Automatic classification of wooden cabinet doors using computer vision

Abstract

This thesis describes the use of computer vision techniques for distinguishing wooden components in a manufacturing environment. The components considered here are kitchen cabinet doors, which are produced in many different styles and sizes, and travel on a conveyor at 30 feet per minute. An automatic classification system has been developed which can classify doors reliably. The system includes a host computer with video digitizer, two laser sources, and three video cameras to obtain profile images. This thesis describes the careful design of illumination and sensing geometry, the profile-based feature extraction process, and the classification method. The system exists as a laboratory prototype, and has been successfully tested with a large number of samples.