Detection of Water Fraud in Natural Lime Juice using Fuzzy Logic Table Look-Up Scheme

For the first time in case of fraud detection in Iran, the amount of water as a fraud in natural lime juice was detected using fuzzy modeling system in the present study. Four parameters consisting pH, EC (electrical conductivity), contrast and homogeneity (obtained from the samples’ images) were considered as input and the amount of water in natural lime juice was introduced as output of the fuzzy model system. Zero, 5, 10, 15, 25, 30 and 40% of water were used for system training and 7.5, 12.5, 20 and 35 % were used for testing the system. Mamdani’s fuzzy inference system, triangle membership functions and 7 rules were used for fuzzy modeling. The performance of the model was determined using R2 and RMSE statistical parameters. As a result, high value of R2(0.99) as well as low value of RMSE (1.67) demonstrated that fuzzy inference system can be used as an efficient method to estimate the amount of water fraud in natural lime juice.