adaptive image steganography with LSB embedding using image features

Abstract

Low distortion and high embedding capacity is one of the main topics in the area of steganography. There is many research based on these issues. Some of them focus on the high embedding capacity while others focus on low distortion area. Both have their pros and cons but low distortion area is more important because of the following reasons. First, it is less susceptible to detection. Second, usually the secret information is not very big. In this paper, different schemes of adaptive image steganography having lower distortion with relatively high embedding is proposed, which are based on different block size and number of thresholds used. Image is converted into non-overlapping blocks and valid blocks for information embedding are selected depending on difference between pixels of the block. Difference of pixels in a block is compared with the thresholds, which are adaptive and user can control these thresholds using secret scaling factors. Encrypted information is then embedded in specific pixels of valid blocks using least significant bit (LSB) embedding method. Blocks are then reshaped to generate stego image. All proposed schemes are compared with each other for best performance. These schemes are also analyzed and compared with other conventional schemes of image steganography.