Rahayu, Nur Putri (2017) Segmentasi Pembuluh Darah Pada Citra Fundus Menggunakan Isotrophic Undecimate Wavelet Transform dan Fuzzy-Region Growing. Masters thesis, Institut Teknologi Sepuluh Nopember.
Preview |
Text
5115201018-Master-Theses.pdf - Published Version Download (1MB) | Preview |
Abstract
Segmentasi pembuluh darah pada citra retina menjadi hal yang substansial dalam dunia kedokteran, karena citra retina dapat digunakan untuk mendeteksi beberapa penyakit, seperti: diabetic retinopathy, hypertension, dan cardiovascular.
Dokter membutuhkan waktu yang lama untuk mendeteksi pembuluh darah retina terutama pembuluh darah tipis
karena diproses secara manual, sehingga diperlukan metode yang dapat membantu proses diagnosis agar lebih cepat.
Pada penelitian ini menggabungkan metode isotropic undecimated wavelet transform (IUWT) dan fuzzy-region growing (FRG) untuk segmentasi pada citra fundus. Metode IUWT digunakan sebagai segmentasi awal, sedangkan FRG digunakan untuk segmentasi akhir pada pembuluh darah tebal dan tipis. Input awal yaitu citra fundus yang diambil green channel selanjutnya diproses dengan IUWT dan FRG. Output hasil segmentasi IUWT dan FRG yaitu pembuluh darah tebal dan tipis.
Skenario uji coba menggunakan dua skenario yang pertama uji coba perbandingan dengan citra groundthruth, skenario kedua perbandingan hasil segmentasi IUWT dan fuzzy-region growing. Rata-rata metode IUWT dan FRG lebih tinggi yaitu 92 % daripada metode IUWT dan thresholding sebesar 83 %.
=================================================================
Segmentation of the blood vessels in the retinal image becomes substantially in medicine, because the retinal image can be used to detect several diseases, such as diabetic retinopathy, hypertension, and cardiovascular. The doctor take a long time to detect retinal blood vessels especially thin blood vessels due to be processed manually, so we need automatic method that can help the diagnostic process to make it faster.
In this research combines methods isotropic undecimated wavelet transform (IUWT) and fuzzy region growing (FRG) for segmenting the image of the fundus. IUWT used to obtain the first segmentation, then the FRG used to finish the segmentation in the thick and thin blood vessels. The initial input is green channel. The result of segmentation are thick and thin blood vessels.
There are two scenario: the first scenario is used to trial the groundthruth comparison with the image, second scenario is used to comparison IUWT segmentation and fuzzy-region growing with the previous method IUWT and thresholding. Accuracy results show a better average yield of 92% for FRG and 83% for thresholding. The IUWT and FRG methods have been shown to segment blood vessels thick and thin both visually and accurately.
| Item Type: | Thesis (Masters) |
|---|---|
| Uncontrolled Keywords: | Blood vessel segmentation in fundus, Isotropic Undecimated Wavelet Transform dan Fuzzy-Region Growing, Segmentasi Pembuluh darah Fundus |
| Subjects: | Q Science > QA Mathematics > QA248_Fuzzy Sets Z Bibliography. Library Science. Information Resources > ZA Information resources > Z699.5 Information storage and retrieval systems |
| Divisions: | Faculty of Information Technology > Informatics Engineering > 55101-(S2) Master Thesis |
| Depositing User: | Putri Nur Rahayu |
| Date Deposited: | 17 Oct 2017 01:51 |
| Last Modified: | 08 Mar 2019 01:57 |
| URI: | http://repository.its.ac.id/id/eprint/46104 |
Actions (login required)
 |
View Item |