Rekontruksi Citra Panorama Berbasis Data Geospasial Menggunakan Sift (Scale Invariant Feature Transform)

Hermansyah, Adi (2019) Rekontruksi Citra Panorama Berbasis Data Geospasial Menggunakan Sift (Scale Invariant Feature Transform). Masters thesis, Institut teknologi sepuluh november surabaya.

Text 07111650050201_Master_Thesis.pdf Restricted to Repository staff only Download (6MB)

Abstract

Rekontruksi merupakan pengembalian sesuatu ketempat semula, Penyusunan kembali dari bahan-bahan yang ada dan disusun kembali sebagaimana adanya atau kejadian semula. Dimana pada penelitian ini menggunakan metode SIFT, match dan homography. Berdasarkan dari hasil eksperiment ini, sudah mampu merekontruksi citra panorama. Setelah dilakukan proses stitching pada data geospasial ditemukan lah sebuah masalah dimana data GPS nya hilang dan hasilnya masih terdapat overlap. Dimana data awal memiliki data GPS. Dari hasil eksperiment ini, sudah mampu merekontruksi citra panorama berbasis data geospasial menggunakan SIFT (Scale Invariant Feature Transform). Di ketahui bahwa dari hasil penelitian ini sudah bisa mengetahui titik koordinat x dan y pada pada citra awal terhadap citra hasil stitching sehingga dapat diketahui posisi citra awal terhadap image hasil stitching. Untuk itu dilakukan match antara image awal dengan image hasil stitching sehingga didapatlah hasil dari match SIFT dengan keypoint pada match image dengan image hasil stitching yaitu image A = 7 , image B = 16, image C = 27 dan image D = 32, Dengan begitu dapat disimpulkan hasil matching tersebut telah mendapatkan kesamaan keypoint dengan total keypoint sebanyak 82, rata-rata keypoint pada saat match 20,5 dan rata-rata waktu running program saat match 0,0385 second. Dengan diketahui posisi koordinat image awal terhadap image hasil stitching maka diketahui juga posisi longitude dan latitude pada image hasil stitching. Dengan hasil citra A (7,16,51.2675 S / 112,47,32.1102 E), citra B (7,16,51.4336 S/ 112,47,32.2728 E), citra C (7,16,51.6024 S / 112,47,32.3975 E) dan citra D (7,16,51.8191 S / 112,47,32.5617 E). Titik koordinat GPS pada citra A, dimana x,y = (205, 164) dan x ’, y’ = (603,15), pada citra B dimana x,y = (194, 175) dan x’, y’ = (598,70), citra C dimana x,y = (196, 110) dan x’, y’ = (599,71) dan citra D dimana x,y = (28, 110) dan x’, y’ = (607,130).
===================================================================================================================================
Reconstruction can mean the return of something in its place of origin, rearrangement of existing material and rearranged as it is or the original event. Where in this study used the SIFT, match and homography method. Based on the results of this experiment, it was able to reconstruct panoramic images. After the stitching process on geospatial data found was a problem where the GPS data is lost and the results are still there overlap. Where the initial data has GPS data. From the experimental results, we have reconstructed the panoramic data based on geospatial data using SIFT (Scale Invariant Feature Transform) method, in which the keypoint results obtained by matching the original image and stitched image. It can be seen from key point results from four images namely image A = 7, image B = 16, image C = 27 and image D = 32. By summing the key points from those images, the total number of keypoints are 82. The average of keypoints will be 20.5 and the average of run time of the program is 0.0385 second. By knowing the position of initial image coordinate on the stitched image, we know where the position of longitude and latitude is. The stitched image from image A located in (70,16’,51.2675’’S/1120,47’,32.1102’’E), image B located in (70,16’,51.4336’’S/1120,47’,32.2728’’E), image C located in (70,16’,51.6024’’S/1120,47’,32.3975’’E) and image D located in (70,16’,51.8191’’S/1120,47’,32.5617’’E). Meanwhile, GPS coordinates point in image A lies in x,y=(205,164) and x',y'=(603,15), in image B lies in x,y=(194,175) and x',y'=(598,70), image C located in x,y=(196,110) and x',y'=(599,71) and image D located in x,y=(28,110) and x',y'=(607,130).

Item Type:	Thesis (Masters)
Uncontrolled Keywords:	geospatial data, homography, image, panoramic, reconstruction, SIFT
Subjects:	T Technology > T Technology (General) > T57.5 Data Processing
Divisions:	Faculty of Electrical Technology > Electrical Engineering > 20101-(S2) Master Thesis
Depositing User:	Adi Hermansyah
Date Deposited:	23 Jul 2026 08:06
Last Modified:	23 Jul 2026 08:06
URI:	http://repository.its.ac.id/id/eprint/65102

Actions (login required)

View Item