Conjugate-Gradient Method for Soliving Inverse Scattering with Experimental Data

Lobel, P.; Kleinman, R. E.; Pichot, Ch.; Blanc‐Féraud, Laure; Barlaud, Michel

doi:10.1109/map.1996.511954

Cited by 58 publications

(28 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The investigation area is assumed of square shape and side equal to 0.06 m. It is partitioned into 20×20 square subdomains. The algorithm is initialized by using a rough estimate obtained by means of a back-propagation algorithm (Lobel, Kleinman, Pichot, Blanc-Feraud, & Barlaud, 1996). The inner loop (Landweber algorithm) is stopped after a fixed number of iterations = .…”

Section: Preliminary Reconstruction Resultsmentioning

confidence: 99%

Non-Invasive Microwave Characterization of Dielectric Scatterers

Costanzo¹,

Massa²,

Pastorino³

et al. 2012

Microwave Materials Characterization

View full text Add to dashboard Cite

Section: Preliminary Reconstruction Resultsmentioning

confidence: 99%

Non-Invasive Microwave Characterization of Dielectric Scatterers

Costanzo¹,

Massa²,

Pastorino³

et al. 2012

Microwave Materials Characterization

View full text Add to dashboard Cite

“…To-date there have been a good number of algorithms developed for tomographic image reconstruction including both Born-based methods and iterative methods [8,42,[49][50][51][52][53][54][55][56][57][58][59][60][61]. Iterative image reconstruction algorithms based on Gauss-Newton method [52], Newton-Kantorovich method [50,51], quasi-Newton method [53], Conjugate gradient method [60,61] and sequential quadratic programming method [55] have all been applied in microwave tomography.…”

Section: Technique For Solving the Inverse Scattering Problemmentioning

confidence: 99%

“…Iterative image reconstruction algorithms based on Gauss-Newton method [52], Newton-Kantorovich method [50,51], quasi-Newton method [53], Conjugate gradient method [60,61] and sequential quadratic programming method [55] have all been applied in microwave tomography. These methods would have different complexity, give different rates of convenience in the iterative process, produce images of different qualities, and have different levels of sensitivity to the quality of data.…”

Section: Technique For Solving the Inverse Scattering Problemmentioning

confidence: 99%

Microwave Tomography for Industrial Process Imaging: Example Applications and Experimental Results

Wang

2017

IEEE Antennas Propag. Mag.

View full text Add to dashboard Cite

Abstract---This paper describes the implementation of microwave tomography for industrial process applications. Microwave tomography for industrial process imaging has different requirements from that for medical imaging. In addition to spatial resolution, high temporal resolution or real-time imaging is also important for high speed processes, flows or rapid reactions. Depending on the specific application, both quantitative imaging and qualitative imaging may be needed. Qualitative imaging would be sufficient to display distributions, patterns or shapes, which may be adequate for some applications. Quantitative imaging would however be more informative, giving images with quantitative dielectric contrast or permittivity values from which other physical parameters such as density, moisture content and phase fraction may be derived. With the microwave tomography approach described, several example applications in industrial processes are demonstrated, and a number of experimental imaging results are presented. Keyword---microwavetomography, microwave imaging, instrumentation, image reconstruction I. INTRODUCTIONOMOGRAPHIC imaging is a technique that reconstructs the internal distribution of an object in terms of a physical parameter and presents it as an image with a grey or color scale for easy visualization. Instead of a single point measurement, it offers a unique imaging method to the oil, chemical, pharmaceutical and food industries to visualize industrial processes without invading the pipes and vessels and provide valuable information for control and optimization. Since the concept of tomographic imaging was put into practice in the 1970's [1,2], there has been a significant development in tomography instruments based on X-ray, -ray, magnetic resonance, optical, positron emission, ultrasound, electrical resistance, electrical capacitance, electromagnetic induction and microwaves [3,4]. Among them, X-ray and magnetic resonance tomographic imaging instruments, which are often known as X-ray CT and MRI respectively, have been widely used in the medical sector. X-ray CT, MRI and optical tomography can all produce images of high spatial resolution. But X-ray CT is only applicable to the imaging of objects with large density contrast. MRI is only useful for imaging objects which contain water or hydrogen atoms. Optical tomography is applicable only to optically transparent substances or gases where optical attenuations are measureable. X-ray, MRI and optical instruments are all bulky and have therefore not found applications in industrial imaging. Positron emission and -ray imaging techniques can also produce high resolution images, but both involve the use of radiative tracer particles or ionizing radiation. They are therefore not welcome in the industrial environment.On the other hand, electrical resistance and electrical capacitance tomographic techniques, which are often referred to as ERT and ECT respectively, have been applied to the imaging of industrial processes either in vessels or pipelines. ERT an...

show abstract

“…Among the various algorithms developed, one class comprises gradient methods [Lobel et al, 1996; Kleinman and van den Berg, 1992Berg, , 1993] that cast the inverse problem as an optimization problem, which iteratively converge to the actual profile by minimizing a cost function. The other class of iterative methods comprises Newton-type algorithms that linearize the system at every step.…”

Section: Introductionmentioning

confidence: 99%

A hybrid scheme for inverse scattering of electrically large regions

Rao

Carin

2000

Radio Science

View full text Add to dashboard Cite

Abstract. An efficient algorithm is presented for two-dimensional inverse scattering from electrically large regions. The technique is a hybrid combination of the modified-gradient (MG) method and a beam-tracing-based iterative Born method. The beam-tracing-based inversion cannot resolve sharp contrasts, but such are subsequently recovered by the MG algorithm. In particular, the hybrid scheme feeds the iterative Born reconstructed profile as the initial guess to the modified-gradient algorithm. The hybrid scheme has a highly improved convergence rate and also overcomes the limitations of the iterative Born and modified-gradient schemes, when each is applied alone. Results are presented for several lossless profiles. Issues addressed include the types of profiles that can be successfully imaged, the range of parameters over which the scheme is effective, and performance when noisy data are considered. IntroductionDiffraction tomography is an imaging technique that accounts for the underlying wave physics, and it has been developed and studied in the context of the Born and Rytov approximations [Devaney, 1981[Devaney, , 1982[Devaney, , 1983 The objective of this work is to establish an efficient hybrid method that combines the speed of the BTIB scheme and the accuracy of the MG algorithm, while reducing the drawbacks of both. This improvement is 315

show abstract

Conjugate-Gradient Method for Soliving Inverse Scattering with Experimental Data

Cited by 58 publications

References 6 publications

Non-Invasive Microwave Characterization of Dielectric Scatterers

Non-Invasive Microwave Characterization of Dielectric Scatterers

Microwave Tomography for Industrial Process Imaging: Example Applications and Experimental Results

A hybrid scheme for inverse scattering of electrically large regions

Contact Info

Product

Resources

About