I am using The YOLOv7 for object detection, and I trained it on a custom dataset and I use it in real time with a fullHD Camera.
My problem is that when using it, it uses a 640*480 cropped image.
I tried to tweak the code by using cap.set(cv2.CAP_PROP_FRAME_WIDTH, 1920) cap.set(cv2.CAP_PROP_FRAME_HEIGHT, 1080) but it doesn't work.
What is the solution? Thanks
