MMW Radiometric Image Feature Extraction Method of Concealed Objects Based on Watershed Algorithm
Qi Wang, Guang Feng Zhang · Applied Mechanics and Materials · 2012
Because of all-time, all-weather ability to work and strong penetration, millimeter-wave radiometer is used to inspect concealed objects hidden under clothing. Due to the restrictions of antenna, equipment and environment, the resolution of image is low and the contours of concealed objects are not clear. To improve the quality of radiometric images and the accuracy of edge detection, we firstly process the image using threshold segmentation, then, a method of combining morphology self-adaptive filter and watershed algorithm is proposed to do the further processing and feature extraction. The experiment results indicate that morphology self-adaptive algorithm and watershed algorithm can complete the edge extraction of the concealed objects well with the real objects.