Image encryption method based on chaotic fuzzy cellular neural networks

• Main Authors: Ratnavelu, Kurunathan, Kalpana, M., Balasubramaniam, P., Wong, K., Paramesran, Raveendran
• Format: Article
• Published: Elsevier 2017
• Subjects: QA Mathematics; T Technology (General)
• Online Access: http://eprints.um.edu.my/17550/; https://doi.org/10.1016/j.sigpro.2017.05.002
• Institution: Universiti Malaya

In this work, an image encryption method is proposed based on fuzzy cellular neural network (FCNN). First, the shortcomings of FCNN in encrypting image are identified, and the FCNN model is then modified to address these shortcomings. Specifically, a theoretical framework is developed to identify the values of the parameters of FCNN to generate chaotic signals, which are in turn utilized to encrypt the image. The encryption method is designed where an encrypted pixel is generated based on the corresponding plaintext pixel together with the neighbouring encrypted pixels. The proposed method has a key sensitivity in the order of 10−10 to achieve adequate security robustness. Further evaluations on standard test images verified and confirmed that the proposed encryption method is robust against plaintext-only (i.e., brutal force) and chosen-plaintext attacks.