Al Musayyab, Said (2016) Forensik Digital Deteksi Pemalsuan Copy-Move Citra Dengan Menggunakan Metode Block Matching. Undergraduate thesis, Institut Teknologi Sepuluh Nopember.
Preview |
Text
5112100177-Undergraduate_Thesis.pdf - Accepted Version Download (2MB) | Preview |
Abstract
Semakin berkembangnya teknologi manipulasi citra, semakin
mudah seseorang untuk melakukan pemalsuan citra.
Pemalsuan citra dapat menimbulkan kesalahpahaman bagi
seseorang yang melihat citra tersebut. Salah satu jenis
pemalsuan citra yaitu copy-move, di mana bagian pada citra
disalin, kemudian ditempelkan pada bagian lain pada citra
yang sama. Tugas akhir ini akan mengimplementasikan deteksi
pemalsuan copy-move dengan menggunakan metode block
matching. Metode block matching dibagi menjadi dua
pendekatan, yaitu exact match dan robust match. Kedua metode
akan mengambil blok ke seluruh bagian citra, kemudian
membandingkan blok-blok yang ditemui. Exact match
mengambil blok melalui representasi piksel, sedangkan robust
match mengambil blok melalui nilai DCT (Discrete cosine
transform). Blok yang sama akan dideteksi sebagai serangan
copy-move. Metode robust match dapat menghasilkan kualitas
yang lebih baik dibandingkan dengan exact match. Dari sampel
data set yang digunakan, robust match menghasilkan kualitas
rata-rata 75% dengan hasil terbaik sebesar 97%.
==================================================================================================================
In recent years, development of image editing makes
people easier to manipulate image. Image manipulation can
raise misleading to someone looking at the image. One example
of image forgery is copy-move, where some region of image
copied and pasted on other position in the same image. This
final project will implement copy-move forgery detection using
block matching method. Block matching method devided into
two approaches, exact match and robust match. Block matching
method will generate overlapping blocks to all regions of
image. Then it will compare the blocks, and identic blocks will
detected as copy-move area. Exact match will use pixel
representation and robust match will use DCT (Discrite Cosine
Transform) for generating blocks. From the experiments,
robust match produce average quality 75% with the best up to
95% which is better than exact match.
| Item Type: | Thesis (Undergraduate) |
|---|---|
| Additional Information: | RSIf 006.42 Alm f |
| Uncontrolled Keywords: | Forensik digital; copy-move; citra |
| Subjects: | T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK7882.P3 Pattern recognition systems |
| Divisions: | Faculty of Information Technology > Informatics Engineering > 55201-(S1) Undergraduate Thesis |
| Depositing User: | ansi aflacha |
| Date Deposited: | 23 Apr 2020 04:51 |
| Last Modified: | 23 Apr 2020 04:51 |
| URI: | http://repository.its.ac.id/id/eprint/75892 |
Actions (login required)
 |
View Item |