The relationship between the appearance of pomegranate fruit and color and size of arils based on image processing

Pomegranate fruit are usually separated and graded manually. Also, color and size of the arils, are not measurable without removing the fruit peel. The objective of this study was to develop a method for grading pomegranate fruit based on color and size of arils using image processing and artificial intelligence. The physical characteristics of 200 fruit were measured and photographed, the fruit peels were cut, the arils photographed, and then categorized into three grades by an expert. The images were processed and modeled using three artificial intelligence algorithms. An Artificial Neural Network (ANN) Model with an accuracy of 98%, a correlation coefficient of 0.943 and a MSE of 0.008 was recognized as optimal. Accuracy of 95.5% and 75.5% was obtained for Adaptive Neuro Fuzzy Inference System (ANFIS) and Response Surface Methodology (RSM), respectively. The correlation coefficient and MSE of the ANFIS Model was 0.918 and 0.011, respectively while the RSM Model had a correlation coefficient of 0.622 and a MSE of 0.052.