Banana Maturity Classification Using Hybrid Features On Various Classifiers

Abstract

Fruits are grown based on the climatic conditions and its requirement. Ripening is an important phase in horticulture and a fruit is categorized as either climacteric or non-climacteric. Climacteric fruit ripens even after harvesting whereas non-climacteric does not ripen off the tree. Therefore at a particular unripe stage all the fruits are harvested. This paper describes the banana ripening process and explains how image processing is utilized for banana ripeness classification using Histogram of Oriented Gradients (HOG), zernike and resnet features.To bring an automated ripeness recognition system in image processing Naïve Bayes (NB), K-Nearest Neighbor (KNN), Decision Tree and Random Forest (RF) classifiers are analyzed. Experimental results prove that the proposed hybrid features using RF classifier gives most accurate results.