opencv 图像特征检测算法akaze

对象检测代码 hog+svm，压缩包里含有行人，车辆等分类器，可以直接检测-

用python3+opencv3做的中国车牌识别，包括算法和客户端界面。(Chinese license plate recognition using python3+opencv3, including algorithm and client interface.)

基于Python的分水岭算法可以自动分割连接的图像(Python - based watershed algorithm can automatically segment connected images)

说明：  高精度图像计算,使用OpenCV进行图像处理，深度学习比对(GaoJingDu Picture JiSuan)

整合了retinex 直方图均衡化 基于对数log 基于拉普拉斯算子对视频进行实时增强

这是我做毕设时，用到的机器视觉关于OPENCV方面的文献，主要是尺寸测量方面的(This is what I do when the complete set, used in machine vision OPENCV the literature on the main aspects is the size measurement)

opencv 开发学习资料 只是一部分的类使用说明(Only a part of the class to use shows that the development of learning materials opencv)

算法思想来自于网上资源，先使用图像边缘和车牌颜色定位车牌，再识别字符。车牌定位在predict方法中，为说明清楚，完成代码和测试后，加了很多注释，请参看源码。车牌字符识别也在predict方法中，请参看源码中的注释，需要说明的是，车牌字符识别使用的算法是opencv的SVM， opencv的SVM使用代码来自于opencv附带的sample，StatModel类和SVM类都是sample中的代码。SVM训练使用的训练样本来自于github上的EasyPR的c++版本。由于训练样本有限，你测试时会发现，车牌字符识别，可能存在误差，尤其是第一个中文字符出现的误差概率较大。(Bits in the predict method, for clarity, add a lot of comments after completing the code and testing, see the source code. License plate character recognition is also included in the predict method. Please refer to the comments in the source code. It needs to be noted that the algorithm used for license plate character recognition is SVM of opencv. The training sample used for SVM training comes from the c++ version of EasyPR on github. Due to limited training samples, you may find errors in license plate character recognition when testing, especially the error probability of the first Chinese character is large.)

I want ot just dowlnoad one thing please.... one

对视频进行火焰检测，并实现对火焰的跟踪。(The video is detected by flame, and the flame is tracked)