This submission includes three demos that were presented in a phased approach in the Image Webinar recorded on January 19th, 2006.
Phase 1: We developed an algorithm to find and label broken pieces of candy in a static image (labelDefectsRGBImage.m).
Phase 2: We tested the algorithm with live images from a web camera. We added some flexibility to the original algorithm with the thresholding tool submitted to MATLAB® Central by Robert Bemis (labelDefectsLiveImage.m).
Phase 3: We used GUIDE to develop a graphical user interface that let us more easily work with live images and run the algorithm on them (shape.m and shape.fig).
I would like to thank Brian Madsen and Inpakala Simon for the contributions to this demo.