基于人工神经网络的红外目标识别与跟踪，运行环境为VISUAL STUDIO OPENCV3环境-Based on artificial neural network infrared target recognition and tracking, the operating environment for the VISUAL STUDIO OPENCV3 environment

用vs2013 opencv3.0 实现了视频中火焰检测，test视频是打火机，和小火堆，都可以进行识别，是很好的颜色识别。(With vs2013 opencv3.0 video flame detection, test video is a lighter, and small fire, can be identified, is a good color recognition.)

opencv3.3以上的版本调用caffe模型提前训练好的模型进行图片识别。(The above version of opencv3.3 calls the Caffe model in advance training model for image recognition.)