Tensor Tomographic Imaging for Foliage Penetration (FOPEN)

该博士学位的主要目的是开发解决雷达成像面临的挑战的基本工具，特别是在林业环境中。例如，我们如何减轻合成孔径雷达（SAR）图像中由于感兴趣的物体而产生的电磁波散射的非线性影响？更精确的传感器系统的发展，加上便携式设备计算能力的提高，导致了捕获丰富数据集的硬件的爆炸式增长。低频电磁波穿透树叶较好，但导致较低的分辨率图像使用已知的图像形成算法。此外，各向异性物体散射电磁波的接收多径/反弹增加了植被密集成像区域的复杂性。据设想，多极化、多频率、多视角和多静态几何的丰富数据结构将简化获取感兴趣区域的体积图像的过程，并增强对场景的描述，即张量断层扫描。由于树叶环境中电磁波的影响，高分辨率图像的检索被归类为非线性反问题，特别是麦克斯韦方程组的反边值问题（BVPs）。

The main aim of this PhD is to develop the tools that are fundamental in resolving the challenges faced by Radar Imaging, especially in a forestry environment. For instance, how would we alleviate the non-linear effects of Electromagnetic wave scattering due to objects of interests in Synthetic Aperture Radar (SAR) images? Developments in more accurate sensor systems coupled with the increase in computational power for portable devices give rise to an explosion in the hardware to capture rich datasets.Low Frequency Electromagnetic waves penetrate foliage better, but result in lower resolution images using known image formation algorithms. Furthermore, the received multi-path/bounce of Electromagnetic waves scattered by anisotropic objects adds complexity in imaging areas of dense vegetation. It is envisaged that a rich data structure of multi-polarisation, multi-frequency, multi-look and multi-static geometry will ease in the process of acquiring a volumetric image of the region of interest with an enhanced description of the scene, i.e. tensor tomography. Due to the effects of Electromagnetic waves in a foliage environment, the retrieval of a high resolution image is classed as a nonlinear Inverse Problem, specifically as Inverse Boundary Value Problems (BVPs) for Maxwell's equations.