Fruit recognition and quality detection based on different classifiers: a case study of Apple

Abstract

The ability to identify the fruits based on quality in the food industry is very important now a days where every person has become health conscious. There are different types of apples available in the market. However, to identify best quality apples is cumbersome task. Therefore, the need arises to develop an algorithm that can detect the apples and classify it in order to increase its market value. Moreover, recognizing and classifying the both types of apples which are harvested and which are not harvested with same algorithm is not possible in one go. Therefore, the researchers proposed two different types of approaches for recognizing the apples in the images, one in which apples are harvested and other in which apples are not harvested. For the type of images in which apples are not harvested, one approach is used in which analysis is performed on the basis of color and shape. For the type of images, in which apples are not on the tree, another approach is used in which analysis is performed on the basis of edge and shape. After recognizing the apple in the respective image, quality is checked by different classifier (SVM and KNN). Proposed apple recognizing technique analyzes, identifies and classifies apple successfully up to 89.16% accuracy by SVM and up to 93.33% by KNN.