STEGANOGRAFI CITRA BERTINGKAT CASCADED IMAGE STEGANOGRAPHY

ERIK IMAN HERI UJIANTO, Agus Harjoko

2015 | Tesis | FMIPA

Information hiding is an important matter especially in protecting information from unauthorized party. The purpose of information hiding is to protect the secrecy in which certain mechanism is used in order to prevent the curiosity of other unauthorized party. One of the technique to hide information is steganography. This research proposes Cascaded Image Steganography (SCB - Steganography Citra Bertingkat) method based on spatial and transformation domains. This SCB method utilizes 2 steganography algorithms and 2 steps cover (double cover). This method consists of two main processes which are encoding and decoding. The encoding process consists of two phases which are Encoding level-1 and Encoding level-2. Similarly, the decoding process also consists of 2 two phases which are Decoding level-1 and Decoding level-2. One of the benchmarks to measure steganography method is its robustness. The robustness test is then performed against stego image that undergo images processing operations such as filtering, cropping, rotating, and compression. Several measurement parameters used in this SCB methods are the capacity level, fidelity level, recovery level, and imperceptibility level. The capacity level is measured by detecting the capacity of stego object as well as the stego image. During the encoding process, it is detected many characters can be hid, and then the remaining spaces in the cover image is counted. The fidelity level is measured using SNR to show the quality of the tested image. The SNR measurement of filtered stego image is 53.6361 dB which is considered good image quality. The SNR of stego image experiencing JPEG compression is 52.7693 dB which also considered good image quality. Both SNR measurements are indicating good level of fidelity. The imperceptibility level is measured using visual observation from respondents of survey. The imperceptibility level of stego image experiencing Median Filtering and JPEG compression are both good. The recovery level of extracted Label image in order from the best to the worst are Rotating, Median Filtering, JPEG Compression, and Cropping.

Kata Kunci : Cascaded Image Steganography, Image, Encoding, Decoding, Picture Quality