hausdorff111

程序进行了两张图片的简单合并，适合对控件的使用进行学习(Procedures carried out two simple merger picture, suitable for the control of the use of learning)

PCB板检测的大概流程如下：首先存储一个标准PCB板图像作为参考标准，然后将待检测的PCB板图像进行处理，比较与标准PCB图像的差异。根据差异的情况来判断缺陷类型。(The approximate process of PCB board detection is as follows: first, a standard PCB board image is stored as a reference standard, and then the PCB board image to be detected is processed to compare the difference between the standard PCB image and the standard image. The type of defect is judged according to the difference.)

说明：  一个立体视觉匹配程序，采用matlab编写(A stereo vision matching program, prepared using matlab)

基于SVM与人工神经网络的车牌识别算法，使用了OpenCV的图像处理函数，在VS2013 + OpenCV 2.4.9平台上实现(SVM algorithm based on license plate recognition and artificial neural network, using OpenCV image processing function, implemented on VS2013+ OpenCV 2.4.9 platform)

模糊C均值分类，用于图像处理中 很好用的，极力推荐 (Fuzzy C-means classification, used for image processing in a very good use, and strongly recommend)

1.首先从文件菜单打开程序目录下的car.avi视频,系统将显示视频的第一帧。 2.点击背景提取菜单，稍后会提取出背景并显示在打开的第二个窗口中。 3.当提取出背景后，点击检测跟踪菜单，对车辆进行检测和跟踪。 4.点击轨迹绘制菜单，绘制车辆的轨迹。 注：由于使用OpenCV打开AVI视频文件需要视频解码器，故在程序运行前请在电脑上先安装DivX类型的的视频解码器，否则将出现无法打开文件的错误。本程序使用的是OpenCV2.1版本，在没有安装此版本软件的机器上运行时需要将cv210.dll,cvaux210.dll,cvcore210.dll,highgui210.dll,ml210.dll这五个文件拷贝至*.exe可执行文件下。(First, open the CR. AVI video from the file menu and the system will display the first frame of the video. 2. Click on the Background extract menu and the background will be extracted later and displayed in the open second window. 3. When the background is extracted, click on the Test Tracking menu to test and track the vehicle. 4. Click on the trajectory drawing menu to map the trajectory of the vehicle. Note: Since a video decoder is required to open an AVI video file using OpenCV, install a DivX-type video decoder on the computer before the program runs, otherwise an error will occur that can not open the file. The program uses OpenCV2 .1, which requires five files to be copied to *, cv210. DLL, cvaux 210. DLL, cvcore 210. DLL, highgui 210. DLL, ML 210. DLL. Exe under executable.)

vtk碰撞检测程序，基于obb包围盒的碰撞检测程序，需要用cmake进行编译(vtk collision detection procedure, based on obb bounding box collision detection program needs to be compiled with cmake)

说明：  机器学习专门研究计算机怎样模拟或实现人类的学习行为，以获取新的知识或技 能，为包括人脸识别领域的人工智能的发展做出了极大的贡献。本文探索如何应用机器学习中的一些技术，使计算机更好地完成人脸识别领域中的人脸检测和人脸验证。 在人脸检测方面，针对如何快速、准确地检出人脸的问题，基于DLIB中特征模型，可以快速提取检测出人脸，并且提取出人脸特征点,主要作用就是快速检测定位人脸。在训练过程中，引入了ResNet机器学习算法，该算法采用多层卷积神经网络结构，对人脸进行多层特征提取和描述，得到人脸特征描述符。通过测试可以证实了上述方法能够检测和识别人脸。(In terms of face detection, for the problem of how to detect faces quickly and accurately, based on the feature model in DLIB, you can quickly extract and detect faces, and extract face feature points, the main role is to quickly detect and locate faces. In the training process, the VGG-FACE machine learning algorithm is introduced. This algorithm uses a multi-layer convolutional neural network structure to perform multi-layer feature extraction and description on the face to obtain the face feature descriptor. Tests can confirm that the above method can detect and recognize human faces.)

说明：  目前一种快速稳健的卷积神经网络yolov3的代码，在protorch平台上运行(At present, the code of yolov3, a fast and robust convolutional neural network, runs on the platform of protorch)

利用MATLAB进行数值计算，得出HE模的色散曲线。(The dispersion curve of HE mode is obtained by numerical calculation by MATLAB.)