Answers are grounded in this paper and cite their sources.
Computerization has an essential role in various fields, one of which is remote sensing. His ability to process data quickly and accurately and receive and store data produces information based on a given program. Edge detection techniques in remote sensing imagery have developed rapidly in line with advances in computer technology. This technique is essential, one of which is used for preliminary surveys of coastline changes. A well-detected land will assist the surveyors in the field study. This study aims to determine three edge detection methods' performance, including Roberts' Filter, segmentation, and band ratio on Landsat 8 imagery in producing land edges directly adjacent to the sea area. The use of a 2x2 matrix on the Roberts' filter, the range of values from 0.52 to 0.56 in the segmentation method, and the green band ratio - NIR Infrared resulted in varying accuracy. The results of this study show that the non-fusion image, the resulting edge is less sharp and coarse than the fusion image in the Roberts' filter method. The best method is produced by Band Ratio, where the edge detection results are close to the real image conditions. Meanwhile, unfavorable results were produced in the segmentation method. So that the band ratio has reliable performance for the detection of land objects bordering the ocean.
No takes yet. Share an insight, caveat, or question.
Luhur Moekti Prayogo (2021) studied this question.
Synapse has enriched 7 related papers. Consider them for comparative context: