Joint image encryption and compression schemes based on graph transform

Classical image compression coding has poor texture edge image block coding performance and non-sparse representations, and current encryption schemes are often difficult to achieve a good balance between encryption capability and compression performance, as well as coding format compatibility issues. In this paper, a joint scheme for image compression and encryption based on graph changes is proposed. The scheme firstly realizes adaptive selection of Laplace matrix using image texture and graph transformation, and performs optimal transformation selection. Secondly, group substitution and exclusive or operation are performed on the DC coefficients, and zero run length reduction method and substitution operation are performed on the AC coefficients to realize one-round encryption of frequency domain information. Then the color image layer structure is used for intra- and inter-layer interference to realize the second round of encryption. The experimental results prove that the scheme not only highly matches the image characteristics, but also has good compression and encryption characteristics, and is able to resist known plaintext attacks.