An Optimum Method for Noise Reduction and Quality Improvement of the Passive Millimeter Wave Images Based on Nonsubsampled Shearlet Transform and Improved Adaptive Median Filter

Today, moving prohibited objects is one of the security threats in high-traffic places like airports, and it is essential to have an efficient system to deal with them. Current systems, which often focus on metal detection, may not only consider unsuspicious items such as coins as threats but also fail to detect new explosives that are not metal-based. A passive millimeter wave (PMMW) imaging system is an image formation method through the passive detection of energy naturally emitted from an object in the millimeter wave range. This system generates images of concealed objects in the human body by measuring the radiometric temperature emitted from the front scene in the millimeter wave band as a safe security tool. The image produced by this system is noisy and has low resolution, and it is not possible to show all its details compared to the visible image. In this paper, a method is presented for noise reduction in PMMW images and better detection of hidden targets compared to the background of the human body. In this method, first, the noise is detected then an adaptive median filter is proposed to remove the low-density salt-and-pepper noise. Next, another proposed algorithm based on the Shearlet transform is presented to reduce the Gaussian noise in the high-frequency components of the image. The implementation results show that the proposed algorithm efficiency is up to 6% better than the latest methods used for noise reduction and improve the resolution of real images received from a PMMW system.