Optimal cross-polarization spatial-depth matrix for millimeter-wave concealed threat detection by using polarization,...

Active millimeter-wave (MMW) technology has been widely applied in personnel security screening. However, the potential of polarization information has yet to be fully exploited in current systems. To address this issue, this paper proposes a direct detection method that relies solely on cross-polarization power information analysis without utilizing phase information, enabling effective detection of concealed threats due to the significant difference in cross-polarization echo power between concealed threats and the human body surface, while offering the potential for a higher-flexibility detection range and a simpler system architecture compared to imaging techniques. By applying the inverse Fourier transform to the echo power spectrum to obtain the depth spectrum, combined with spatial scanning, to our knowledge a new data representation termed the spatial-depth matrix is derived, characterizing the spatial and depth distributions of the target. Additionally, a method for obtaining the optimal cross-polarization angle under linear polarization is developed by fitting the power versus incident polarization angle data to a theoretical curve, enhancing the contrast between the human body and concealed threats. Based on the two-dimensional features of the spatial-depth matrix, this study establishes quantitative discrimination criteria, including the standard deviation of the spatial spectrum gradient (SDSSG) and the depth spectrum peaks (SDDSP). Based on our experimental data, the SDSSG and SDDSP for individuals carrying concealed threats exhibited changes of over 50% and 157%, respectively, compared to normal individuals. This significant difference facilitates concealed threat detection through thresholding using these two metrics.