Adaptive Steganography Using Improve Bit-plane Complexity Segmentation

Abstract

One of the primary challenges in Internet data transmission lies in safeguarding data from unauthorized access by potential attackers. The goal of content-adaptive steganography is to conceal data inside the image's intricate texture. This research introduces an improved algorithm for concealing messages within color images. The developed method incorporates bit-plane slicing and the RSA algorithm as its foundation, aiming to achieve a heightened level of security for data hiding. The algorithm's uniqueness lies in its adaptability, where the threshold is determined based on both the text and image characteristics. Subsequently, the public key is employed for encryption the thresholds, while the private key is utilized for decryption it. Performance criteria such as Mean Squared Error (MSE), Peak Signal to Noise Ratio (PSNR) and the Structural Similarity Index (SSIM) are using to assess the quality of the developed algorithm. The results indicate that when 90,000 bits are conceal in the image, it yields an acceptable PSNR value of 60.5749, MSE of 0.0569, and SSIM of 0.9996. The developed algorithm excels in data hiding, as evidenced by its favorable comparison with other studies.

Keywords

• steganography
• adaptive steganography
• bit plane complexity segmentation (bpcs)
• rsa