A LIGHTWEIGHT CNN ARCHITECTURE FOR LANDCLASSIFICATION ON SATELLITE IMAGES
Abstract
Abstract: Image matching is a fundamental method for gathering 1 ground control points (GCPs) by establishing a link betweenincoming images and reference image map chips. It is a required step in the automatic precise geo-registration of satellite images.To increase georeferencing accuracy, reference chips on the photographs must be properly aligned. With a limited supply of chips,the need for a higher matching success rate rises. This study compares incoming satellite photographs to reference chips derivedfrom aerial colour ortho-images. It is difficult to match the two datasets since they have different spectral responses and textures. Toimprove the matching success rate, we employ pan-sharpened satellite pictures. The results revealed a greater matching success ratewith pansharpened images due to the similar spectral spectrum and enhanced spatial resolution. As a consequence, pansharpenedpictures may help to improve picture matching success rates in automatic accurate georeferencing of high-resolution satellitephotographs
Full Text PDF
IMPORTANT DATES
Submit paper at ijasret@gmail.com
Paper Submission Open For |
October 2024 |
UGC indexed in (Old UGC) |
2017 |
Last date for paper submission |
30th October, 2024 |
Deadline |
Submit Paper any time |
Publication of Paper |
Within 15-30 Days after completing all the formalities |
Publication Fees |
Rs.6000 (UG student) |
Publication Fees |
Rs.8000 (PG student)
|
|