Formulation and Evaluation of Coating Content and Their Distribution on Chips Products by Automated Inspection Using Fuzzy-Based Image Processing

This study proposed a method to develop an intelligent system using an adaptive neuro-fuzzy inference system (ANFIS)-based image processing which predicts the flavor distribution of the chips ring. This study was divided into three stages and different variables in each stage to obtain the best combination. These variables were the number of nozzles employed (one, two and three nozzles), flavor solution temperature (4, 50 and 60C) and the flavor spraying air flow rate (2.5, 3.0 and 3.5 psi). The results showed that the chips rings produced by the combinations of the three nozzles, 50C and with different air flowing rates exhibited significantly higher color distribution, flavor and overall acceptance compared with all other treatments (P < 0.05). These results were supported by the flavor distribution images that were taken for each combination. The ANFIS predicted the experimental results with 100% accuracy. ANFIS processing indicated that the best treatment was the one that was produced by using three nozzles for spraying the flavor under 50C with 2.5 psi air flow rate. Practical Applications This system was developed by using an adaptive neuro-fuzzy inference system (ANFIS)- based image processing to rapidly predict the flavor distribution of the chips ring for accepting or rejecting defects.