Pengenalan Citra Logo Kendaraan Menggunakan Metode Gray Level Co-Occurence Matrix (Glcm) dan Jst-Backpropagation

Abstract

A car is a vehicle that has a varied shape or model but the difference is the brand or logo. Vehicle logos have their own meaning and meaning for car industry companies. The logo should have a practical and effective or efficient function so that the logo form is part of the marketing and branding program of the car industry company [1]. There are three types of car logos that are now known, in the form of symbols, text, or a combination between the two. The logo is always in the front and back of the car body and usually has a lighter color than the color of the vehicle. One that supports the development of technology is how to recognize a vehicle either from the brand, shape, model and color of the vehicle. Some references that are deemed feasible to help this research include utilizing the weaknesses and weaknesses of the results of previous research, including a paper entitled. Scale Invariant Feature Transform (SIFT) [2]. SIFT is combined with Logistic Regression [3] based on Gradient Orientation Histogram (HOG). Logo Recognition Using Probabilistic Neural Networks [4]. Therefore, the researchers wanted to focus on the logo recognition using the extraction of the Gray Level Co-occurrence Matrix (GLCM) feature. Testing and training testing using ANN-Backpropagation. From the results of this study the best accuracy obtained 95.7%, so that GLCM and ANN-Backpropagation can recognize the image of the vehicle logo. DOI : 10.29408/jit.v1i1.894