Registrasi Permukaan Tiga Dimensi Berbasis Angular Invariant Modified

Koeshardianto, Meidya (2010) Registrasi Permukaan Tiga Dimensi Berbasis Angular Invariant Modified. Masters thesis, Institut Teknologi Sepuluh Nopember.

Text 2208205203-Master_Thesis.pdf Restricted to Repository staff only Download (9MB) | Request a copy

Abstract

Registrasi citra menjadi langkah yang sangat penting dalam semua tugas analisis gambar dimana informasi yang diperoleh dari berbagai sumber data seperti penggabungan citra, deteksi perubahan serta perbaikan citra. Bahkan dalam perkembangannya, registrasi citra juga dapat digunakan sebagai identifikasi obyek dengan obyek yang lain. Dalam penelitian ini membahas tentang registrasi citra tiga dimensi menggunakan fitur angular invariant. Registrasi dapat dilakukan pada obyek sebelum terdeformasi. Akan tetapi fitur ini kurang efektif pada obyek terdeformasi. Selanjutnya, fitur angular invariant di modifikasi dengan menambahkan fitur koefisien Fourier. Hasilnya fitur angular invariant modified lebih efektif dibandingkan dengan angular invariant tanpa koefisien Fourier. Beberapa kendala dihadapi dalam proses registrasi terutama pada obyek homogen seperti tabung dan bola. Fitur ini kurang dapat memberikan ciri khas pada obyek — obyek homogen. Maksimum deformasi yang dilakukan sebesar 0,4 pada obyek beetle dan 0,25 pada obyek holes.
=======================================================================================================================================
Image registration become step which of vital importance in all analysis duties draw where obtained information from various data source like merger of image, detect change and also repair of image. Even in growth, image registration also can be used as to identify object with other object. In this research study concerning 3D image registration use angular invariant feature. Registration can be conducted for image before deformation. However this feature is less effective for deformation image. Hereinafter, angular invariant feature modification by added Fourier coefficient feature. The result of feature more compared to effective angular invariant modified of angular invariant without Fourier coefficient. Some constraints faced in registration especially for homogeneous images like ball and tube. This feature give less individuality for homogeneous images. Maximum conducted deformation equal to 0,4 for beetle image and 0,25 for holes image.

Item Type:	Thesis (Masters)
Additional Information:	RTE 006.6 Koe r 2010 3100010040525 WEEDING
Uncontrolled Keywords:	Angular invariant, Registration, Feature, Fourier
Subjects:	T Technology > TA Engineering (General). Civil engineering (General) > TA1637 Image processing--Digital techniques. Image analysis--Data processing.
Divisions:	Faculty of Industrial Technology > Electrical Engineering > 20101-(S2) Master Thesis
Depositing User:	Anis Wulandari
Date Deposited:	26 Aug 2026 05:04
Last Modified:	26 Aug 2026 05:04
URI:	http://repository.its.ac.id/id/eprint/133442

Actions (login required)

View Item