A common hurdle in image steganography is balancing the amount of hidden data (embedding capacity) with how much the image’s appearance changes. Liu et al.6 developed an edge-detection method that modifies data embedding based on whether an image block contains edges. While this approach effectively keeps images looking good, it sacrifices embedding capacity by using two bits in each "reference" pixel one to mark an edge and one for secret data. This capacity loss is especially noticeable in images with many edges or high-frequency details. This paper introduces an enhanced steganography technique designed to overcome this limitation by using reference pixels more efficiently. Instead of two bits, our proposed method uses only one bit for the edge flag, freeing up the remaining bits in the reference pixel for additional secret data. We also employ an adaptive strategy where blocks with edges embed more data than non-edge blocks, striking a better balance between capacity and imperceptibility. Tests on standard images show that our scheme significantly boosts embedding capacity by up to 2.62 bits per pixel, all while maintaining or even improving the image’s visual quality. This means we can hide more information with less overhead and greater efficiency, making our method more suitable for secure, high-capacity communication.
Fatima et al. (Mon,) studied this question.