<![CDATA[Based on previous observations, a series of steps using digital image processing methods is proposed for the automatic detection and counting of islands to avoid inaccuracies from satellite imagery by leveraging morphological properties of object. The need for accurate spatial data regarding the number of islands in Indonesia is crucial for various developmental purposes. Many small islands known to have beautiful landscapes remain unaccounted for due to the vast territorial waters of the country, posing challenges to manual evaluation of the numbers and distributions. Remote sensing methods offer a viable solution for efficiently counting and inventorying islands. Therefore, this study aimed to detect islands in Kepulauan Seribu, located north of DKI Jakarta, through the thresholding-based segmentation process and count the total number using morphological information. The methodology applied was Object-Based Image Analysis (OBIA), including image gray-scaling, thresholding, morphological operations, connected component labeling, and region-based object counting. The results obtained showed 111 islands, compared to direct observation of image from which 104 were found, with detection accuracy of 93.27%. The discovery not only contributes valuable insights into the specific region but also serves as a basis for potentially applying digital image processing methods on a larger scale to recalculate the number of Indonesian islands. Such recalculations could play a crucial role in informing and guiding future developmental initiatives. ]]>
Copyrights © 2024
