GPR Image Recovery Effect on Faster R-CNNBased Buried Target Detection

다국어 초록 (Multilingual Abstract)

Measurements acquired through ground-penetrating radar (GPR) may contain missing information that needs to be recovered before the implementation of any post-processing method, such as target detection, since buried target detection methods fail and cannot produce desired results if the input GPR image contains missing information. This study proves that the recovery of missing information in a GPR image has a direct influence on the performance of subsequent target detection methods. Thus, state-of-the-art matrix completion methods are applied to the GPR image with missing information in both pixel- and column-wise cases with different missing rates, such as 30% and 50%. After the GPR image is successfully recovered, the faster region-based convolutional neural network (Faster R-CNN) target detection method is applied. The performance correlation between matrix completion accuracy and the target detection method’s confidence score is analyzed using both quantitative and visual results. The obtained results demonstrate the importance of GPR image recovery prior to any post-processing implementation, such as target detection.

참고문헌 (Reference)

1 C. Warren, "gprMax : open source software to simulate electromagnetic wave propagation for ground penetrating radar" 209 : 163-170, 2016

2 Z. Wen, "Solving a low-rank factori-zation model for matrix completion by a nonlinear successive over-relaxation algorithm" 4 (4): 333-361, 2012

3 X. F. Li, "Seismic data reconstruction with fractal interpolation" 51 (51): 855-861, 2008

4 D. Kumlu, "Operations Research for Military Organizations" IGI Global 375-399, 2009

5 O. Lopez, "Offthe-grid low-rank matrix recovery and seismic data reconstruction" 10 (10): 658-671, 2016

6 D. Kumlu, "Missing data recovery via deep networks for limited ground penetrating radar measurements" 14 (14): 754-, 2022

7 R. H. Keshavan, "Matrix completion from a few entries" 56 (56): 2980-2998, 2010

8 G. Shabat, "Interest zone matrix approximation" 23 : 678-702, 2012

9 D. Kumlu, "Ground penetrating radar data reconstruction via matrix completion" 42 (42): 4607-4624, 2021

10 D. J. Daniels, "Ground Penetrating Radar" The Institution of Engineering and Technology 2004

11 S. Ren, "Faster R-CNN : towards real-time object detection with region proposal networks" 28 : 91-99, 2015

12 M. T. Pham, "Buried object detection from B-scan ground penetrating radar data using Faster-RCNN" 6804-6807, 2018

13 Y. Xu, "An alternating direction algorithm for matrix completion with nonnegative factors" 7 (7): 365-384, 2012