Digital Watermarks for Videos Based on a Locality-Sensitive Hashing Algorithm

Abstract

Sensitive information in images is leaked during attacks, resulting in the malicious acquisition of personal privacy. To improve the robustness of attacking defence for video images, a digital watermarking algorithm based on locality-sensitive hashing (LSH) is designed. The video signal was decomposed using a one-dimensional wavelet transform. According to the Yeung Mintzer (Y-M) algorithm, a marker watermark W1 was embedded in the low-frequency subband to identify image tampering. The data string of hash function values and the exclusive OR (XOR) result of identification watermark W2 were embedded into the HH high-frequency subband, which was used to identify and counter the pseudo-authentication attacks such as collage and Vector Quantization(VQ). The singular value decomposition (SVD) algorithm was used to decompose the hashmapped watermark and adaptively adjust the embedding strength of the watermark. The position-sensitive hash algorithm proposed has good invisibility for embedding digital watermarks into images, with an average accuracy of approximately 97% for feature matching of digital images. The PSNR value of the image embedded with the watermark is approximately 49 dB. At the 50th minute of the experiment, the regulatory factor value of the research method was 0.3. Under different attack modes, the correlation coefficient between the watermark extracted by this method and the original watermark image is greater than 0.85. Due to the low compression quality of JPGE, the correlation coefficient between the watermark and the initial watermark is greater than 0.6, and its error rate is less than 0.10bit.