Author : Saeed AL-Mansoori and Fatima AL-Marzouqi

The coastline is a vitally important characteristic of the environment and effects many different aspects of civilization. Globally it becomes extremely essential to monitor the coastline zones, as it reflects on economic development, environmental protection and infrastructure of the country. The coastline indicator was defined as the segment of the earth where the sea or ocean meets the land. In this paper, automatic technique is proposed to extract the coastline from satellite images. The proposed procedure is based on classifying the land and water components by applying a spectral band ratios method with two normalized difference indices; Normalized Difference Vegetation Index (NDVI) and Normalized Difference Water Index (NDWI), and then extracts the coastline as a vector data by converting the classified image into a binary image format followed by employing the Sobel edge operator. The accuracy of the results is assessed in comparison with the ground truth observations. This technique is fitting through a case study over the coastline of Dubai to monitor the changes in the coastline covering a time span of six years from 2009 till 2015. The results revealed an increase in the Dubai coastline by 6%.

Keywords: Coastline extraction, DubaiSat-1, DubaiSat-2, Band Ratios, Normalized Difference Vegetation Index (NDVI), Normalized Difference Water Index (NDWI), Ground Sampling Distance (GSD).

Article published in International Journal of Current Engineering and Technology, Vol.6, No.4 (Aug-2016)