METODE GABUNGAN DWT-SVD UNTUK PENGUJIAN KETAHANAN IMAGE WATERMARKING
Daftar Isi:
- An intellectual property in this modern era is a form of digital data that has developed rapidly and made life easier. However, the impact of the dissemination of digital data into business land illegally and is very detrimental, namely by plagiarism and or theft of information. To protect the intellectual property of digital products, many methods have been developed to overcome the above problems. One method developed is Watermarking. This study proposes a combined method of Discrete Wavelet Transform (DWT) and Singular Value Decomposition (SVD) as a method of proposing to apply the quality of the Watermarking Technique. The DWT technique functions as the application of the embedding process to make it easier and SVD functions to authenticate the image based on the extracted watermarking correlation value. The testing of this research uses Lossy compression of 10%, Noise Salt n Pepper, Cropping, Blur Gaussian 0.5 and 1.0, and Sharpening Unmask 50,100, and 200 (%). The results for compression testing are PSNR of 46,0054 dB and MSE of 1,686. Then Noise Salt n Pepper with PSNR test results of 21.6963 and MSE of 441.1522. Then, testing using Cropping imagery produced PSNR of 13,423 dB and MSE of 3038.65. For Blur Gaussian 0.5 and 1.0, respectively, PSNR of 28,32431 dB and 24,3182 dB and MSE of 104,455 and 257.7 respectively. As well as the last test, Unmask Sharpening 50, 100, and 200 (%) each produced PSNR of 17,9052 dB, 17.83 dB, and 17.62 dB while MSE were 1098,815, 1124,503 and 1141.82, respectively. The test results show that the use of the combined method of Discrete Wavelet Transform (DWT) and Singular Value Decomposition (SVD) can overcome noise / interruptions in the image even though noise / interruption results in MSE values outside of the upper limit of 1.0 but the PSNR value remains in the range of 0 ~ 100 dB which has the highest MSE of 3038.65 with PSNR holds at 13,423 dB. Keywords: Discrete Wavelet Transform (DWT)-Singular Value Decomposition (SVD), Robustness, Noise, Watermarking