![Sample our Computer Science journals, sign in here to start your access, latest two full volumes FREE to you for 14 days]({"type":"image" , "src" : "/sda/5928/TOC-Subject-Sample-2021-Computer-Science.jpg"})
ABSTRACT
Satellite images taken in hazy conditions tend to have low brightness and contrast, and to address this problem, a simple and optimised method using near-infrared (NIR) histogram equalisation has been proposed, in order to improve the brightness and contrast of the SPOT-5 images taken by satellite. This helps to improve the detection of shoreline with a higher efficiency in edge detection, and produces results with a high success rate of more than 90%. In order to remove haze from SPOT-5 images, a false colour composite is used to prevent misclassified pixels of the shoreline, and ensure minimal error in image classification. The presence of an NIR channel is suitable for the detection of vegetation and waterbodies. The NIR channel is important for this study, in ensuring that contrast between vegetated areas and sea can be clarified, due to the high reflectance of vegetation leaves in the NIR wavelength region. This research used three areas of interest to show different results in shoreline detection. The results showed that the proposed method performed better than existing methods, when dealing with the low brightness or contrast ratio of SPOT-5 images.
Acknowledgements
The authors would like to acknowledge the Malaysian Remote Sensing Agency which provided SPOT-5 data, and the Malaysian Government and Department of Survey and Mapping Malaysia (JUPEM) which provided topography and gazettes relating to the Tanjung Piai area. Tanjung Piai Johor National Park staff provided useful information about the study site. Thanks are also expressed to the supervisory team for helpful discussions during this research.
Disclosure statement
No potential conflict of interest was reported by the authors.