MRF, based on Li-Yi Wei of paper.)

说明：  用canny算子来检测图像边缘，VC++(Using canny operator to detect image edge)

主要是用于雷达目标跟踪中的数据融合，有很好的作用(It is mainly used for data fusion in radar target tracking.)

车道偏离识别的程序，采用的是动态规划算法，对弯道也有比较高的识别率(lane departure algorithm, which has good performance even for curve lane.)