High speed object recognition system

A high speed object recognition system is designed and developed for recognizing a set of handtools. The system is not only capable of handling the scaling and orientation problems, but also can recognize partially occluded objects. Corners and curves which are detected based on different order of discontinuities are selected as the primitive local features of an object. After feature detection, a back propagation network (BP net) is employed as a classifier. A set of general handtools is selected to evaluate the performance of the system and the results are encouraging. The recognition time is less than 6 seconds with an IBM AT computer