基于darknet的视频实时检测

darknet需要编译opencv，darkflow对视频无法实时检测（摄像头可以）
参考https://github.com/jplnasa5/Tracking-with-darkflow

1）改写run.py

#FLAGS.demo = "camera" # video file to use, or if camera just put "camera"

FLAGS.demo = "/Users/sisyphus/Tracking-with-darkflow/darkflow/video5.mp4"

FLAGS.model = "darkflow/cfg/yolo.cfg" # tensorflow model

FLAGS.load = "darkflow/bin/yolo.weights" # tensorflow weights

2）cd到文件夹，运行

python run.py