Performance Evaluation and Comparative Analysis of Watermarking Algorithm Based on Adaptive Prediction Method

Now a days digital watermarking appeared as a solution for copyright detection, protection and maintenance of important data. This paper deals with a new reversible watermarking algorithm based on adaptive prediction error expansion, which can recover original image after extracting the hidden data. Embedding capacity of such algorithm depend on the prediction accuracy of the predictor. The method can embed secret data into 3×3 image block order by exploiting the pixel redundancy within each block. It has been observed that proposed method of reversible watermarking provide much better results in terms of Peak Signal to Noise Ratio (PSNR) and Mean Square Error (MSE) in comparison to existing literature.