Automated Visual Inspection for Bottle Caps Using Fuzzy Logic

Automated Visual Inspection System (AVIS) has the capability to investigate large numbers of manufactured goods quickly and accurately. In addition, this system operates with a high level of reliability and consistency in their tasks. This study proposed an AVIS for detecting cap situations by using fuzzy logic classifiers. The objectives of this research are to develop an applicable image processing algorithm, together with a feature extraction technique, and to detect the cap for plastic bottles which is based on the average of distances. Three types of classification were compared for detecting the bottle caps. They are Mamdani, Sugeno, and production rule. The system was evaluated in a real time environment. The results are 97.91%, 97.5%, 96.66% accuracy for Mamdani, Sugeno, and production rule respectively.