Hybridization of Discrete Cosine Transform and Principal Component Analysis to Achieve Digital Watermarking
Ashish Dixit, Rajesh Agarwal, B. K Sharma · 2023
This study presents an original method for the watermarking of grayscale images that are based on the discrete cosine transform in sun bands (SB-DCT) and principal component analysis (PCA). When applied to each non-overlapping image block, the SB-DCT is utilized to extract the image's features. These structures are used as a dataset for the PCA analysis that is carried out. PCA analysis inserts the binary watermark logo into image portions. that have a low level of variation. Because PCA has the best approximation function and the highest elaborating characteristic. capable of recovering the original watermark after demonstrating the effectiveness of several image processing methods According to the extensive testing results, the watermarked image has a High PSNR and a normalized correlation (NC). the retrieved watermark's high Bit Correct Ratio (BCR) value demonstrates the resilience and imperceptibility of the suggested approach When compared to cutting-edge approaches.