PENGOLAHAN CITRA DIGITAL DAN LOGIKA FUZZY DALAM IDENTIFIKASI TINGKAT KEMATANGAN BUAH

Abstract

This study aims to implement digital image processing and fuzzy logic in identifying the maturity level of red dragon fruit and describing the level of accuracy. The process carried out is to change the type of red dragon fruit image from the red green blue (RGB) type to the grayscale type. Furthermore, extraction is carried out using MATLAB to obtain information from the image. Information from the image is contrast, correlation, energy, homogeneity, mean, variance, entropy. There are 7 information used as fuzzy input. The input fuzzy model uses triangular and trapezoidal membership functions to construct fuzzy rules on 28 data, so there are 28 fuzzy rules. After the fuzzy rules are obtained, then the inference and defuzzification processes are carried out. The inference used is the mamdani method. The result of defuzzification is the value for the maturity level of red dragon fruit which is divided into four categories, namely raw, half-ripe, ripe, rotten. The fuzzy model that has been built will test the model by determining the level of accuracy and error of the model. With the results of 96.42% with an error of 3.57%.