Reconstruction Of Blurred Images Research Articles

MP-NeRF: More refined deblurred neural radiance field for 3D reconstruction of blurred images

Neural Radiance Fields (NeRF) has gained prominence in the domain of 3D reconstruction. Despite its popularity, NeRF algorithm typically require clear, static images to function effectively, leading to reduced performance when dealing with real-world scenarios that present non-ideal conditions such as complex reflections, low dynamic range, dark scenes and blurriness resulting from camera motion or defocus photography. The resilience of NeRF to such blurred inputs has not been sufficiently examined, presenting a gap in current research, which also tends to neglect the role of 3D scene context in image deblurring. Addressing these challenges, we introduce Multi-branch Fusion Network and Prior-based Learnable Weights NeRF (MP-NeRF), a novel approach engineered for the precise reconstruction of scenes from blurred images. MP-NeRF innovatively applies distinct priors for various blur types, thus better interpreting the blur formation mechanism. It integrates a Multi-branch Fusion Network (MBFNet) with a Prior-based Learnable Weights (PLW), which collectively enhance the capture of complex scene details, including textural and pattern information. Our experimental findings validate that MP-NeRF considerably enhances the visual accuracy of reconstructions achieved through NeRF, yielding substantial improvements in Peak Signal-to-Noise Ratio (PSNR), Structural Similarity Index Measure (SSIM), and Learned Perceptual Image Patch Similarity (LPIPS) compared to state-of-the-art models. The source code for MP-NeRF is made available at the following URL: https://github.com/luckhui0505/MP-NeRF.

Sparse deep predictive coding captures contour integration capabilities of the early visual system.

Both neurophysiological and psychophysical experiments have pointed out the crucial role of recurrent and feedback connections to process context-dependent information in the early visual cortex. While numerous models have accounted for feedback effects at either neural or representational level, none of them were able to bind those two levels of analysis. Is it possible to describe feedback effects at both levels using the same model? We answer this question by combining Predictive Coding (PC) and Sparse Coding (SC) into a hierarchical and convolutional framework applied to realistic problems. In the Sparse Deep Predictive Coding (SDPC) model, the SC component models the internal recurrent processing within each layer, and the PC component describes the interactions between layers using feedforward and feedback connections. Here, we train a 2-layered SDPC on two different databases of images, and we interpret it as a model of the early visual system (V1 & V2). We first demonstrate that once the training has converged, SDPC exhibits oriented and localized receptive fields in V1 and more complex features in V2. Second, we analyze the effects of feedback on the neural organization beyond the classical receptive field of V1 neurons using interaction maps. These maps are similar to association fields and reflect the Gestalt principle of good continuation. We demonstrate that feedback signals reorganize interaction maps and modulate neural activity to promote contour integration. Third, we demonstrate at the representational level that the SDPC feedback connections are able to overcome noise in input images. Therefore, the SDPC captures the association field principle at the neural level which results in a better reconstruction of blurred images at the representational level.

Deconvolved Image Restoration From Auto-Correlations.

The recovery of a real signal from its auto-correlation is a wide-spread problem in computational imaging, and it is equivalent to retrieve the phase linked to a given Fourier modulus. Image-deconvolution, on the other hand, is a funda- mental aspect to take into account when we aim at increasing the resolution of blurred signals. These problems are addressed separately in a large number of experimental situations, ranging from adaptive astronomy to optical microscopy. Here, instead, we tackle both at the same time, performing auto-correlation inversion while deconvolving the current object estimation. To this end, we propose a method based on I -divergence optimization, turning our formalism into an iterative scheme inspired by Bayesian-based approaches. We demonstrate the method by recovering sharp signals from blurred auto-correlations, regardless of whether the blurring acts in auto-correlation, object, or Fourier domain.

Total Variation Based Parameter-Free Model for Impulse Noise Removal

AbstractWe propose a new two-phase method for reconstruction of blurred images corrupted by impulse noise. In the first phase, we use a noise detector to identify the pixels that are contaminated by noise, and then, in the second phase, we reconstruct the noisy pixels by solving an equality constrained total variation minimization problem that preserves the exact values of the noise-free pixels. For images that are only corrupted by impulse noise (i.e., not blurred) we apply the semismooth Newton's method to a reduced problem, and if the images are also blurred, we solve the equality constrained reconstruction problem using a first-order primal-dual algorithm. The proposed model improves the computational efficiency (in the denoising case) and has the advantage of being regularization parameter-free. Our numerical results suggest that the method is competitive in terms of its restoration capabilities with respect to the other two-phase methods.

Computational multi-focus imaging combining sparse model with color dependent phase mask.

A method for extended depth of field imaging based on image acquisition through a thin binary phase plate followed by fast automatic computational post-processing is presented. By placing a wavelength dependent optical mask inside the pupil of a conventional camera lens, one acquires a unique response for each of the three main color channels, which adds valuable information that allows blind reconstruction of blurred images without the need of an iterative search process for estimating the blurring kernel. The presented simulation as well as capture of a real life scene show how acquiring a one-shot image focused at a single plane, enable generating a de-blurred scene over an extended range in space.

Image deblurring process based on separable restoration methods

This paper proposes a method for reconstruction of blurred images damaged by a separable motion blur. The method can be used after the application of currently developed image restoration algorithms. Our approach is based on the usage of least squares solutions of certain matrix equations which define the separable motion blur. The method uses appropriately selected matrices besides the Moore–Penrose inverse. The method is tested by reconstructing a set of images after the removal of blur caused by uniform and separable motion. The quality of the restoration is observable by a human eye. The measurements such as the Improvement in Signal to Noise Ratio and the Peak in Signal to Noise Ratio have been increased significantly in comparison with the classical image restoration methods as well as the image restoration proposals based on the usage of the Moore–Penrose inverse.

Application of the pseudoinverse computation in reconstruction of blurred images

We present a direct method for removing uniform linear motion blur from images. The method is based on a straightforward construction of the Moore-Penrose inverse of the blurring matrix for a given mathematical model. The computational load of the method is decreased significantly with respect to other competitive methods, while the resolution of the restored images remains at a very high level. The method is implemented in the programming package MATLand respective numerical examples are presented.

Image deconvolution and coded masks in neutron radiography

Nearly all neutron radiography facilities use a setup similar to a single pinhole camera. With such facilities, a compromise between high flux at the detector plane and high spatial resolution has to be made: high resolution can be obtained with a small aperture diameter D and a large distance L between aperture and object. As flux is proportional to ( D/ L) 2, high resolution affords a long exposure time and yields a bad signal to noise ratio. Small L/ D ratios provide a low noise level but the resulting image will be blurred. Investigations were carried out whether this situation can be improved by reconstruction of blurred images that were recorded with high fluence. Several image deconvolution algorithms were tested for this purpose. It is shown that the application of those algorithms in neutron radiography is useful only in some special cases: for a given setup the resolution can be increased considerably in some cases. But unless there are restrictions for choosing L and D freely, there is always a combination of L and D that provides a better resolution without algorithmic reconstruction. The quality of reconstructed images strongly depends on the point spread function of the optical system. The point spread function is an issue of the aperture layout. Therefore the performances of different aperture shapes were simulated by Monte Carlo method. It is shown that optimized apertures (like multiple hole masks) improve considerably the quality of reconstructed images compared to a single hole system.

A Kalman algorithm for reconstruction of blurred radar images

A new Kalman algorithm is suggested to remove the blurring of radar image of surface, when the image is formed by an airborne radar in the high-resolution mode of operation.

Reconstruction of blurred images by controlled formation of spatial solitons

Diffracted digital images are reconstructed by excitation of spatial solitons in a bulk chi((2)) crystal, in the regime of single-pass parametric amplification of the quantum noise.