A Fast Back-Projection Algorithm Based on Cross Correlation for GPR Imaging
A Fast Back-Projection Algorithm Based on Cross Correlation for GPR Imaging
复制标题
DOI：
10.1109/lgrs.2011.2165523
复制
发表时间：
2012-03
影响因子：
4.8
通讯作者：
Lin Zhou;Chunlin Huang;Yi Su
中科院分区：
文献类型：
--
作者：
Lin Zhou;Chunlin Huang;Yi Su
在探地雷达成像中，经典的BP算法在层状介质成像方面具有便捷性和鲁棒性。然而，经典的BP算法耗时长，且存在大量伪影，对后续的检测和识别等工作产生不利影响。提出了一种新的BP算法，该算法速度快，抑制伪影效果好。首先，采用一种经认可的近似方法计算折射点位置，计算速度快，精度高。然后，在经典BP算法中使用查找表来减少冗余。为了实现有效的伪影抑制，提出了一种基于互相关的方法。现场数据的实验结果表明，本文提出的BP算法在运算速度和伪影抑制方面都优于经典算法。
In ground-penetrating radar imaging, the classic back-projection (BP) algorithm has an excellent reputation for imaging in layered mediums with convenience and robustness. However, the classic BP algorithm is time consuming and with a lot of artifacts, which have adverse effects on the following work like detection and recognition. A novel BP algorithm, which is both fast and with good effect of suppressing artifacts, is proposed in this letter. At first, an approved approximation method is used to calculate the position of refraction point with remarkable speed and satisfactory accuracy. Then, a lookup table is used to reduce the redundancy in classic BP algorithm. In order to achieve effective artifact suppression, a cross-correlation-based method is introduced. Experimental results of field data present the superiority of the proposed BP algorithm over its classic counterpart both in operation speed and artifact suppression.