Kompresi Citra Menggunakan Metode Fraktal

Arifin, Muchammad and , Diah Priyawati, S.T., M.Eng. (2018) Kompresi Citra Menggunakan Metode Fraktal. Skripsi thesis, Universitas Muhammadiyah Surakarta.

[img] PDF (Surat Pernyataan Publikasi)
Surat Pernyataan.pdf
Restricted to Repository staff only

Download (349kB) | Request a copy
[img] PDF (Naskah Publikasi)

Download (1MB)


At this time the need for information technology so rapidly. The existence of image compression is needed in the world of digital photos. This study aims to perform image compression with fractal method to reduce storage capacity with good image quality. Fractal compression is a compression that resembles the image itself as a whole. The method of quadtree decomposition is used to determine the block size varies according to the image feature and is separated into several different blocks. Self-similarity is done to find the resemblance of the whole block. After compression, huffman coding is applied to the compressed image so that its performance will be better. In this case different quantity sizes are found by skipping images with different dimensions and formats, and the image format is good for compression. So this study uses the calculation of MSE and PSNR to determine the success of fractal compression with other compression methods. This study was also compared with the DCT compression method. Fractal compression method yields better value that is the value for the mean of MSE is 11,07 dB and the mean value of PSNR is 38,48 dB with the average value of compression ratio is 95.05%.

Item Type: Karya ilmiah (Skripsi)
Uncontrolled Keywords: Fractal, huffman, MSE, PSNR, quadtree decomposition, ratio
Subjects: Q Science > Q Science (General)
Q Science > QA Mathematics > QA75 Electronic computers. Computer science
Q Science > QA Mathematics > QA76 Computer software
Divisions: Fakultas Ilmu Komunikasi dan Informatika > Teknik Informatika
Date Deposited: 25 May 2018 07:25
Last Modified: 25 May 2018 07:25
URI: http://eprints.ums.ac.id/id/eprint/62670

Actions (login required)

View Item View Item