Diversity of Plum Stones Based on Image Texture Parameters and Machine Learning Algorithms

The objective of this study was to evaluate the usefulness machine learning based on image texture parameters discriminate plum stone cultivars. plums cultivars ‘Emper’, ‘Kalipso’, and ‘Polinka’ were sampled. For each cultivar, one hundred images stones acquired using a digital camera. Processing included conversion individual color channels, segmentation, region interest (ROI) determination, parameter extraction. Then, discriminant analysis, including selection building discriminative models for evaluation diversity cultivars, carried out. obtained results discrimination very accurate confirmed effectiveness processing cultivar diversity. most satisfactory results, reaching 96.67% average accuracy three (97% 96% ‘Polinka’), built combined textures selected from all channels IBk classifier. developed procedure can be practical importance correct identification avoiding their mixing preserve uniformity.