SELF EMBEDDING FINGERPRINTWATERMARKING AND STEGANOGRAPHY USING LSB,RSA, 3-DWT AND MLPSVM ALGORITHM
Mili Singh, Jitendra Sheetlani · IJITR International Journal of Innovative Technology and Research - IJITR International Journal of Innovative Technology and Research · 2016
Digital watermarking methods are used to biometric document protect from attacks of either intentional and accident. These attacks are intended to either circumvent the afforded security throughsystem or to normal functioning deter of the system. In this research,verification of fingerprint methods with watermarking technology combination to provide copyright protection and authentication of digital images is proposed. In proposed algorithm for higher performancepresent self embedding fingerprint watermarking and steganography using least significant bit (LSB), RSA, three level discrete wavelet transform (DWT) and multilayer perceptron support vector machine (MLPSVM) classification. In this study, we take cover image and create self watermark images using bi-cubic interpolation (BCI) method for memory saving.The proposed approach gives perfect results from previous work. The experimental database contains 3 person fingerprint images with 8 different images. In this study, we calculate false matching ratio (FMR), false rejection ratio (FRR), accuracy, time and peak signal noise ratio (PSNR). The experimental outcomes for FMR and FRR is 0.5833and 0.7500. PSNR value is varying from 70-95%. Accuracylies between 0 and 3. Also, we applied attacks on watermarked image (like noise and rotate). Also, calculate time of each image and it lies between 65-76%.