Knowledge-based inspection of electric lamp caps

Abstract It is vitally important to check the caps of electric lamps before they reach the market. A faulty, protruding filament wire would be extremely dangerous. Other possible faults include partial soldering and holes in the solder. This paper describes an automated visual inspection system which can detect all these faults. Most industrial vision systems are tailor-made to specific applications. However, a possible solution to the quest for a more adaptable system arises from the use of knowledge-based techniques—if such a system were to be adapted to a different application, only the knowledge base would have to be altered. A few rule-based systems do exist, but are generally not applicable to real-time operation. The highly efficient method described here can achieve on-line, real-time performance.

[1]  Anil K. Jain,et al.  A Rule Based Approach for Visual Pattern Inspection , 1988, IEEE Trans. Pattern Anal. Mach. Intell..

[2]  Dragutin Petkovic,et al.  A Rule-Based System for Verifying Engineering Specifications in Industrial Visual Inspection Applications , 1987, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[3]  K. Jambunathan,et al.  Development of an intelligent front end: An experience , 1991 .

[4]  Ashok K. Agrawala,et al.  Equivalence of Hough curve detection to template matching , 1977, Commun. ACM.

[5]  Ramesh C. Jain,et al.  Automatic Solder Joint Inspection , 1988, IEEE Trans. Pattern Anal. Mach. Intell..