AI-DRIVEN CHAOTIC SYSTEM FOR SECURE IMAGE ENCRYPTION AND DECRYPTION IN SIMULINK

Abstract: An efficient ANN-based chaotic system for image encryption and decryption, implemented in Simulink and MATLAB using the Xilinx System Generator (XSG) tool. A feed-forward neural network (FFNN) with one hidden layer and Tangent Sigmoid activation ensures high accuracy, validated through a low mean square error. The generated Verilog code from XSG is executed in Vivado for pixel distribution, enabling hardware-level optimization. Security tests, including histogram analysis, PSNR, and SSIM, demonstrate robust encryption performance. The system’s analysis and visualization, conducted within Simulink, MATLAB, and Vivado, confirm its effectiveness. This approach showcases ANN-based chaotic systems for secure and real-time image processing.

Keywords: Artificial Neural Networks (ANN), Chaotic Systems, Simulink, MATLAB.