Klasifikasi Hewan Berbasis Fitur Gray Level Co-Occurrence Matrix dengan Artificial Neural Network

Abstract

Dogs and cats are animals commonly treated as pets by many people. Humans have the ability to differentiate various things, and this ability when converted to a form of system is called Computer Vision. Computer Vision has many applications such as image processing that can be used for various things and one of the techniques in image processing is image classification. Image classification is a problem that aims to organize objects that are then observed into predefined categories. The approach utilized to construct the model involves employing an Artificial Neural Network (ANN) using Gray Level Co-occurrence Matrix (GLCM) as the method for extracting the features. The data used are images of cats and dogs which will be extracted using GLCM using various parameters that include distance and angles. The extracted feature will then be used to train a model and accuracy of each model will be measured to find the best parameter result. In this study, the best parameter that results in the best accuracy is 1 for the distance and 0°, 45°, 90° for the combination of angles resulting in 79% accuracy.

Keywords: Animal, Image, Gray Level Co-Occurrence Matrix, Classification, Artificial Neural Network