Robust digital image watermarking using cuckoo search optimization and probabilistic neural network.

In this new age, because of the exceptional achievement of the global computer network, the trading of digital media over the web has turned out to be unbelievably easy. However, securing the exclusive rights of the owner while trading digital media is an important issue. Digital image watermarking is a method to ensure copyright protection, security, and authenticity of data. In this paper, a novel technique is presented which is optimized, secure, and robust. The Watermark is set solidly in the discrete cosine transform domain. It is also encrypted based on the proposed block shuffling algorithm to increase security. The method was examined against various attacks, and it succeeded in maintaining statistical significance in terms of robustness and imperceptibility, as the average Peak Signal to Noise Ratio value is 65 dB, and the average Normalized Correlation value is close to one after apply all possible attacks. The results have been tested on MATLAB 2020a. [ABSTRACT FROM AUTHOR]