A Chaotic Image Encryption Based on Random Noise and Arnold Cat Maps

The increasing prevalence of data theft incidents, particularly involving sensitive images like identification documents or medical records, poses significant risks to individuals and organizations. To mitigate these risks, employing robust image encryption techniques is crucial. This study proposes a solution whereby strong encryption methods, including random noise based on ensemble Gaussian and salt and pepper techniques, along with the Arnold Cat Map (ACM), are applied to protect image data. The effectiveness of this approach is demonstrated through measurements of Mean Squared Error (MSE), Peak Signal-to-Noise Ratio (PSNR), Unified Average Changing Intensity (UACI), and Number of Pixel Change Rate (NPCR). Results indicate a notable increase in UACI values from the Random Noise to the Proposed Encryption stage, suggesting heightened security against unauthorized access. Specifically, for the Lena image, the MSE and PSNR at the Random Noise stage are (7.778, 22.131 dB), at the Image Scrambling stage are (19.111, 9.86 dB), and at the Proposed Encryption stage are (33.211, 4.922 dB). Similarly, for the Peppers image, the MSE and PSNR at the Random Noise stage are (9.898, 18.131 dB), at the Image Scrambling stage are (22.776, 9.311 dB), and at the Proposed Encryption stage are (36.854, 4.944 dB). Additionally, for the Baboon image, the MSE and PSNR at the Random Noise stage are (9.297, 19.131 dB), at the Image Scrambling stage are (22.41, 9.722 dB), and at the Proposed Encryption stage are (37.812, 4.989 dB). NPCR values remain consistently high across all encryption stages and images, indicating reliable pixel change detection. For future research, exploring advanced encryption algorithms and hybrid approaches could further enhance image security and foster innovation in data protection methods.