more pre-processing on captured frame #3
Description
currently the captured frame that is passed to the AI has no pre-processing unless a resizing and grey scaling. We need to apply more pre-processing and filtering to make it more similar to the learned images. For instance, the hand needs to be detected and cut off from the image and the background needs to be clear from any extra objects so that the passed in image would be similar to the ones in the asl_guide.png but in grey scale.
