Separable Reversible Data Hiding in Encrypted Images With High Capacity Based on Median-Edge Detector Prediction

Li Liu,Anhong Wang,Chin-Chen Chang

doi:10.1109/access.2020.2972736

Li Liu, Anhong Wang + Show 1 more

Open Access

https://doi.org/10.1109/access.2020.2972736

Copy DOI

Abstract

In recent years, reversible data hiding technology in encrypted images has become an important topic for researchers. In addition, separable reversible data hiding technology in encrypted images also has been developed extensively because of its better practicability. In this paper, a new scheme of separable reversible data hiding in encrypted image is proposed based on the median-edge detector prediction. First, the predicted value of each pixel is calculated by using the median-edge detector predictor, and the difference between the original pixel and the predicted pixel is recorded according to the threshold parameter, and a label map is generated. Second, the original image is encrypted. Third, the differences and the secret bits are embedded into the pixels according to certain rules. After being compressed by using Run Length Coding and Huffman Coding, the label map is spliced to the bottom of the encrypted image. Finally, the stego-image is generated. When the receiver possesses the stego-image, which are the encrypted image that contain the hidden data, the desired image can be extracted or decrypted in accordance with a different key. The experimental results showed that our proposed scheme achieved better hiding capacity than the other schemes.

Highlights

In recent years, with the rapid development of computer networks and multimedia technology, especially the advent of cloud technology, more and more digital images are stored and processed in the cloud platform
We propose a separable reversible data hiding scheme in encrypted image based on the median-edge detector (MED) prediction (SRDH-EI-MED)
EXPERIMENTAL RESULTS we discuss the experiments we conducted to evaluate the performance of the proposed scheme and to compare the proposed scheme with several earlier Rottondi .technology in encrypted images (RDH-EI) schemes

Summary

INTRODUCTION

With the rapid development of computer networks and multimedia technology, especially the advent of cloud technology, more and more digital images are stored and processed in the cloud platform. In order to recover the original image completely, the label of each pixel must be recorded in binary code as extra data and embedded into the encrypted image. The remaining extra data and reference pixels are embedded into the encrypted image by substituting multiple MSBs according to the label map, as shown in (6): t xe(i, j) mod 27−t + (bs × 27−s), xe(i, j) = 8 s=0. A. DIFFERENCE VALUE AND LABEL MAP GENERATION For an original image I that has the size of M × N , the predicted values px(i, j) of the pixels (except the pixels in the first row and the first column) can be calculated by using the MED predictor [23].

ENCRYPTION OF AN IMAGE

EXPERIMENTAL RESULTS

CONCLUSION