Proposed Image Restoration Research Articles

Image restoration based on vector quantization for robotic automatic welding

For the purpose of robot automatic welding based on a laser stripe welding seam tracking system, a weld noise image restoration algorithm based on vector quantization and synthetic fractal noise is proposed to solve the problems caused by strong noise interference and limited noise dataset in welding seam tracking. The weld noise image restoration is a process of noise loss. To achieve this process, a weld noise image restoration learning model based on feature encoding, vector quantization, and feature decoding is constructed, which includes feature encoding to extract image features, vector quantization to produce noise loss, and feature decoding to complete image restoration. To enhance the generalization ability of the weld noise image restoration model and overcome the difficulty of limited data collection in on-site welding, a synthetic welding noise model is proposed based on fractal theory, multidimensional Gaussian distribution, and random region generation algorithm. A large-scale training dataset is generated by randomly initializing parameters, and the potential mapping relationship between the weld noise image and the noise-free image is constructed. Compared with the limited dataset obtained in on-site welding, the synthetic dataset makes the image restoration model more generalizable, and the similarity between the restored welding seam image and the original image reaches 0.85. In the welding seam tracking experiment, for multiple sets of different data, the average tracking accuracy of the welding seam tracking algorithm based on image restoration is 0.2 mm, which verifies that the proposed image restoration algorithm can improve the robustness of automatic welding.

Raindrop Removal With Light Field Image Using Image Inpainting

In this paper, we propose a method that removes raindrops with light field image using image inpainting. We first use the depth map generated from light field image to detect raindrop regions which are then expressed as a binary mask. The original image with raindrops is improved by refocusing on the far regions and filtering by a high-pass filter. With the binary mask and the enhanced image, image inpainting is then utilized to eliminate raindrops from the original image. We compare pre-trained models of several deep learning based image inpainting methods. A light field raindrop dataset is released to verify our method. Image quality analysis is performed to evaluate the proposed image restoration method. The recovered images are further applied to object detection and visual localization tasks.

Morphological Component Image Restoration by Employing Bregmanized Sparse Regularization and Anisotropic Total Variation

Image deblurring is a fundamental problem in imaging field which often needs to recover the important structure of images. This paper addresses the image deblurring problem by considering an image as a combination of its cartoon (the piecewise smooth part of the image) and texture (the oscillation part of the image) components. To recover both of these parts, we propose the use of coupled analysis-based sparse representations to regularize the cartoon structure and the texture part of the image. We apply anisotropic total variation with a quadratic term to enhance the edges existing in the cartoon part. Furthermore, we develop a multivariable Bregman optimization method to solve the proposed image restoration model by combining the alternating minimization method and the split Bregman iteration. The experiments show that the proposed algorithm not only performs well for image decomposition, but also outperforms the previously established methods in terms of the visual residual error, the structure similarity index and the peak signal-to-noise ratio for image deblurring.

Image-Based Artefact Removal in Laser Scanning Microscopy.

Recent developments in laser scanning microscopy have greatly extended its applicability in cancer imaging beyond the visualization of complex biology, and opened up the possibility of quantitative analysis of inherently dynamic biological processes. However, the physics of image acquisition intrinsically means that image quality is subject to a tradeoff between a number of imaging parameters, including resolution, signal-to-noise ratio, and acquisition speed. We address the problem of geometric distortion, in particular, jaggedness artefacts that are caused by the variable motion of the microscope laser, by using a combination of image processing techniques. Image restoration methods have already shown great potential for post-acquisition image analysis. The performance of our proposed image restoration technique was first quantitatively evaluated using phantom data with different textures, and then qualitatively assessed using in vivo biological imaging data. In both cases, the presented method, comprising a combination of image registration and filtering, is demonstrated to have substantial improvement over state-of-the-art microscopy acquisition methods.

Motion Blur Image Restoration Algorithm Based on the Theory of Reaction Diffusion

Image restoration is of great significance in the digital image processing. When imaging equipment obtains the image, if the imaging equipment and the scene occur relative motion in the exposure time, it could lead to actually get blurred image. Due to motion blur, the got image loses real value. So the research of motion blur image restoration has important application value. This paper proposes a motion blur image restoration algorithm based on the theory of reaction diffusion. The algorithm uses the reaction diffusion equations related mathematical knowledge as the background, build the motion blur image restoration model and design motion blur image restoration algorithm. Through the experiment, the algorithm compares with the common motion blur restoration algorithm (RI-BM3D, BM3DDEB, IDD-BM3D and ForWaRD). The subjective and objective restoration results of the proposed image restoration algorithm are obviously improved.

Image denoising using combined higher order non-convex total variation with overlapping group sparsity

It is widely known that the total variation image restoration suffers from the stair casing artifacts which results in blocky restored images. In this paper, we address this problem by proposing a combined non-convex higher order total variation with overlapping group sparse regularizer. The hybrid scheme of both the overlapping group sparse and the non-convex higher order total variation for blocky artifact removal is complementary. The overlapping group sparse term tends to smoothen out blockiness in the restored image more globally, while the non-convex higher order term tends to smoothen parts that are more local to texture while preserving sharp edges. To solve the proposed image restoration model, we develop an iteratively re-weighted $$\ell _1$$ based alternating direction method of multipliers algorithm to deal with the constraints and subproblems. In this study, the images are degraded with different levels of Gaussian noise. A comparative analysis of the proposed method with the overlapping group sparse total variation, the Lysaker, Lundervold and Tai model, the total generalized variation and the non-convex higher order total variation, was carried out for image denoising. The results in terms of peak signal-to-noise ratio and structure similarity index measure show that the proposed method gave better performance than the compared algorithms.

Restoration of longitudinal laser tomography target image from inhomogeneous medium degradation under common conditions.

In order to overcome the shortages of the target image restoration method for longitudinal laser tomography using self-calibration, a more general restoration method through backscattering medium images associated with prior parameters is developed for common conditions. The system parameters are extracted from pre-calibration, and the LIDAR ratio is estimated according to the medium types. Assisted by these prior parameters, the degradation caused by inhomogeneous turbid media can be established with the backscattering medium images, which can further be used for removal of the interferences of turbid media. The results of simulations and experiments demonstrate that the proposed image restoration method can effectively eliminate the inhomogeneous interferences of turbid media and achieve exactly the reflectivity distribution of targets behind inhomogeneous turbid media. Furthermore, the restoration method can work beyond the limitation of the previous method that only works well under the conditions of localized turbid attenuations and some types of targets with fairly uniform reflectivity distributions.

Image Restoration Based on Statistic PSF Modeling for Improving the Astrometry of Space Debris

Space debris is a special kind of fast-moving, near-Earth objects, and it is also considered to be an interesting topic in time-domain astronomy. Optical survey is the main technique for observing space debris, which contributes much to the studies of space environment. However, due to the motion of registered objects, image degradation is critical in optical space debris observations, as it affects the efficiency of data reduction and lowers the precision of astrometry. Therefore, the image restoration in the form of deconvolution can be applied to improve the data quality and reduction accuracy. To promote the image processing and optimize the reduction, the image degradation across the field of view is modeled statistically with principal component analysis and the efficient mean point-spread function (PSF) is derived from raw images, which is further used in the image restoration. To test the efficiency and reliability, trial observations were made for both low-Earth orbital and high-Earth orbital objects. The positions of all targets were measured using our novel approach and compared with the reference positions. The performance of image restoration employing our estimated PSF was compared with several competitive approaches. The proposed image restoration outperformed the others so that the influence of image degradation was distinctly reduced, which resulted in a higher signal-to-noise ratio and more precise astrometric measurements.

Nonconvex-TV Based Image Restoration with Impulse Noise Removal

In this paper, we propose and study a nonconvex data fitting term and a total variation regularization term for image restoration with impulse noise removal. The proposed model is different from existing image restoration models where the data fitting term is based on the $\ell_1$- or $\ell_2$- norm, and the regularization term is based on the total variation, $\ell_1$-norm, or some nonconvex functions. Theoretically, we analyze the properties of minimizers of the proposed objective function with the nonconvex data fitting term and the total variation regularization term. We show that minimizers can preserve piecewise constant regions or match with the data points perfectly. This property is particularly useful for impulse noise removal. The proposed image restoration model can be solved by the proximal linearized minimization algorithm, and the global convergence of the iterative algorithm can also be established according to Kurdyka--Łojasiewicz property. The performance of the proposed model is tested ...

Non-Local Regularized Variational Model for Image Deblurring under Mixed Gaussian-Impulse Noise

The presence of mixed Gaussian-impulse noise makes image deblurring much more challenging in real-world applications. To guarantee high-quality deblurred images, a non-local regularized variational model is proposed in this paper for restoration of blurred images corrupted by mixed Gaussian-impulse noise. The proposed image restoration model is mainly composed of a mixed L^(1,2) data-fidelity term and a non-local total variation regularizer. The nonlocal regularizer is capable of suppressing staircase-like artifacts and preserving fine structures since it takes full advantage of high degree of self-similarity and redundancy within images. However, the mixed data-fidelity term and non-local regularizer make the image deblurring problem arithmetically difficult to solve. To guarantee solution stability and efficiency, an alternating minimization algorithm is developed to solve the resulting optimization problem. In particular, the original optimization problem can be decomposed into two simpler subproblems. Each subproblem is then efficiently solved using existing numerical method. Comprehensive experiments on grayscale and color test images have been carried out to compare our proposed method with several state-of the- art image restoration methods. Experimental results demonstrated that our proposed method was more capable of removing both blurring effect and mixed Gaussian-impulse noise, as well as preserving fine image details.

A novel image restoration scheme based on structured side information and its application to image watermarking

This paper presents a new image restoration method based on a linear optimization model which restores part of the image from structured side information (SSI). The SSI can be transmitted to the receiver or embedded into the image itself by a digital watermarking technique. In this paper we focus on a special type of SSI for digital watermarking where the SSI is composed of mean values of 4×4 image blocks which can be used to restore manipulated blocks. Different from existing image restoration methods for similar types of SSI, the proposed method minimizes image discontinuity according to a relaxed definition of smoothness based on a 3×3 averaging filter of four adjacent pixel value differences, and the objective function of the optimization model has a second regularization term corresponding to a second-order smoothness criterion. Our experiments on 100 test images showed that given complete information of the SSI, the proposed image restoration technique can outperform the state-of-the-art model based on a simple linear optimization model by around 2dB in terms of an average Peak Signal-to-Noise Ratio (PSNR) value and around 0.04 in terms of a Structural Similarity Index (SSIM) value. We also tested the robustness of the image restoration method when it is applied to a self-restoration watermarking scheme and the experimental results showed that it is moderately robust to errors in SSI (which is embedded as a watermark) caused by JPEG compression (the average PSNR value remains above 16.5dB even when the JPEG QF is 50), additive Gaussian white noises (the average PSNR value is approximately 18.4dB when the noise variance σ2 is 10) and image rescaling assuming the original image size is known at the receiver side (e.g. the average PSNR value is approximately 19.6dB when the scaling ratio is 1.4).

A nonmonotone adaptive projected gradient method for primal-dual total variation image restoration

This paper studies image restoration problems from noisy and blurred observation. Based on a primal-dual total variation model, a nonmonotone adaptive projected gradient method is proposed and tested. By introducing an auxiliary variable, the proposed method implements image restoration by de-blurring and de-noising alternatively at each iteration. Convergence result of the proposed method is established. Numerical results illustrate the efficiency of this method and indicate that it is competitive to some state-of-the-art algorithms in the literature, such as FISTA and FTVd.

Iterative image restoration using nonstationary priors

In this paper, we propose an algorithm for image restoration based on fusing nonstationary edge-preserving priors. We develop a Bayesian modeling followed by an evidence approximation inference approach for deriving the analytic foundations of the proposed restoration method. Through a series of approximations, the final implementation of the proposed image restoration algorithm is iterative and takes advantage of the Fourier domain. Simulation results over a variety of blurred and noisy standard test images indicate that the presented method comfortably surpasses the current state-of-the-art image restoration for compactly supported degradations. We finally present experimental results by digitally refocusing images captured with controlled defocus, successfully confirming the ability of the proposed restoration algorithm in recovering extra features and rich details, while still preserving edges.

Development of Camera-Based Measurement System for Crane Spreader Position using Foggy-degraded Image Restoration Technique

In this paper, a foggy-degraded image restoration technique with a physics-based degradation model is proposed for the measurement system. When the degradation model is used for the image restoration, its parameters and a distance from the spreader to the camera have to be previously known. In the proposed image restoration technique, the parameters are estimated from variances and averages of intensities on two foggy-degraded landmark images taken at different distances. Foggy-degraded images can be restored with the estimated parameters and the distance measured by the measurement system. On the basis of the experimental results, the performance of the proposed foggy-degraded image restoration technique was verified.

An Adaptive Variational PDE Based Image Restoration Model Using Hopfield Neural Network

An adaptive variational partial differential equation (PDE) based aproach for restoration of gray level images degraded by a known shift-invariant blur function and additive noise is presented. The restoration problem of a degraded image is solved by minimizing this model, and this minimizing problem is realized by using Hopfield neural network. In the proposed image restoration model, an adaptive regularization parameter is developed instead of the constant regularization parameter used in previous PDE model. The value of the adaptive regularization parameter changes according to different regions of the image to remove noises and preserve edge better. Several computer simulation results show that the image restoration results of the proposed model both look better and have better SNR (Signal to Noise Ratio) than the previous variational PDE based model.

Digital Image Restoration Based on Pixel Simultaneous Detection Probabilities

Here an image restoration on the basis of pixel simultaneous detection probabilities (PSDP) is proposed. These probabilities can be precisely determined by means of correlations measurement [NIMA 586 (2008) 314-326]. The proposed image restoration is based on the solution of matrix equation. Non-zero elements of Toeplitz block matrix with ones on the main diagonal, is determined using PSDP. The number of non zero descending diagonals depends on the detector construction and is not always smaller than 8. To solve the matrix equation, the Gaussian elimination algorithm is used. The proposed restoration algorithm is studied by means of the simulated images (with and without additive noise using PSDP for General Electric Senographe 2000D mammography device detector) and a small area (160 times 160 pixels) of real images acquired by the above mentioned device. The estimation errors of PSDP and the additive noise magnitude permits to restore images with the precision better than 3% for the above mentioned detector. The additive noise in the real image is present after restoration and almost has the same magnitude. In the restored small area (16 times 16 mm) of real images, the pixel responses are not correlated. The spatial resolution improvement is also analyzed by the image of an absorber edge.

Image deconvolution in digital autoradiography: A preliminary study

Digital autoradiography (DAR) is a powerful method to determine quantitatively the "small-scale" (i.e., submillimeter) distribution of a radiotracer within a tissue section. However, the limited spatial resolution of the DAR image, due to blurring by the point spread function (PSF), can result in a poor correlation with tissue histology and immunohistochemistry. The authors attempt to overcome this limitation by recovering the radiotracer distribution by image deconvolution using the Richardson-Lucy algorithm and a measured PSF obtained from a small radioactive source on hydrophobic microscope slide. Simulation studies have shown that the deconvolution algorithm reliably recovers the pixel values corresponding to the radioactivity distributions. As an example, the proposed image restoration approach has been tested with DAR images of different radiolabeled markers on tumor sections obtained from clinical and preclinical animal model studies. Digital autoradiograms following deconvolution show improved sharpness and contrast relative to the unprocessed autoradiograms.