Paper Title
Light Spectrum Speckle Analysis in Roughness Material Identification by Using Naïve Bayes Classifier Based GLCM Method

Abstract
Speckle imaging is a technique used in various sectors of human life. The process can be used to analyze object roughness. Speckle imaging use light and observes speckle patterns that formed from light interference on surface plane. The speckle image technique is very safe and does not require any contact which is easy to detect roughness of any object. Two sandpapers were used as objects in this study. The image processing in this study began with pre-processing method, image segmentation, feature extraction and then classification process. The feature extraction process uses GLCM method. The classification process is using Naïve Bayes classifier method. Based on experimental result, it was found that visible light spectrum affects image feature which changed for the GLCM method. The feature extraction its pixel mean and its pixel contract which is increasing due to decrease of wavelength spectrum. The accuracy of the naïve bayes classifier increases if the wavelength shorter. Pixel differences marked by feature extraction in the identification of object roughness due to differences in wavelengths affect naïve bayes classification which are recommended for lower spectrum wavelengths. Keywords - Grey Level, Image Processing, Light Spectrum and Roughness