DCT-based Robust Reversible Watermarking Technique based on histogram Modification

In this paper, a strong, reversible image watermarking technique based on discrete cosine transform (DCT) and histogram shifting is proposed, where it overcomes the following concerns: (i) Reversing the cover object to its starting appearance is the primary goal of the reversible watermarking system. (ii) Military, medical, and standard law enforcement images are the main types of images that require distortion and reinstatement of the cover object following the watermark extraction. (iii) Lack of robustness and cover image-dependent embedding capacity are the primary concerns about reversible watermarking. Decompose the cover object into blocks that don't overlap in the first stage to insert a binary watermark bit into every block that is converted. These binary bits of watermark are embedded by altering a single set of middle substantial AC coefficients. To restore the cover image, subsequently using the histogram bin shifting method, a location map is created and integrated within the cover image. On the extracting side, at first, a location map is extracted from the image using the histogram bin shifting technique. In the following step, the image's watermark is recovered, and a reversed image has been generated using a location map. To verify the robustness property, several image processing attacks are tested with the suggested reversible watermarking approach, and favorable results are attained. The proposed scheme using the Lena image achieved 46.62 imperceptibility for 4096 embedding capacities. To methodically evaluate the proposed approach, it is compared with two current reversible watermarking systems, where they achieved 39.10 and 37.90 imperceptibility with 4.4 × 103 and 256 embedding capacities, respectively. The experimental results affirmed that the suggested method exhibits superior performance relative to these existing techniques.