Robust Spatial-Domain Watermarking Methods Based on a Weighting Table with Fine-Tune Technique

Abstract

Spatial-domain digital watermarking methods are generally considered as having poor performance after geometric distortion (such as cropping and scaling), common signal processing (such as JPEG and filtering), or subterfuge attacks (such as changing the least-significant bits). On the other hand, frequency-domain watermark techniques, in addition to their high computation complexity, usually require original image data during extraction. The methods, without referring to the original image during watermark extraction, are based on a predefined weighting table to record the magnitude difference of image pixel values and a novel fine-tune process to gradually change pixel values in order to reduce the impact on image quality. The performance can be further improved by properly selecting reference pixels for each partitioned 8 ´8 block and determining the reliability metrics for proper weighting during watermark extraction. Experimental results show promising robustness performance compared to other spatial-domain and frequency-domain methods.