Point Spread Function Deconvolution Research Articles

How to Coadd Images. II. Anti-aliasing and PSF Deconvolution

We have developed a novel method for co-adding multiple under-sampled images that combines the iteratively reweighted least squares and divide-and-conquer algorithms. Our approach not only allows for the anti-aliasing of the images but also enables Point-Spread Function (PSF) deconvolution, resulting in enhanced restoration of extended sources, the highest peak signal-to-noise ratio, and reduced ringing artefacts. To test our method, we conducted numerical simulations that replicated observation runs of the China Space Station Telescope/ the VLT Survey Telescope (VST) and compared our results to those obtained using previous algorithms. The simulation showed that our method outperforms previous approaches in several ways, such as restoring the profile of extended sources and minimizing ringing artefacts. Additionally, because our method relies on the inherent advantages of least squares fitting, it is more versatile and does not depend on the local uniformity hypothesis for the PSF. However, the new method consumes much more computation than the other approaches.

Enhancing the resolution of three‐dimensional migration images based on space‐variant point spread function deconvolution

AbstractImproving the resolution of seismic migration images plays an important role for geophysical interpreters to characterize underground reservoirs. However, the classical image domain least‐squares migration method based on the local‐stationary assumption cannot obtain a satisfactory high‐resolution seismic image due to the significant spatial variant characteristics of the point spread function. To mitigate this problem, we proposed a high‐resolution point spread function deconvolution method and applied it to two‐dimensional cases. Nevertheless, extending the two‐dimensional method to three‐dimensional problems directly would fail due to the intrinsic complexity in three‐dimensional cases. In this study, we resolve the differences encountered in the point spread function deconvolution method for two‐ and three‐dimensional cases and provide specific strategies for achieving high‐resolution imaging with low computational cost when extending the point spread function deconvolution method to three‐dimensional cases. The main schemes include (1) incorporating the analytical expression of the point spread function to guide the generation of three‐dimensional point spread function distributions, (2) extending the point spread function filter calculation method from a two‐dimensional square to a three‐dimensional rectangular prism and (3) interpolating to obtain more compact point spread functions for reducing migration artefacts. Results from the three‐dimensional synthetic Overthrust model and field data set demonstrate that our techniques could effectively enhance the spatial resolution of the migration images with reduced migration artefacts. With these specific strategies, the space‐variant point spread function deconvolution algorithm shows superior performance on three‐dimensional cases at a much lower computational cost compared with the classical least‐squares migration method and the local‐stationary deblurring method. Synthetic tests and real data applications confirm that the space‐variant point spread function deconvolution method has distinct advantages over both two‐ and three‐dimensional problems and can be widely adopted in practice.

Multi-axis electromagnetic sample handler for multi-view imaging and point-spread function deconvolution of organoids

Multi-axis electromagnetic sample handler for multi-view imaging and point-spread function deconvolution of organoids

PsfDeconNet: High-Resolution Seismic Imaging Using Point-Spread Function Deconvolution With Generative Adversarial Networks

PsfDeconNet: High-Resolution Seismic Imaging Using Point-Spread Function Deconvolution With Generative Adversarial Networks

Implementing the Point Spread Function Deconvolution for Better Molecular Characterization of Newly Diagnosed Gliomas: A Dynamic 18F-FDOPA PET Radiomics Study.

Purpose: This study aims to investigate the effects of applying the point spread function deconvolution (PSFd) to the radiomics analysis of dynamic L-3,4-dihydroxy-6-[18F]-fluoro-phenyl-alanine (18F-FDOPA) positron emission tomography (PET) images, to non-invasively identify isocitrate dehydrogenase (IDH) mutated and/or 1p/19q codeleted gliomas. Methods: Fifty-seven newly diagnosed glioma patients underwent dynamic 18F-FDOPA imaging on the same digital PET system. All images were reconstructed with and without PSFd. An L1-penalized (Lasso) logistic regression model, with 5-fold cross-validation and 20 repetitions, was trained with radiomics features extracted from the static tumor-to-background-ratio (TBR) and dynamic time-to-peak (TTP) parametric images, as well as a combination of both. Feature importance was assessed using Shapley additive explanation values. Results: The PSFd significantly modified 95% of TBR, but only 79% of TTP radiomics features. Applying the PSFd significantly improved the ability to identify IDH-mutated and/or 1p/19q codeleted gliomas, compared to PET images not processed with PSFd, with respective areas under the curve of 0.83 versus 0.79 and 0.75 versus 0.68 for a combination of static and dynamic radiomics features (p < 0.001). Without the PSFd, four and eight radiomics features contributed to 50% of the model for detecting IDH-mutated and/or 1p/19q codeleted gliomas, respectively. Application of the PSFd reduced this to three and seven contributive radiomics features. Conclusion: Application of the PSFd to dynamic 18F-FDOPA PET imaging significantly improves the detection of molecular parameters in newly diagnosed gliomas, most notably by modifying TBR radiomics features.

An Efficient and Stable High‐Resolution Seismic Imaging Method: Point‐Spread Function Deconvolution

AbstractBy fitting observed data with predicted seismograms, least‐squares migration (LSM) computes a generalized inverse for a subsurface reflectivity model, which can improve image resolution and reduce artifacts caused by incomplete acquisition. However, the large computational cost of LSM required for simulations and migrations limits its wide applications for large‐scale imaging problems. Using point‐spread function (PSF) deconvolution, we present an efficient and stable high‐resolution imaging method. The PSFs are first computed on a coarse grid using local ray‐based Gaussian beam Born modeling and migration. Then, we interpolate the PSFs onto a fine‐image grid and apply a high‐dimensional Gaussian function to attenuate artifacts far away from the PSF centers. With 2D/3D partition of unity, we decompose the traditional adjoint migration results into local images with the same window size as the PSFs. Then, these local images are deconvolved by the PSFs in the wavenumber domain to reduce the effects of the band‐limited source function and compensate for irregular subsurface illumination. The final assembled image is obtained by applying the inverse of the partitions for the deconvolved local images. Numerical examples for both synthetic and field data demonstrate that the proposed PSF deconvolution can significantly improve image resolution and amplitudes for deep structures, while not being sensitive to velocity errors as the data‐domain LSM.

Deep learning-based point-spread function deconvolution for migration image deblurring

High-resolution seismic imaging is crucial for deriving an accurate geologic section and ensuring successful petroleum exploration. However, traditional high-resolution imaging methods, such as least-squares reverse time migration (LSRTM) and migration deconvolution, usually involve high computational costs for approximating the inverse Hessian matrix with limited improvements in resolution. To obtain high-resolution migration images and reduce computing costs, we present a deep learning-based point-spread function (PSF) deconvolution method. We decompose the large Hessian matrix into small dispersed PSFs and design a convolutional neural network (CNN) to automatically predict the deconvolution operator of every single PSF. The deconvolution operator eliminates the PSF effect and balances the insufficient illumination of the acquisition geometry. To obtain an efficient CNN model for predicting deconvolution operators, we trained 2500 pairs of PSFs and their corresponding deconvolution operators collected from part of a model using the regular PSF deconvolution method. After training the network, we could predict all deconvolution operators in a few seconds for the following processing of migration images. The results on synthetic and field-data applications indicate that our method could provide a deblurred migration image comparable to that of the LSRTM approaches with significantly reduced computational and memory costs.

Image feature extraction and galaxy classification: a novel and efficient approach with automated machine learning

ABSTRACT In this work, we explore the possibility of applying machine learning methods designed for 1D problems to the task of galaxy image classification. The algorithms used for image classification typically rely on multiple costly steps, such as the point spread function deconvolution and the training and application of complex Convolutional Neural Networks of thousands or even millions of parameters. In our approach, we extract features from the galaxy images by analysing the elliptical isophotes in their light distribution and collect the information in a sequence. The sequences obtained with this method present definite features allowing a direct distinction between galaxy types. Then, we train and classify the sequences with machine learning algorithms, designed through the platform Modulos AutoML. As a demonstration of this method, we use the second public release of the Dark Energy Survey (DES DR2). We show that we are able to successfully distinguish between early-type and late-type galaxies, for images with signal-to-noise ratio greater than 300. This yields an accuracy of $86{{\ \rm per\ cent}}$ for the early-type galaxies and $93{{\ \rm per\ cent}}$ for the late-type galaxies, which is on par with most contemporary automated image classification approaches. The data dimensionality reduction of our novel method implies a significant lowering in computational cost of classification. In the perspective of future data sets obtained with e.g. Euclid and the Vera Rubin Observatory, this work represents a path towards using a well-tested and widely used platform from industry in efficiently tackling galaxy classification problems at the peta-byte scale.

Point-spread Function Deconvolution of the IFU Data and Restoration of Galaxy Stellar Kinematics

Abstract We present a performance test of the point-spread function (PSF) deconvolution algorithm applied to astronomical integral field unit (IFU) spectroscopy data for restoration of galaxy kinematics. We deconvolve the IFU data by applying the Lucy–Richardson algorithm to the 2D image slice at each wavelength. We demonstrate that the algorithm can effectively recover the true stellar kinematics of the galaxy, by using mock IFU data with a diverse combination of surface brightness profile, signal-to-noise ratio, line-of-sight geometry, and line-of-sight velocity distribution (LOSVD). In addition, we show that the proxy of the spin parameter λ R e can be accurately measured from the deconvolved IFU data. We apply the deconvolution algorithm to the actual SDSS-IV MaNGA IFU survey data. The 2D LOSVD, geometry, and λ R e measured from the deconvolved MaNGA IFU data exhibit noticeable differences compared to the ones measured from the original IFU data. The method can be applied to any other regular-grid IFU data to extract the PSF-deconvolved spatial information.

Effect of Point Spread Function Deconvolution in Reconstruction of Brain 18F-FDG PET Images on the Diagnostic Thinking Efficacy in Alzheimer's Disease.

Purpose: This study aims to determine the effect of applying Point Spread Function (PSF) deconvolution, which is known to improve contrast and spatial resolution in brain 18F-FDG PET images, to the diagnostic thinking efficacy in Alzheimer's disease (AD).Methods: We compared Hoffman 3-D brain phantom images reconstructed with or without PSF. The effect of PSF deconvolution on AD diagnostic clinical performance was determined from digital brain 18F-FDG PET images of AD (n = 38) and healthy (n = 35) subjects compared to controls (n = 36). Performances were assessed with SPM at the group level (p < 0.001 for the voxel) and at the individual level by visual interpretation of SPM T-maps (p < 0.005 for the voxel) by the consensual analysis of three experienced raters.Results: A mix of large hypometabolic (1,483cm3, mean value of −867 ± 492 Bq/ml) and intense hypermetabolic (902 cm3, mean value of 1,623 ± 1,242 Bq/ml) areas was observed in the PSF compared to the no PSF phantom images. Significant hypometabolic areas were observed in the AD group compared to the controls, for reconstructions with and without PSF (respectively 23.7 and 26.2 cm3), whereas no significant hypometabolic areas were observed when comparing the group of healthy subjects to the control group. At the individual level, no significant differences in diagnostic performances for discriminating AD were observed visually (sensitivity of 89 and 92% for reconstructions with and without PSF respectively, similar specificity of 74%).Conclusion: Diagnostic thinking efficacy performances for diagnosing AD are similar for 18F-FDG PET images reconstructed with or without PSF.

Blind deconvolution in astronomy with adaptive optics: the parametric marginal approach

ABSTRACT One of the major limitations of using adaptive optics (AO) to correct image post-processing is the lack of knowledge about the system’s point spread function (PSF). The PSF is not always available as direct imaging on isolated point-like objects, such as stars. The use of AO telemetry to predict the PSF also suffers from serious limitations and requires complex and yet not fully operational algorithms. A very attractive solution is to estimate the PSF directly from the scientific images themselves, using blind or myopic post-processing approaches. We demonstrate that such approaches suffer from severe limitations when a joint restitution of object and PSF parameters is performed. As an alternative, here we propose a marginalized PSF identification that overcomes this limitation. In this case, the PSF is used for image post-processing. Here we focus on deconvolution, a post-processing technique to restore the object, given the image and the PSF. We show that the PSF estimated by marginalization provides good-quality deconvolution. The full process of marginalized PSF estimation and deconvolution constitutes a successful blind deconvolution technique. It is tested on simulated data to measure its performance. It is also tested on experimental AO images of the asteroid 4-Vesta taken by the Spectro-Polarimetric High-contrast Exoplanet Research (SPHERE)/Zurich Imaging Polarimeter (Zimpol) on the Very Large Telescope to demonstrate application to on-sky data.

Imaging through motional scattering layers via PSF reshaping and deconvolution

Speckle deconvolution is a forcefully effective technique for imaging through scattering layers. However, this technique is not suitable for imaging through motional scattering layers, since the motion of the scattering layer after the point-spread-function (PSF) measurement makes the recovered image become blurred. To recover the image, the motion of the scattering layer is analyzed in four dimensions: rotation around the optical axis, translation along the optical axis, rotation around the cross-axis, and translation along the cross-axis. As a result, the speckles output from the scattering layer are distorted in four ways: speckle rotation, speckle scaling, speckle stretch, and speckle shift. According to these speckle distortions, the object image can be sharply recovered again by numerically reshaping the premeasured PSF. Moreover, the motion of the scattering layer can be roughly determined during the PSF reshaping process. This method effectively breaks the limitation of the deconvolution technique and has its meanings.

Single Underwater Image Restoration Based on Adaptive Transmission Fusion

Underwater images are often deteriorated with blurring, darkness, poor visual quality of low contrast, and color diminishing. This is mainly due to the fact that the light is exponentially attenuated while traveling through water and the strength of attenuation is color dependent. After constructing a simplified image formation model, this paper proposes a new strategy for single underwater image restoration. In light of different perspectives, two distinct transmission coefficient estimation approaches have been developed. One is based on the optical characteristics while the other relies on the essence of image processing knowledge. Subsequently, these two transmission maps are fused to produce the final outcome, which is adaptively weighted by their respective saliency maps. The obtained signal radiance is dissolved through point spread function deconvolution and color compensation to produce the final scene radiance. A variety of underwater images with various scenarios were exploited to evaluate the restoration performance. Experimental results demonstrated the superiority of the proposed algorithm over other competitive methods for underwater image restoration.

Single Underwater Image Restoration Based on Depth Estimation and Transmission Compensation

For undersea activities, underwater vehicles usually carry optical imaging systems for recording. The captured images and videos frequently suffer from two displeasing problems: First, color distortion; and second, poor visibility. This is mainly because that the light is exponentially attenuated while penetrating through water and the strength of attenuation is color dependent. This paper develops an effective single underwater image restoration framework based on the depth estimation and the transmission compensation. To address the consequences of scattering and absorption, our restoration scheme consists of five major phases: 1) background light estimation; 2) submerged dark channel prior; 3) transmission refinement and radiance recovery; 4) point spread function deconvolution; and 5) transmission and color compensation. A wide variety of underwater images with various scenarios were exploited to assess the restoration performance of the proposed algorithm. Various evaluation metrics were employed to analyze the experimental results. It was suggested that this new restoration algorithm outperformed many state-of-the-art methods both qualitatively and quantitatively. In addition, potential applications regarding autopilot and three-dimensional visualization were demonstrated. We believe that our underwater image restoration technique is promising in many undersea activities that require high-quality images.

The Orbit of the Close Companion of Polaris: Hubble Space Telescope Imaging, 2007 to 2014*

Abstract As part of a program to determine the dynamical masses of Cepheids, we have imaged the nearest and brightest Cepheid, Polaris, with the Hubble Space Telescope (HST) Wide Field Planetary Camera 2 and Wide Field Camera 3. Observations were obtained at three epochs between 2007 and 2014. In these images, as in HST frames obtained in 2005 and 2006, which we discussed in a 2008 paper, we resolve the close companion Polaris Ab from the Cepheid Polaris Aa. Because of the small separation and large magnitude difference between Polaris Aa and Ab, we used point-spread function deconvolution techniques to carry out astrometry of the binary. Based on these new measurements, we have updated the elements for the 29.59 year orbit. Adopting the distance to the system from the recent Gaia Data Release 2, we find a dynamical mass of 3.45 ± 0.75 M ⊙ for the Cepheid, although this is preliminary and will be improved by CHARA measurements covering periastron. As is the case for the recently determined dynamical mass for the Cepheid V1334 Cyg, the mass of Polaris is significantly lower than the “evolutionary mass” predicted by fitting to evolutionary tracks in the Hertzsprung–Russell diagram. We discuss several questions and implications raised by these measurements, including the pulsation mode, which instability-strip crossing the stars are in, and possible complications such as rotation, mass loss, and binary mergers. The distant third star in the system, Polaris B, appears to be older than the Cepheid, based on isochrone fitting. This may indicate that the Cepheid Polaris is relatively old and is the result of a binary merger, rather than being a young single star.

Improving the detection efficiency and modulation transfer function of lens-coupled indirect X-ray imaging detectors based on point spread functions simulated according to lens performance parameters.

Lens-coupled indirect X-ray imaging detectors have the advantage of high resolution and the disadvantage of low detection efficiency. Using thicker single-crystalline films (SCFs) can improve the detection efficiency. However, the image quality will become worse due to the degradation of the point spread function (PSF) and modulation transfer function (MTF). This disadvantage can be improved by deconvolution with the PSF, which is unknown. In this article, a method was established to acquire the PSF based on a simulation of the imaging process for a lens-coupled indirect X-ray imaging detector. Because the structural parameters of commercial lenses cannot usually be obtained, the PSFs were calculated from lens performance parameters. PSFs were calculated using the conditions of 12 keV X-ray energy, 10× and 40× magnification objectives and 4.6 µm- and 20 µm-thick GGG:Tb scintillators. These were then used to deconvolve images of an Xradia resolution test pattern taken under the same conditions. The results show that after deconvolution the MTF had been clearly improved for both the 4.6 µm- and 20 µm-thick SCFs, indicating that the image has better quality than before deconvolution. Furthermore, a PSF deconvolution was performed on mouse brain tissue projection images, and the original and deconvolution projection images were used to perform computed-tomography reconstruction; the result proved that the method was effective for improving the image quality of low-contrast samples. Therefore, this method shows promise in allowing the use of thick SCFs to improve the detection efficiency while maintaining good image quality.

Little Blue Dots in the Hubble Space Telescope Frontier Fields: Precursors to Globular Clusters?

Abstract Galaxies with stellar masses and specific star formation rates yr−1 were examined on images of the Hubble Space Telescope Frontier Field Parallels for Abell 2744 and MACS J0416.1-02403. They appear as unresolved “Little Blue Dots” (LBDs). They are less massive and have higher specific star formation rates (sSFRs) than “blueberries” studied by Yang et al. and higher sSFRs than “Blue Nuggets” studied by Tacchella et al. We divided the LBDs into three redshift bins and, for each, stacked the B435, V606, and I814 images convolved to the same stellar point-spread function (PSF). Their radii were determined from PSF deconvolution to be ∼80 to ∼180 pc. The high sSFRs suggest that their entire stellar mass has formed in only 1% of the local age of the universe. The sSFRs at similar epochs in local dwarf galaxies are lower by a factor of ∼100. Assuming that the star formation rate is for efficiency , gas mass M gas, and free-fall time, t ff, the gas mass and gas-to-star mass ratio are determined. This ratio exceeds 1 for reasonable efficiencies, and is likely to be ∼5 even with a high of 0.1. We consider whether these regions are forming today’s globular clusters. With their observed stellar masses, the maximum likely cluster mass is , but if star formation continues at the current rate for before feedback and gas exhaustion stop it, then the maximum cluster mass could become .

Viability of Point Spread Function Deconvolution for SEM

Viability of Point Spread Function Deconvolution for SEM

SYNCHRONIC CORONAL HOLE MAPPING USING MULTI-INSTRUMENT EUV IMAGES: DATA PREPARATION AND DETECTION METHOD

ABSTRACT We describe a method for the automatic mapping of coronal holes (CHs) using simultaneous multi-instrument EUV imaging data. Synchronized EUV images from STEREO/EUVI A and B 195 Å and SDO/AIA 193 Å are preprocessed, including point-spread function deconvolution and the application of data-derived intensity corrections which account for center-to-limb variations (limb brightening) and inter-instrument intensity normalization. We systematically derive a robust limb-brightening correction that takes advantage of unbiased long-term averages of data and respects the physical nature of the problem. The new preprocessing greatly assists in CH detection, allowing for the use of a simplified variable-connectivity, two-threshold region-growing image segmentation algorithm to obtain consistent detection results. We generate synchronic EUV and CH maps, and show a preliminary analysis of CH evolution. Several data and code products are made available to the community (www.predsci.com/chd): for the period of this study (2010 June 10 to 2014 August 18), we provide synchronic EUV and CH map data at 6 hr cadence, data-derived limb-brightening corrections for STEREO/EUVI A and B 195 Å and SDO/AIA 193 Å, and inter-instrument correction factors to equate their intensities. We also provide the CH image segmentation code module (ezseg) implemented in both FORTRAN-OpenMP and GPU-accelerated C-CUDA. A complete implementation of our CH detection pipeline in the form of a ready-to-use MATLAB driver script euv2chm utilizing ezseg is also made available.

Cardiac motion compensation and resolution modeling in simultaneous PET-MR: a cardiac lesion detection study

Cardiac motion and partial volume effects (PVE) are two of the main causes of image degradation in cardiac PET. Motion generates artifacts and blurring while PVE lead to erroneous myocardial activity measurements. Newly available simultaneous PET-MR scanners offer new possibilities in cardiac imaging as MRI can assess wall contractility while collecting PET perfusion data. In this perspective, we develop a list-mode iterative reconstruction framework incorporating both tagged-MR derived non-rigid myocardial wall motion and position dependent detector point spread function (PSF) directly into the PET system matrix. In this manner, our algorithm performs both motion ‘deblurring’ and PSF deconvolution while reconstructing images with all available PET counts. The proposed methods are evaluated in a beating non-rigid cardiac phantom whose hot myocardial compartment contains small transmural and non-transmural cold defects. In order to accelerate imaging time, we investigate collecting full and half k-space tagged MR data to obtain tagged volumes that are registered using non-rigid B-spline registration to yield wall motion information. Our experimental results show that tagged-MR based motion correction yielded an improvement in defect/myocardium contrast recovery of 34–206% as compared to motion uncorrected studies. Likewise, lesion detectability improved by respectively 115–136% and 62–235% with MR-based motion compensation as compared to gating and no motion correction and made it possible to distinguish non-transmural from transmural defects, which has clinical significance given the inherent limitations of current single modality imaging in identifying the amount of residual ischemia. The incorporation of PSF modeling within the framework of MR-based motion compensation significantly improved defect/myocardium contrast recovery (5.1–8.5%, p < 0.01) and defect detectability (39–56%, p < 0.01). No statistical difference was found in PET contrast and lesion detectability based on motion fields obtained with half and full k-space tagged data.