Multispectral spectroscopic analysis for classification

Document Type : Primary Research paper

Authors

1 Charotar University of Science and Technology, CHARUSAT Space Research and Technology Center, V T Patel Deptartment of Electronics & Communication Engineering Changa, Ta-Petlad, Anand, Gujarat 388421, India

2 Charotar University of Science and Technology, CHARUSAT Space Research and Technology Center, Chamos Matrusanstha Department of Mechnical Engineering, Changa, Ta-Petlad, Anand, Gujarat 388421, India

3 Charotar University of Science and Technology, Devang Patel Institute of Advance Technology and Research , Dept. of Information technology ,Changa, Ta-Petlad, Anand, Gujarat 388421, India

Abstract

In recent competitive scenarios, multispectral imaging systems are widely used in the area of information retrieval and object classification. This technology has made satellite applications versatile by providing multiband information in specific application sectors like material classification, agriculture industry, and food processing industry. With this experiment, we have made an attempt to address the material classification and access milk purity by acquiring data in Visible and Near Infrared bands between 400 to 1000nm. The experiments were carried out on samples of copper, aluminium, basalt fiber composite and milk. Using machine learning techniques the classification model has been developed. Results are discussed in paper.

Keywords