Rahmad, Cahya (2005) Fair Share Amount Pada Kuantisasi Vektor Untuk Kompresi Citra Berwarna. Masters thesis, Institut Teknologi Sepuluh Nopember.
|
Text
5102201009-Master_Thesis.pdf Restricted to Repository staff only Download (4MB) |
Abstract
Citra digital merupakan sekumpulan titik gambar atau piksel yang terletak pada koordinat f(x,y). Dalam bentuk aslinya, citra memerlukan kapasitas memori yang sangat besar, sehingga membutuhkan media penyimpanan yang besar dan lebar pita yang besar dalam proses transmisi. Oleh sebab itu, diperlukan proses kompresi citra. Untuk mengompresi data dengan teknik kuantisasi vektor, diperlukan sebuah codebook yang dihasilkan dari vektor-vektor yang dibangkitkan dari data asal. Metode Fair Share Amount merupakan metode untuk menghasilkan codebook dengan cara membagi vektor-vektor yang ada ke dalam berbagai area, kemudian dipilih vektor yang menjadi wakil dari setiap area tersebut. Pada penelitian ini dilakukan kuantisasi vektor pada citra berwarna dengan metode Fair Share Amount serta dilakukan perbaikan pada sebagian langkah algoritmanya. Pengujian pada berbagai citra berwarna menggunakan perangkat lunak MATLAB meliputi pengujian CR (Compression Ratio), e-RMS (Root Mean Square Error), dan SNR (Signal-to-Noise Ratio), menunjukkan bahwa metode Fair Share Amount dapat mengompresi citra berwarna dengan hasil yang baik. Dari pengujian menggunakan algoritma yang diperbaiki (Expanded Fair Share Amount) diperoleh kinerja yang lebih baik daripada metode Fair Share Amount
==================================================================================================================================
A digital image is a collection of pixels located at coordinates f(x,y). In its original form, an image requires a large memory capacity, resulting in the need for large storage media and wide bandwidth for the transmission process. Therefore, an image compression process is required. To compress data using vector quantization, a codebook is needed, which is generated from vectors derived from the original data. The Fair Share Amount method is a method for generating a codebook by dividing existing vectors into several areas and then selecting a vector to represent each area. In this research, vector quantization is performed on color images using the Fair Share Amount method, along with improvements to several steps of the algorithm. Testing of various color images using MATLAB software includes CR (Compression Ratio), e-RMS (Root Mean Square Error), and SNR (Signal-to-Noise Ratio) measurements. The results show that the Fair Share Amount method is able to compress color images with good results. Testing using the improved algorithm (Expanded Fair Share Amount) shows better performance than the original Fair Share Amount method.
| Item Type: | Thesis (Masters) |
|---|---|
| Additional Information: | RTIf 006.6 Rah f |
| Uncontrolled Keywords: | kompresi citra, fair share amount, Expanded fair share amount, generate codebook, e-RMS, SNR, : Image Compression, Fair Share Amount, Expanded fair Share Amount |
| Subjects: | Q Science > QA Mathematics > QA76.9.D33 Data compression (Computer science) |
| Divisions: | Faculty of Industrial Technology > Informatics Engineering > 55101-(S2) Master Thesis |
| Depositing User: | magang . |
| Date Deposited: | 17 Sep 2026 04:38 |
| Last Modified: | 17 Sep 2026 04:38 |
| URI: | http://repository.its.ac.id/id/eprint/144639 |
Actions (login required)
![]() |
View Item |
