An SVM Based Secural Image Steganography Algorithm for IoT

Abstract

With the fast development of IoT network, there are more and more images generated by sensors and other devices, which increases the transmission expenses. By adopting image steganography, the images can deliver more information than they could. Therefore, the transmission expenses could be significantly reduced. However, the safety and quality of steganographic algorithms is not promising nowadays. To improve this situation, we propose an SVM-based steganography algorithm. The algorithm takes advantage of four features, including the variance of the image, the overall difference, the shape context matching and the smoothness. The analysis and experimental results show that the information hiding algorithm can effectively optimize the information steganography and anti-steganography analysis, which could be used in IoT.