关键词：压缩传感;探地雷达;地雷;GPR
摘 要：This paper investigates the implementation of compressive sensing (CS) for handheld GPR sensor system for landmine detection. The system collects GPR data in random and sparse spatial positions, resulting in the limited resolution of the GPR image. Compressive sensing which is widely known to be able to reconstruct sparse information from limited number of measurements is applied for this system. In this study, we present landmine detection by model based CS that exploit the physical characteristic of the landmine in CS processing. The performance is evaluated using real data sets acquired in a landmine field in Cambodia. The results show that the model based CS is able to give better visualization of a buried PMN-2 landmine compared to the conventional CS processing. It is also observed that the accuracy of the block sparsity level, which is the estimation of how many targets exist in the area, is an important factor for the correct image reconstruction by the model based CS.
目 录：