pingjiahanshu

说明：  利用非局部均值的算法对图像做去噪，速度比较快，但效果一般，存在一定局限性(Using non-local mean algorithm to denoise image is faster, but the effect is general and there are some limitations.)

社会力模型，基于社会力的行人流模型，包括行人间作用力，行人与墙体等障碍物的作用力等。(Social force model)

给图像加椒盐和高斯噪声，以及空域滤波（中值和均值滤波）！(to Image pretzels and Gaussian noise, and the spatial filter (mean values and filter)!)

图像处理作业：基于高频补偿算法实现图像滤波及分析 数值分析作业：四种方法实现求解根号3的值(Image processing operations: image filtering and analysis of numerical analysis operations: four methods based on high-frequency compensation algorithm to solve the root of the value)

说明：  卷积神经网络（Convolutional Neural Network）是一个专门针对图像识别问题设计的神经网络。它模仿人类识别图像的多层过程：瞳孔摄入像素；大脑皮层某些细胞初步处理，发现形状边缘、方向；抽象判定形状（如圆形、方形）；进一步抽象判定（如判断物体是气球）。(Convolutional neural network is a neural network specially designed for image recognition. It imitates the multi-layer process of human recognition image: pupil ingestion of pixels; some cells in the cerebral cortex preliminary processing to find shape edges and directions; Abstract judgment of shape (such as circles and squares); and further Abstract judgment (such as judging objects are balloons).)

图像金字塔配准，利用金字塔分解实现对图像的多分辨率的分析(Registration of the image pyramid, multi-resolution image pyramid decomposition analysis)

说明：  bp人工神经网络模式识别教程示例，研究人脸识别 手势识别有帮助(bp nervel network trainning and recognition ,e.g:face & dynamic hand gesture detection and face recognition .I hope it would be very useful for researcher !)

该工具箱可实现matlab编程和zemax通信，同时借助于matlab 强大的编程能力以及zemax的光学设计能力，为光学领域工作的科研人员提供了新的途径。(This program can realize the communication between the matlab and zemax, very helpful for optical engineer and student whose major in optical science.)

集成ITK、VTK、MFC读取DICOM的代码，建成win32程序即可。运行完美(Integrated ITK, VTK, MFC to read DICOM code, built win32 program can. Works perfect)

采集RGB-D深度图像生成点云矩阵，属于RGB-D SLAM的前端处理过程(Collecting depth image to generate point cloud)