divide

红外与可见光实时融合算法 ------------------------------------ 若不能读取xvid视频，请安装XviD-1.1.3-28062007.exe。 ----------------------------------- 请双击*.bat文件看程序运行结果。 也可以在命令行下，输入“evs3 参数文件.xml 配准.xml”，以融合指定的多源视频图像。 ---------------------------------- 首先，离线地时空配准多源视频，并保存结果到xml文件。然后融合配准后的图像。融合过程是，通过分解图像为Laplacian金字塔，分别对图像的高频和低频按照规则合并，最后将Laplacian金字塔重新组合为图像。(Real time fusion algorithm for infrared and visible light------------------------------------If you cannot read XviD video, please install XviD-1.1.3-28062007.exe.-----------------------------------Double-click the*.bat file to see the program running results.Can also be in the command line, " evs3 input parameter file .Xml .Xml registration ", by the fusion of the specified source video image.----------------------------------First of all, offline space-time registration source video, and save the results to a XML file. After registration and fusion image. Through the decomposition of the image fusion process, Laplacian Pyramid, were on the high-frequency and low frequency image in accordance with the rules of the merger, Laplacian Pyramid reassembled for image.)

说明：  自由曲面光学透镜的图像变换面形测量系统与精度建模(Free-form optical lens image transformation shape measurement system with precision modeling)

检测鼠标信息，C++Builder精选编程学习源码，很好的参考资料。(Detecting mouse information, C++Builder select learning programming source code, a good reference.)

Large Deviations for Gaussian Queues - Modelling Communication Networks - Michel Mandjes

说明：  雷达遥感图像的小波去噪程序，可有效去除噪声(radar remote sensing image wavelet denoising procedures, which can effectively remove noise)

说明：  在VC中实现了使用Moravec算子对遥感影像进行特征点的提取，缺点是速度较慢。不过提取效果不错！(Implemented in VC using Moravec operator of remote sensing images feature point extraction, the disadvantage is slow. However, extracting good results!)

可以动态调节运行环境的参数，重要参数的提取，利用自然梯度算法。( Can dynamically adjust the parameters of the operating environment, Extract important parameters, Use of natural gradient algorithm.)

图像去噪，包括小波去噪、均值滤波、中值滤波、自适应维纳滤波等(Image denoising)

说明：  ORL人脸图像库，共40人，每人10幅图像，其中每人的前5幅作为训练样本，后5幅作为测试分类样本，统计正确分类率。分类准则为最近邻规则。 真实的图像尺寸为112x92，列向量堆积对应人脸库矩阵的每一列。 (ORL face image database, a total of 40 per 10 images, each of which the first five as training samples, after the 5 categories as a test sample, correct classification rate statistics. Classification criteria for the nearest neighbor rule. The real image size is 112x92, the corresponding column vector face database matrix accumulation of each column.)

使用canny算子进行边缘检测的代码，效果很好(Using the canny edge detection operator code, good results)