This document describes a system for recognizing static hand gestures in American Sign Language (ASL). It discusses:
1) Using image processing techniques like segmentation, morphological filtering, and contour tracing to extract features from images of hand gestures. Features include centroid, shape, and number of fingers.
2) A classification approach using rule-based recognition based on the extracted features to identify the gesture as a particular letter in the ASL alphabet.
3) The implementation of the system in MATLAB, including steps for image capture, preprocessing, feature extraction, and classification. Experimental results demonstrating recognition of the letter "A" are shown.