Hi, I right now have a python program that detects objects from live video stream, following Hello AI world and NVIDIA’s youtube channel. I have modified it so that on key “q” press, the class it is currently detecting is printed out to the terminal. However, I would like to modify it so that on button press, the current frame which is being streamed in should be captured and saved in some directory. Furthermore, I would like it so that on detecting a “person” it should save the frame and save it in a directory. I am struggling now with how to capture the frame. I just need to substitute the printing of the class with the “snapshot” and I will have the first step done. Any suggestions on how I can just take the current frame and save it as an image?
(I am a intern, still learning the basics of the nano and AI in general)