Path: Top

Segmentasi Multiresolusi pada Citra Natural dengan Representasi Scale Space Nonlinear.

Jurnal Teknologi dan Manajemen Informatika, Fakultas Teknologi Informasi, Universitas Merdeka Malang, volume 4, nomor 3, Desember 2006.
Journal from JIPTUNMERPP / 2008-01-09 09:44:55
Oleh : Chastine Fatichah, Nanik Suciati, Shofiya Syidada, Faculty_of_Information_Technology_-_Merdeka_Univer
Dibuat : 2006-12-01, dengan file

Keyword : Multiresolution segmentation, scale space liniar representation, image stack, scale space non liniar representation, region growing, hierarchical tree, total variation, partial differential.

Segmentation is an important part in the analysis of digital imaging. It segments the region on the basis of homogeneity criterian that is the gray-level. Scale space representation arranges image to become several images forming multi-resolution multiscale. This study implemented the method of multiresolution natural image segmentation using scale nonlinier representation. This method would result in multiscale image being separated on each level. Segmentation of region growing separated images into regions being connected to ISO level set. Image stack was yielded by doing nonlinier diffusion using approach to nonlinier partial differential equation that was total variation. Multi-resolution segmentation resulted in hierarchical tree image by connection segmentation resulted in hierarchical tree image by connection two image regions at adjecent level. On the basis of tryout result performance to several grayscale images in size of 128 x 128, it found that this method resulted in several different segmented model images on each level and maintained the edge longer using total varian. The method was dependent upon scale/iteration parameter and the appropriate selection of threshold value. The result segmentation on noised image showed that this method was able to maintain the edge on nolsed images.

Deskripsi Alternatif :

Segmentation is an important part in the analysis of digital imaging. It segments the region on the basis of homogeneity criterian that is the gray-level. Scale space representation arranges image to become several images forming multi-resolution multiscale. This study implemented the method of multiresolution natural image segmentation using scale nonlinier representation. This method would result in multiscale image being separated on each level. Segmentation of region growing separated images into regions being connected to ISO level set. Image stack was yielded by doing nonlinier diffusion using approach to nonlinier partial differential equation that was total variation. Multi-resolution segmentation resulted in hierarchical tree image by connection segmentation resulted in hierarchical tree image by connection two image regions at adjecent level. On the basis of tryout result performance to several grayscale images in size of 128 x 128, it found that this method resulted in several different segmented model images on each level and maintained the edge longer using total varian. The method was dependent upon scale/iteration parameter and the appropriate selection of threshold value. The result segmentation on noised image showed that this method was able to maintain the edge on nolsed images.

Beri Komentar ?#(0) | Bookmark

PropertiNilai Properti
ID PublisherJIPTUNMERPP
OrganisasiF
Nama KontakDra. Wiwik Supriyanti, SS
AlamatJl. Terusan Halimun 11 B
KotaMalang
DaerahJawa Timur
NegaraIndonesia
Telepon0341-563504
Fax0341-563504
E-mail Administratorperpus@unmer.ac.id
E-mail CKOwsupriyanti@yahoo.com

Print ...

Kontributor...

  • Editor: