Insertion of watermark using a novel hybrid algorithm for secure multimedia digital evidence

Insertion of watermark using a novel hybrid algorithm for secure multimedia digital evidence

연구 분야: Analysis

학회: Progress in Artificial Intelligence

Watermarking digital images has become a popular and effective way to safeguard digital image content from fraudulent alterations and to protect copyright laws over the past few years. Digital content with restricted distribution can be well protected by watermarking, and following decryption, authorized parties are fully permitted to handle sensitive information. Digital watermarking faces challenges in sharing data while protecting content ownership and copyright. Data-concealing method embeds message into digital information to prove ownership. Therefore, a novel hybrid algorithm is proposed for watermarking the secret key in the digital evidence. In this designed model, the digital evidence which involves audio, image, video and text are watermarked with the novel hybrid algorithm as discrete wavelet transform and discrete Tchebichef transform. The hybrid algorithm has two parts, one divides the digital evidence into two lower rank and upper rank components and the other one will scramble the watermark image. Then the entire data is decomposed into separate sub-bands in order to provide multiple resolution representation such as LL, LH, HL and HH. The LL sub-band as well as the coefficient belongs to the low frequency domain. Hence the multimedia watermark will be inserted in the low frequency domain. The extraction of the watermark is performed by using the median filer and the inverse of the hybrid algorithm. An evaluation and comparison is conducted for the proposed and the existing approaches. The performance metrics includes PSNR, NC, and SSIM of the proposed model are 22.606 dB, 0.97 and 0.79. Thus, the designed model operates more effectively than existing methods.