Novel Dominant Color Subband Image Encryption in Visual Sensor Network for Smart Military Surveillance System

Abstract

Internet of Things plays a major role in improving the levels of intelligence for several defense and military applications. There are myriads of visual sensor nodes deployed in various domain areas to acquire complete situational awareness. These sensor nodes must transmit the real-time information obtained from multiple locations to identify the threats in the danger-prone zones. The keyframes obtained from the visual sensor nodes should reach the command and control room in a confidentially manner. The lightweight scheme proposed for encryption in the visual sensor network is the novel Dominant color subband image encryption. This scheme involves encrypting only the dominant subband from an entire RGB image using the novel hybrid Chaotic method. The hybrid chaotic method combines the distinctive features of Quadratic, Ginger breadman, and Baker’s map. This hybrid scheme develops a complex, diverse chaotic structure that is unpredictable. A comparative analysis of this encryption scheme against other chaotic methods was done. It was found that the Novel Dominant color subband image encryption algorithm performs significantly better than the other existing methods to achieve real-time security in vision sensors. This method takes advantage of having lesser computation time and increased PSNR values, which indicates the higher quality of the images after decryption.