Enhancing Steganography by Image Segmentation and Multi-level Deep Hiding

Abstract

In this paper, we present Modify Deep Hiding Extraction Algorithm (MDHEA) that is a steganography algorithm with Multi-Level Steganography (MLS) and color image segmentation. Through experimental results, MDHEA shows improvement in the results of previous works by securing encrypted secret data against attacks. We use segmentation to choose the appropriate segment, pass it on the cover image, calculate the value of the change at the pixel of the segment and select the best segment and its location in the cover image based on the least effect. MDHEA applies multi-level steganography to hide the confidential data in color images to ensure the integrity of the hidden data and obtain the largest volume of hidden data without distorting the image of the stego image. To reduce distortion in the cover image due to hiding a large amount of secret data and obtaining a high-quality stego image after hiding the secret data, we implement the Blue Smoothing Algorithm (BSA) to achieve smoothing the largest possible number of pixels in the image.