opencv例程

说明：  用OPENCV和C++来实现SGBM算法进行双目测距的一个程序，普通双目摄像头可以直接使用(A program of sgbm algorithm for binocular distance measurement is implemented by opencv and C + +, which can be directly used by ordinary binocular cameras)

OpenCV 2 Computer Vision Application Programming Cookbook 的源码，采用C++风格，非常适合学习2.0以后的openCV 版本。(OpenCV 2 Computer Vision Application Programming Cookbook source, using C++ style, very suitable for learning 2.0 later OpenCV version. )

360-panorama，360度柱面图像拼接，程序在VS2008环境下开发，opencv2.4.3版本，程序可以直接运行，以将所需依赖项添加在文件夹中，源图像放在source_pic中，最后生成结果在result_pic中。 实现步骤：1先将图像投影到柱面坐标上，2提取图像SIFT描述子，3将SIFT描述子配对，4用RANSAC算法根据配对成功的描述子统计出变换矩阵T，5图像拼接。 对于拼接后的图像，没有做进一步细化处理，使用者可以在生成的图像上再做进一步编辑。(360-panorama 360 degree cylindrical image stitching program developed in VS2008 environment, opencv2.4.3 version, the program can be run directly, add the required dependencies in a folder, the source image on the source_pic in, and finally generate results In result_pic. Implementation steps: first image projected onto the cylindrical coordinates, extract images SIFT descriptor, the SIFT descriptor paired with an the RANSAC algorithm based on the description of the pairing is successful statistics the transformation matrix T 5 image mosaic. For the image with no further thinning, the user can do further editing on the generated image.)

基于OPENCV实现的粒子滤波运动目标跟踪源程序(Particle Filter Moving Object Tracking based on OPENCV)

VS2013 Opencv2.4.11下轮廓检测代码，此代码加入灰度直方图均衡化处理。 (VS2013 Opencv2.4.11 contour detection code, this code to join the gray-level histogram equalization processing. )

实现了人脸检测，利用Haar级联，将检测到的人脸戴上面具。

说明：  图像修复 实现以c++和opencv为环境，修补图像的算法(image inpainting . based on c++ and opencv to restore the picture.)

对马路车流量的统计，外加对每辆经过车辆的车速计算(On road traffic statistics, plus on every passing vehicle speed calculation)

VC+opencv 实现建筑物梯形畸变校正 radon变换 sobel边缘检测 直方图均衡图像增强 radon直线检测 仿射变换等算法(VC+ opencv trapezoidal distortion correction to achieve the building radon transform sobel edge detection image enhancement histogram equalization radon affine transform line detection algorithm)

这是一个基于多摄像头环境下的多目标实时跟踪程序，使用了intel的IPP图像处理库，是实时性得到了很大的提升，代码主要使用粒子滤波算法实现多目标跟踪，同时使用了在线学习方法，提升了系统鲁棒性，还是用了haar特征、adaboost构造强分类器。同时还有论文：MCT_DISTRIBUTED PARTICLE FILTER TRACKING WITH ONLINE MULTIPLE INSTANCE； MCT_Multiple View Discriminative Appearance Modeling