Open Access

ARTICLE

Edge-Based Data Hiding and Extraction Algorithm to Increase Payload Capacity and Data Security

Hanan Hardan^1,*, Osama A. Khashan^2,*, Mohammad Alshinwan¹

1 Faculty of Information Technology, Applied Science Private University, Amman, 11931, Jordan
2 Research and Innovation Centers, Rabdan Academy, Abu Dhabi, 114646, United Arab Emirates

* Corresponding Authors: Hanan Hardan. Email: ; Osama A. Khashan. Email:

Computers, Materials & Continua 2025, 84(1), 1681-1710. https://doi.org/10.32604/cmc.2025.061659

Received 29 November 2024; Accepted 18 March 2025; Issue published 09 June 2025

Abstract

This study introduces an Edge-Based Data Hiding and Extraction Algorithm (EBDHEA) to address the problem of data embedding in images while preserving robust security and high image quality. The algorithm produces three classes of pixels from the pixels in the cover image: edges found by the Canny edge detection method, pixels arising from the expansion of neighboring edge pixels, and pixels that are neither edges nor components of the neighboring edge pixels. The number of Least Significant Bits (LSBs) that are used to hide data depends on these classifications. Furthermore, the lossless compression method, Huffman coding, improves image data capacity. To increase the security of the steganographic process, secret messages are encrypted using the XOR encryption technique before being embedded. Metrics such as the Mean Squared Error (MSE), Peak Signal-to-Noise Ratio (PSNR), and Structural Similarity Index Measure (SSIM) are used to assess the efficacy of this algorithm and are compared to previous methods. The findings demonstrate that the suggested approach achieves high similarity between the original and modified images with a maximum PSNR of 60.7 dB for a payload of 18,750 bytes, a maximum SSIM of 0.999 for a payload of 314,572.8 bytes, and a maximum Video Information Fidelity (VIF) of 0.95 for a payload of 23,592 bytes. Normalized Cross-Correlation (NCC) values are very close to 1. In addition, the performance of EBDHEA is implemented on Secure Medical Image Transmission as a real-world example, and the performance is tested against three types of attacks: RS Steganalysis, Chi-square attack, and visual attack, and compared with two deep learning models, such as SRNet and XuNet.

---

Keywords

---