When we pull video from a live-stream, we receive something like 30 or 60 frames each second. Computed PyTesseract on each frame is inefficient and unoptimized. We should add an command line option to only read every nth frame. Something like:
python score_detection.py --skip_frames 29
I'm already working on this, but I just wanted to mention this.
When we pull video from a live-stream, we receive something like 30 or 60 frames each second. Computed PyTesseract on each frame is inefficient and unoptimized. We should add an command line option to only read every nth frame. Something like:
python score_detection.py --skip_frames 29
I'm already working on this, but I just wanted to mention this.