Dual reversible datahiding using deep neural rhombus histogram in clinical images

Security is vital issues to be communicated for execution of e-healthcare scheme. Till now data hiding has most powerful instrument that can handle previous challenges. Reversible data hiding (RDH) planned to embed secret message in cover image at receiving end, both secret message and original cover image is recovered. Previous RDH methods designed for higher embedding rate. In this work, Deep Neural Rhombus Histogram based Dual Reversible Data Hiding (DNRH-DRDH) is proposed method for addressing both embedding rate as well as visual quality. It utilizes interpolated images reversible data hiding with Block Truncated Rhombus Prediction. Prediction error histogram generated embedded with secret data for obtaining stego images of corresponding interpolated pixels. Lastly, classification output is generated by ReLU activation function in extracted part for secured communication among users. Simulation carried on medical test images. Study reveals of proposed method gives higher embedding rate with superior image quality than existing methods.