K. J. Somaiya

Remarkable advancement in social media websites and chats, sharing photos, audios, and videos has become very easy and dangerous. Digital image watermarking technique has the potential to address the issue of privacy, ownership, and authenticity of the media shared using such medium. Invisible digital image watermarking techniques are necessary which are imperceptible within host image and robust to common signal and image processing attacks. In this paper, we present a watermarking technique for digital images through adaptive texturization, statistical parameters, and Bhattacharya distance. The primary idea is to segment the host image into four different regions based on frequency distribution using discrete wavelet transform. Subsequently important statistical parameters mostly applied in image processing techniques such as mean, standard deviation, skew, kurtosis, and variance are calculated from the wavelet transform coefficients of each region of host and watermark image. These statistical parameters of segmented regions of host and watermark image are then applied to obtain Bhattacharya distance. Wavelet transform coefficients of watermark image are embedded into wavelet transform coefficients of one of the regions of carrier image with minimum Bhattacharya distance through embedding factor. Performance of the technique was tested on multiple host and watermark images under common image processing attacks, which yield better results.