Comparison Between Back Projection Algorithm and Range Migration Algorithm in Terahertz Imaging

Abstract

Back projection algorithm (BPA) and range migration algorithm (RMA) are two common approaches for image reconstruction in synthetic aperture radar (SAR) imaging. Imaging is implemented in spatial domain and spatial frequency domain, respectively and these two methods are often considered to generate the same results. This is true in normal microwave/millimeter wave SAR imaging systems, since the object of interests is located far away and the applied wavelength is long. In this case, the location of the object is the most important issue and there is no impetus to image the object with very fine details. In case of terahertz (THz) imaging, we have found some differences by using the two algorithms above, including the position of the object and resolution in 2D and 3D imaging. Some insights are given and we believe there should be some guidelines regarding the pros and cons before choosing the exact algorithm, especially when high-performance imaging is pursued. INDEX TERMS Ambient environment sensing, THz imaging, image reconstruction.

Real-Time Imaging With Simultaneous Use of Born and Rytov Approximations in Quantitative Microwave Holography

“…The computational complexity of QMH is comparable to that of well-established highly efficient k-space 3-D image reconstruction methods, such as 3-D synthetic aperture radar (SAR) range migration algorithm (RMA) [29]- [32] and closely related microwave holography [1], [2], [33]- [35].…”

“…Another common real-time image reconstruction method is the backprojection algorithm (BPA). In its original implementation, the BPA operates in real (x, y, z) space (see, e.g., [32]) and it has high computational complexity of O(N 2…”

Multi-Phase-Center Sidelobe Suppression Method for Circular GBSAR Based on Sparse Spectrum

“…In this section, the linear frequency modulation (LFM) signal is used as the transmitted signal，to introduce the signal model. Generally, range pulse compression is performed in the frequency domain, so the frequency domain form of the preprocessed echo signal can be expressed as [26,27]. Circular GBSAR can achieve 3D space imaging by superimposing 2D images at different range.…”

3-D Terahertz Imaging Based on Piecewise Constant Doppler Algorithm and Step- Frequency Continuous-Wave Signaling

“…Firstly, the FFT operation requires a uniform spatial sampling in 2-D aperture, leading to an inflexible system configuration [8]. Secondly, the range cell migration (RCM) [9], [17] leads to the samples placed nonuniformly in spatial frequency domain [18]. This has to be compensated by using linear interpolation techniques.…”

“…Therefore, the FFT-based methods have to reconstruct several subimages block by block and then perform image stitching to form a large image. There are also many imaging algorithms based on time-domain correlation [18], [19] which process the received data directly in spatial domain and remove the intrinsic limitations of the FFTbased methods. However, the huge computational cost and data storage limit their practical applications.…”