Modelling the quality of harumanis mango (MA 128) grading using fuzzy logic system / Mohd Syahir Alfathul Amin Mohd Ripin

One of the factors that determining the market values of fruit is the quality. Most of the fruits for export markets are firstly sorted and graded according to the size and appearance of the fruits. The factors that caused the grading of fruits to varies because of lack of expenditure in purchasing the grading machines, human error and well trained personnel. This research focus on the surface area of mango fruits, size of defect and number of defect presence on the mango skin. This study will grade of the Harumanis mango into various categories of classes such as grade A, B or C. This results which in to compare the accuracy of grading Harumanis fruit between manually system and Fuzzy Logic. The result of this study shows that modeling using Fuzzy logic will help to identify and forecast the input data to produce classification result which is of high accuracy. Through this process, the value of each element of input and output can be identify. These valuable data was inserted into Fuzzy Inference System (FIS) through Adaptive Neuro-Fuzzy Inference System (ANFIS). This method was the most suitable for forecasting the model data. The result shows that the classification result from the system achieved 85% accuracy. The grading system using these method can be applied to grade accurately varieties of fruits for domestic and export market.