A hierarchical approach to recognition of finished wooden components

This paper describes a system that can automatically identify textured objects, such as stained wooden parts, using color images. The system integrates texture and color information within a two-stage classification system to assign a class label. The intended application is the identification of wooden cabinet components as they travel on a conveyor in an industrial setting. The system presented has been successfully tested in the laboratory, even though the objects of interest exhibit large variations in texture and color.