摘要

Compared with optical-wave, millimeter-wave has the advantage of penetration through fog, smoke, clothing etc, which can detect weapons concealed underneath clothing. Unfortunately, the images obtained by millimeter-wave imaging systems have very low resolution. Thus image processing methods to improve image quality is necessary. In the paper, a new image algorithm combining Projection onto Convex Sets algorithm (POCS) and nonlinear extrapolation algorithm for improving millimeter-wave images is proposed and the experimental results show it is efficient.