Imaging of buried objects from multi-frequency experimental data using a globally convergent inversion method

Dinh-Liem Nguyen,Michael V. Klibanov,Loc H. Nguyen,Michael A. Fiddy

DOI: https://doi.org/10.48550/arXiv.1705.01219

2017-05-03

Abstract:This paper is concerned with the numerical solution to a 3D coefficient inverse problem for buried objects with multi-frequency experimental data. The measured data, which are associated with a single direction of an incident plane wave, are backscatter data for targets buried in a sandbox. These raw scattering data were collected using a microwave scattering facility at the University of North Carolina at Charlotte. We develop a data preprocessing procedure and exploit a newly developed globally convergent inversion method for solving the inverse problem with these preprocessed data. It is shown that dielectric constants of the buried targets as well as their locations are reconstructed with a very good accuracy. We also prove a new analytical result which rigorously justifies an important step of the so-called "data propagation" procedure.

Numerical Analysis

What problem does this paper attempt to address?

The main problem that this paper attempts to solve is: **Reconstruct the position and dielectric constant of buried objects from multi - frequency experimental data**. Specifically, the authors focus on using the globally convergent inversion method to solve the 3D coefficient inverse problem in order to reconstruct the position and dielectric constant of objects buried in a sandbox from backscattering data in a single incident plane - wave direction.

### Specific background and challenges of the problem:

1. **Application scenarios**:
- The main application target is to image underground - buried explosive devices, such as anti - personnel mines and improvised explosive devices (IEDs). To simulate this scenario, the authors placed objects simulating explosives in a sandbox and collected the backscattering data of these objects using a microwave scattering device.

2. **Data characteristics**:
- The data are multi - frequency backscattering data obtained from a single incident plane - wave direction. These data were collected in the microwave scattering facility at the University of North Carolina at Charlotte.
- Data pre - processing is one of the key steps because the original scattering data contains a lot of noise and other interference signals.

3. **Technical difficulties**:
- Inverse problems are usually nonlinear and ill - posed, which means that the solving process is very challenging.
- Especially for objects buried in sand, the existence of the air/sand interface complicates the signal. Therefore, how to effectively pre - process the data and extract useful information from it is an important research topic.

### Solutions:

1. **Globally convergent inversion method**:
- The authors developed a new globally convergent inversion method, which can find an initial guess point close to the real solution without prior knowledge and iteratively improve this guess step by step.
- This method ensures that data at different frequencies can be effectively utilized through a series of mathematical derivations and numerical implementations.

2. **Data pre - processing**:
- By pre - processing the experimental data, the original data are transformed into a form more suitable for the input of the inversion algorithm. Pre - processing steps include noise removal, signal correction, etc.
- The pre - processed data are regarded as data of targets placed in air, thus simplifying the subsequent calculation process.

3. **Theoretical proof**:
- The paper also provides a new theoretical result (Theorem 4), which strictly proves the rationality of an important step in the "data propagation" process, further enhancing the credibility of the method.

### Results and significance:

- By this method, the authors successfully reconstructed the position and dielectric constant of the buried objects with very high accuracy.
- These results not only help to improve the detection ability of explosives, but also can be used as a supplement to the energy information of radar images, potentially reducing the false alarm rate.

In conclusion, through the innovative globally convergent inversion method and effective data pre - processing technology, this paper solves the key problem of reconstructing buried objects from multi - frequency experimental data, providing strong support for practical applications.