Virtual Eye will constantly take pictures of the world around the user. It will then store this data to a temporary file and analyze that file using template matching to find if it can detect a template image such as a stop sign, and users can customize it by taking pictures of their relatives or loved ones and labeling them as their name. Then if the software finds the face or object in an image, it will alert the user through audio output via headphones.
Template matching files seems unusual. Why did you select this method over more popular image recognition methods like OpenCV? You're using Python and describe what sounds like a portable system, so are you using a Raspberry Pi?
Template matching files seems unusual. Why did you select this method over more popular image recognition methods like OpenCV? You're using Python and describe what sounds like a portable system, so are you using a Raspberry Pi?