Applications > Previous Apps > Flaw Dection

Flaw Detection

This application required that the machine vision system monitor the size as well as the uniform appearance of a mark made as the result of a polishing process. Variations in the dimensions of the mark or the appearance of dark or light spots within the mark were indicative of abnormal process variations and/or contaminants. The occurrence of these problems typically indicated that the part had not been properly polished and should be rejected for rework.

Cycle time for the application was relatively slow and the computational requirements were light. As a result, substantial cost savings were achieved in the implementation of the machine vision system by using one vision system to monitor five separate cameras / work areas.

Successfull completion of this project required the development of custom image processing algortihms since standard algorithms were unable to detect the subtle variations in the burnishing mark.