Jurnal Rekayasa Vol 2 No 2 Artikel 8
Poegoeh Joedhiawan
(Staff Pengajar Jurusan Elektro, Program Studi Teknik, Universitas Jember)
ABSTRACT
In certain applications data and information ment to be understood and accessed by certain people only. Hiding information in image has function in protecting image from profit making of other, such as ownership claimed by other. This can be serious problem for owner. To solve this problem, it is important to hide information by using labeling technique. Digital image labeling process proposed here uses wavelet method and human visual system. This method uses wavelet transformation, error correcting code and exploitation weakness of human system. Embedding label process will be carried on quantized C1 approximation image by regulating transparency and robustness. The image in this method is similar to authentic image, using label extraction to get back embedded label.
The result is invisible label to the naked eye. Experiments denoted the label images have PSNR between 31 dB to 36 dB and label extractions from attacked label image show that the label still clearly seen with average label value similarity at 89,9%.
Keyword : labelling, discrete wavelet transform, error correcting code, human visual system, robustness.
DOWNLOAD NASKAH LENGKAP:
ZIDDU:
SEMOGA BERMANFAAT