Image Compression and Encryption Combining Autoencoder and Chaotic Logistic Map

• Published in: Iranian journal of science (Online) Vol. 44; no. 4; pp. 1091 - 1100
• Main Authors: Ameen Suhail K M, Sankar Syam
• Format: Journal Article
• Language: English
• Published: Shiraz Springer Nature B.V 01.08.2020
• Subjects: Algorithms; Compression; Deep learning; Digital imaging; Encryption; Image compression; Image manipulation; Image processing; Image transmission; Machine learning; Neural networks; Security
• ISSN: 2731-8095; 2731-8109
• DOI: 10.1007/s40995-020-00905-4

The third-party misuse and manipulation of digital images are a threat to the security and privacy of human subjects. In this paper, we report a system which effectively compresses and encrypts images to achieve secure transmission of image data with minimal bandwidth. The proposed system utilizes autoencoder for compression and chaotic logistic map for encryption. Autoencoder is an unsupervised deep learning neural network algorithm which compresses the input vector into a vector of fewer dimensions, which forms the dense representation of input data. This capability of the autoencoder can be used for compressing images. The encryption procedure is applied to the compressed data. The sequences generated by the logistic map are used efficiently in shuffling and encrypting the compressed image form. The security analysis confirms that the system is secure enough for transmitting image data.