Image Steganography in a Karhunen-Loeve Transform Optimization Model

In allusion to such problems as large perceptual distortion and high error rate caused by high image compression ratio in existing steganography technology in the information security field, an image Steganography based on Karhunen-Loeve transform optimization is proposed in this paper. Specifically, the iterative clustering algorithm is adopted for this method to solve the covariance matrix and the clustering mean value, and relevant values are adjusted for image segmentation; then, KLT algorithm is introduced to compress the image data and the least significant bit is adopted to replace the ciphertext data for data hiding. During information extraction, the reverse linear transformation operation and the original pixel matrix are adopted to obtain the effective hidden image information. The experiment result shows: compared with common algorithms, the proposed method has improved capacity and PSNR, and the image data extracted thereby has small distortion.