Michal Šorel

Space-Variant Restoration of Images Degraded by Camera Motion Blur

We examine the problem of restoration from multiple images degraded by camera motion blur. We consider scenes with significant depth variations resulting in space-variant blur. The proposed algorithm can be applied if the camera moves along an arbitrary curve parallel to the image plane, without any rotations. The knowledge of camera trajectory and camera parameters is not necessary. At the input, the user selects a region where depth variations are negligible. The algorithm belongs to the group of variational methods that estimate simultaneously a sharp image and a depth map, based on the minimization of a cost functional. To initialize the minimization, it uses an auxiliary window-based depth estimation algorithm. Feasibility of the algorithm is demonstrated by three experiments with real images.

Space-variant deblurring using one blurred and one underexposed image

We propose a practical method to remove photo blur due to camera shake, which is a typical problem when taking photos in dim lighting conditions such as indoor or night scenes. We use a pair of images, one of them blurred and the other one underexposed or noisy because of high ISO setting. Existing methods assume convolution model, that is the same blur in the whole image. It is seldom true in practice, especially for wide angle lens photos. We apply a space-variant model of blurring valid in many real situations. Results are documented by a photograph of a night scene.

Retinal image restoration by means of blind deconvolution

Retinal imaging plays a key role in the diagnosis and management of ophthalmologic disorders, such as diabetic retinopathy, glaucoma, and age-related macular degeneration. Because of the acquisition process, retinal images often suffer from blurring and uneven illumination. This problem may seriously affect disease diagnosis and progression assessment. Here we present a method for color retinal image restoration by means of multichannel blind deconvolution. The method is applied to a pair of retinal images acquired within a lapse of time, ranging from several minutes to months. It consists of a series of preprocessing steps to adjust the images so they comply with the considered degradation model, followed by the estimation of the point-spread function and, ultimately, image deconvolution. The preprocessing is mainly composed of image registration, uneven illumination compensation, and segmentation of areas with structural changes. In addition, we have developed a procedure for the detection and visualization of structural changes. This enables the identification of subtle developments in the retina not caused by variation in illumination or blur. The method was tested on synthetic and real images. Encouraging experimental results show that the method is capable of significant restoration of degraded retinal images.

Removing Boundary Artifacts for Real-Time Iterated Shrinkage Deconvolution

We propose a solution to the problem of boundary artifacts appearing in several recently published fast deblurring algorithms based on iterated shrinkage thresholding in a sparse domain and Fourier domain deconvolution. Our approach adapts an idea proposed by Reeves for deconvolution by the Wiener filter. The time of computation less than doubles.

Fast convolutional sparse coding using matrix inversion lemma

Convolutional sparse coding is an interesting alternative to standard sparse coding in modeling shift-invariant signals, giving impressive results for example in unsupervised learning of visual features. In state-of-the-art methods, the most time-consuming parts include inversion of a linear operator related to convolution. In this article we show how these inversions can be computed non-iteratively in the Fourier domain using the matrix inversion lemma. This greatly speeds up computation and makes convolutional sparse coding computationally feasible even for large problems. The algorithm is derived in three variants, one of them especially suitable for parallel implementation. We demonstrate algorithms on two-dimensional image data but all results hold for signals of arbitrary dimension. Interesting alternative to sparse coding for shift-invariant signals (images, audio).New fast algorithm, which makes sparse coding feasible even for large problems.3 versions of the algorithm, one of them designed for parallel implementation.Convolution kernels can be learned at several scales simultaneously.Algorithms are demonstrated on images but can be used for arbitrary signals.

Platform motion blur image restoration system

Platform motion blur is a common problem for airborne and space-based imagers. Photographs taken by hand or from moving vehicles in low-light conditions are also typically blurred. Correcting image motion blur poses a formidable problem since it requires a description of the blur in the form of the point spread function (PSF), which in general is dependent on spatial location within the image. Here we introduce a computational imaging system that incorporates optical position sensing detectors (PSDs), a conventional camera, and a method to reconstruct images degraded by spatially variant platform motion blur. A PSD tracks the movement of light distributions on its surface. It leverages more energy collection than a single pixel since it has a larger area making it proportionally faster. This affords it high temporal resolution as it measures the PSF at a specific location in the image field. Using multiple PSDs, a spatially variant PSF is generated and used to reconstruct images.

Superresolution and blind deconvolution of video

In many real applications traditional superresolution methods fail to provide high-resolution images due to objectionable blur and inaccurate registration of input low-resolution images. In this paper, we present a method of superresolution and blind deconvolution of video sequences and address problems of misregistration, local motion and change of illumination. The method processes the video by applying temporal windows, masking out regions of misregistration, and minimizing a regularized energy function with respect to the high-resolution frame and blurs, where regularization is carried out in both the image and blur domains. Experiments on real video sequences illustrate robustness of the method.

Blind restoration of images blurred by complex camera motion and simultaneous recovery of 3D scene structure

We present an algorithm that uses two or more images of the same scene blurred by camera motion for recovery of 3D scene structure and simultaneous restoration of sharp image. Motion blur is modeled by convolution with space-varying mask that changes its scale with the distance of imaged objects. The mask can be of arbitrary shape corresponding to the integral of the camera path during the pick-up time, which can be measured for instance by inertial sensors. This approach is more general than previously published algorithms that assumed shift-invariant blur or fixed, rectangular or Gaussian, mask shape. Algorithm can be easily parallelized and has a potential to be used in practical applications such as compensation of camera shake during long exposures

Restoration of retinal images with space-variant blur

Abstract. Retinal images are essential clinical resources for the diagnosis of retinopathy and many other ocular diseases. Because of improper acquisition conditions or inherent optical aberrations in the eye, the images are often degraded with blur. In many common cases, the blur varies across the field of view. Most image deblurring algorithms assume a space-invariant blur, which fails in the presence of space-variant (SV) blur. In this work, we propose an innovative strategy for the restoration of retinal images in which we consider the blur to be both unknown and SV. We model the blur by a linear operation interpreted as a convolution with a point-spread function (PSF) that changes with the position in the image. To achieve an artifact-free restoration, we propose a framework for a robust estimation of the SV PSF based on an eye-domain knowledge strategy. The restoration method was tested on artificially and naturally degraded retinal images. The results show an important enhancement, significant enough to leverage the images’ clinical use.

Fast Bayesian JPEG Decompression and Denoising With Tight Frame Priors

JPEG decompression can be understood as an image reconstruction problem similar to denoising or deconvolution. Such problems can be solved within the Bayesian maximum a posteriori probability framework by iterative optimization algorithms. Prior knowledge about an image is usually described by the