New Image Steganography Method Based on K-means Clustering
Ismail Kich, El Bachir Ameur, Abdelghani Souhar · 2017
In this paper, a new image steganography method based on k-means clustering for embedding secret messages into a gray image is proposed. The proposed methodology is a combination of two techniques, image clustering and Least Significant Bits replacement. In the embedding process, a cover image is segmented into clusters using the K-means clustering algorithm. Each cluster is partitioned into two regions, smooth and complex, only smooth regions may contain the secret data. In this manner, degradation of the Stego image quality is imperceptible to the human eye. For better protection of secret message, a pseudo-random key mechanism is coupled to the implementation of the LSB method. The secret data can be recovered directly from the image Stego without reference to cover-image. The experimental results show that the proposed method has a high image quality and a good embedding capacity.