Classification of satellite image using maximum likelihood and ISODATA techniques

Chua, Xin Ying (2021) Classification of satellite image using maximum likelihood and ISODATA techniques. Project Report. Universiti Teknikal Malaysia Melaka, Melaka, Malaysia. (Submitted)

	Text (24 Pages) Classification of satellite image using maximum likelihood and ISODATA techniques.pdf - Submitted Version Download (342kB)
	Text (Full Text) Classification of satellite image using maximum likelihood and ISODATA techniques.pdf - Submitted Version Restricted to Repository staff only Download (2MB)

Abstract

Satellite Image Processing is important in Research and Development field. It is taken by the artificial satellite, and the photo taken is processed by computer to extract the data in the photo. Processing the image by using manual methods consumes a lot of time. Besides, the expert must know well about the area covered by the satellite image and the knowledge and familiarity of the expert will directly affect the efficiency and accuracy of the classification. Therefore, image classification by computer is introduced. There are a lot of techniques introduced by the previous researcher to classify the satellite image and the techniques are mainly divided into two types, which are Supervised and Unsupervised Classification. The example of supervised classification method is Maximum Likelihood and unsupervised classification method is ISODATA. The two techniques were used to identify the objects in the satellite image. The process of comparing results based on the two techniques were carried out to identify the best classification techniques. The comparison of the output was conducted based on the percentage of accuracy.

Actions (login required)

View Item