Generalized iterative frequency domain deconvolution technique

Dhaene, Tom; Martens, Luc; Zutter, D. De

doi:10.1109/imtc.1993.382675

Cited by 15 publications

(4 citation statements)

References 3 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…2 illustrates, the noise components of the two waveforms are similarly shaped, while the signal components (centered around s) are approximately 180 out of phase with each other (an explanation is given later in the Additional Remarks and Observations section). Furthermore, the cost given by (11) reaches a minimum with nearly the same filter parameters that minimize the true mean-squared error given by the cost function in (9), as illustrated in Fig. 3.…”

Section: Optimizationmentioning

confidence: 87%

“…Others cannot make use of time-domain weighting because phase information is not included in the cost function [11]. We propose a new cost function having the same form as (9), where an indicated error given by (10) substitutes for (10) where indicates the inverse Fast Fourier transform of . This gives the following cost function: (11) Based on observations of simulated data, the function tends to have a shape similar to that of , particularly when the optimum filter parameters are approached.…”

Section: Optimizationmentioning

confidence: 99%

See 1 more Smart Citation

Noniterative waveform deconvolution using analytic reconstruction filters with time-domain weighting

Roy

Souders

2001

IEEE Trans. Instrum. Meas.

View full text Add to dashboard Cite

A new deconvolution approach is described for reconstructing fast, step-like, or impulsive signals that have been measured with a sampling oscilloscope for which an impulse response estimate is available. The approach uses analytic reconstruction filters to control noise amplification and a new noniterative filter optimization that is based on a calculated "indicated error" function that is similar in shape to the true error. The new method aids in reporting uncertainties of the deconvolution results and it permits the use of time-domain weighting to optimize the measurement of different waveform features. The performance of the proposed approach is compared with that of the Error Energy/Regularization approach that is currently popular.

show abstract

Section: Optimizationmentioning

confidence: 87%

Section: Optimizationmentioning

confidence: 99%

Noniterative waveform deconvolution using analytic reconstruction filters with time-domain weighting

Roy

Souders

2001

IEEE Trans. Instrum. Meas.

View full text Add to dashboard Cite

show abstract

Section: B) Kirchhoff Migrationmentioning

confidence: 99%

“…The fastest classical way of performing deconvolution uses a Wiener inverse filter in frequency domain along with FFT [12]. In our case, we extend it to the frequency-wavenumber domain as follows: where , , and are the complex spectra of the focused image, measured radar volume, and PSF, respectively; IFFT means an inverse FFT; symbol * stands for complex conjugate, and β denotes the regularization parameter originating from the inverse SNR.…”

Section: Regularized 3d Inverse Filteringmentioning

confidence: 99%

3D imaging by fast deconvolution algorithm in short-range UWB radar for concealed weapon detection

Savelyev

Yarovoy

2013

Int. J. Microw. Wireless Technol.

View full text Add to dashboard Cite

3D imaging by fast deconvolution algorithm in short-range UWB radar for concealed weapon detection timofey savelyev 1,2 and alexander yarovoy 1 A fast imaging algorithm for real-time use in short-range (ultra-wideband) radar with synthetic or real-array aperture is proposed. The reflected field is presented here as a convolution of the target reflectivity and point spread function (PSF) of the imaging system. To obtain a focused 3D image, the proposed algorithm deconvolves the PSF out from the acquired data volume with high speed due to fast Fourier transform and implementation in frequency-wavenumber domain. Then the result is tested against two numerical criteria for efficiency, namely error and instability, whose optimal values can be obtained iteratively. Since the PSF differs with distance, the algorithm suits mainly applications with relatively small objects such as concealed weapon detection. Using several PSFs allows us to image a certain range of interest by their successive deconvolution from the same data. Performance of the algorithm has been evaluated experimentally and compared with that of Kirchhoff migration. Measurements were carried out by a 5-25 GHz synthetic aperture radar in the lab, and scenarios included a gun and a ceramic knife in free space, on a large metal plate, and a gun concealed on a dummy under a thick raincoat. The results demonstrate sufficient image quality obtained in a fraction of time.

show abstract