A Novel Approach to Steganography Security: Embedding Encrypted Text in Color Images Using Langton’s Ant Cellular Automaton and Hybrid Chaotic Map

A Novel Approach to Steganography Security: Embedding Encrypted Text in Color Images Using Langton’s Ant Cellular Automaton and Hybrid Chaotic Map

연구 분야: Strategies

학회: SN Computer Science

The proliferation of communication technology has led to the emergence of various threats to secure information transmission. To counteract these threats, steganography has been developed as a technique for hiding confidential information within cover objects, thereby preventing data theft across a rapidly growing network. In this paper, we propose a novel approach that conceals encrypted text messages within a cover image using a two-level encryption algorithm. The first level of encryption involves scrambling the original text messages through an Artificial Neural Network (ANN) architecture, while the second level employs a Hybrid tent-logistic map and complex XOR operations to further encrypt the scrambled text. To ensure the randomness of the cover image coordinates, we use Langton’s Ant cellular automation to generate the necessary coordinates. Our approach utilizes standard colour images as cover images, and the encrypted text messages are concealed within the RGB pixel at random coordinates. The algorithm also integrates key parameters such as entropy, NPCR, UACI, and correlation values to ensure robustness. Additionally, the scheme utilizes SSIM, PSNR, and MSE for image quality analysis, making the detection of hidden data challenging. The use of a larger keyspace makes our encryption process more resilient against different types of attacks. We employ the linear feedback shift register (LFSR) algorithm to control the movement of Langton’s Ant. After the encrypted text message is concealed, both the cover image and stego image are indistinguishable from each other, making it difficult to trace the encrypted text within the colour image through steganalysis. Our experimental results, including the NPCR and PSNR measures, demonstrate that our proposed approach yields better results than other approaches. The NPCR, UACI, SSIM, and PSNR values are evaluated in this work. We got SSIM as 0.9994, PSNR as 52.07, NPCR as 99.97 and UACI as 32.2094 in our study.