Feature Selection Using Artificial Bee Colony for Fruit Classification

Abstract

Extraction of features plays a very significant part in the creation of images. Different image pre-processing methods such as banalization, thresholding, resizing, normalization are added to the captured picture before receiving functionality. After that, techniques of extraction of features are applied to obtain features that will be useful in the classification and recognition of images. This paper aims to propose a fruit classification system based on an Intelligence algorithm, the proposed framework derives extracted features using merge harries corner detection with GLCM feature extraction. After that, the important features are selected using one of the methods of artificial intelligence, where the bee colony algorithm was used. After obtaining the selected features, two powerful classifiers are merged, which are the decision tree and naïve Bayes to classify the input fruit image. This work uses the Fruit 360 dataset. Where 70 percent of the data set was used during the training phase and 30 percent used during the test phase. It was observed that the proposed approach yielded good results and the accuracy was 96 %.