Medical images utilization for significant data hiding based on machine learning

Abstract

This paper uses a machine-learning algorithm to suggest a reliable watermarking approach. The watermarked image propagation through the download physical layer (PhyLyr) of WiMAX standards is simulated in this work (WiMAX is based on the IEEE suite of standards 802.16). By downloading the Physical Layer (PhyLyr) through OFDM, transform coefficients of the host image (Hosting-Image) can be embedded there. Data associated with watermarks are initially encrypted. In a medical image, Support Vector Networks (SVN) are used to classify regions of interest (ROI) and regions of no interest (Non-ROI). The findings of this manuscript revealed that greater signal-to-noise ratios might achieve a 106-bit error rate (Bit-Err-Rate), this means that SNR is exceeds 10.4 dB of SNR. The received cover image’s peak SNR (PSNR) for clinical applications is more than 37 dB.