Photon-Limited Blind Deconvolution Using Unsupervised Iterative Kernel Estimation

The detection of rolling bearing faults is essential to ensure the operational safety of rotating machinery. An effective method for diagnosing rolling bearing faults is the maximum second-order cyclostationarity blind deconvolution (CYCBD) method, which has the ability to extract weak fault periodic pulse features. However, blind deconvolution methods such as CYCBD often fail when composite bearing faults occur in the presence of strong background noise. To overcome this issue, an adaptive blind deconvolution (ABDD) method based on improved CYCBD is proposed. In this method, a finite impulse response (FIR) filter bank is constructed to cover the entire frequency band of the signal, enabling fault frequency segmentation. The improved CYCBD is then used to obtain a filter pattern that locks on the fault frequency. These filter patterns are arranged in descending order based on the correlation kurtosis (CK). The required filtering mode is selected according to the sorting mode, and a spectrum analysis is performed to extract fault features of single and compound faults in the bearing. Simulation and experimental results demonstrate that ABDD effectively extracts bearing composite fault features and outperforms classic feature decomposition and blind deconvolution methods in extracting single and composite fault features.

The main inherent property of medical ultrasound imaging is speckle noise which generally obscures and reduces the diagnostic image resolution and contrast. Consequently, the substantial improvement of ultrasound images is an important prerequisite, whenever ultrasound is used as one of the most utilized diagnostic modalities. The problem of reconstruction of ultrasound images by means of blind deconvolution has long been recognized as a prerequisite in ultrasound imaging processing in the recent decades. Recently, the Total Variation (TV) regularization method has become extremely effective approach for image reconstruction, especially for restoring edges of the blurring image. In this paper, we present a new blind iterative TV deconvolution algorithm for reconstructing ultrasound images from blurry and noisy observations. First, it proposes the initial estimation of the point-spread function (PSF) based on a generalization of two-dimensional homomorphic filtering in cepstrum domain. It is demonstrated that the initial PSF can be effectively estimated by applying a proper smoothing low-pass filtering in cepstrum domain. Second, it introduces a novel blind iterative TV deconvolution which is derived from an alternating minimization algorithm. Fast Fourier Transform (FFT) is used in the pre-iteration computation. The innovative blind deconvolution is based on either concurrent or successive estimation of the PSF function and the image of interest. The iterative scheme is devised to recover the image and simultaneously identify the PSF function. The estimated PSF and restored image will be close to real values in the subsequent iterative deconvolution. Final, tests on phantom and clinical images have proven our novel blind iterative TV deconvolution gives more positive results of higher spatial resolution and better defined tissue structures than other deconvolution methods.

A new blind deconvolution method with additional phase spectral constraints for the blurred image is discussed in this paper. The blurred image is a convolution of an original image and a point-spread function (PSF). The proposed method is based on the following iterative procedures, i.e., a projection onto a frequency space satisfying phase spectrum constraints, a minimization of a cost function and a projection onto an image space satisfying the support and nonnegativity constraints. This method restores the original image and the PSF stably with high accuracy. The effectiveness of the proposed method is verified by applying it to some blind deconvolution problems of images. The experimental results show that the performance of the proposed method is drastically improved in comparison with the conventional blind deconvolution method.

In this paper, a method for MIMO blind deconvolution is proposed. The method is applicable to the case of non-Gaussian i.i.d. input signals. The core of the method is an algorithm which extracts one of the input signals from the information of the output signals only. This algorithm is an non-trivial generalization of the Shalvi-Weinstein algorithm for SISO blind deconvolution and converges in super-exponential rate possibly after finite iterations. By recursive use of this algorithm, extraction of all input signals can be achieved.

Spectrum extraction which is applied to extract one-dimensional (1D) astronomical spectroscopy from twodimensional (2D) multi-fiber spectrum images plays a key role in data processing of 2D optical fiber spectrum. In this paper, we have given a detailed comparison of spectrum extraction methods between the blind deconvolution and the non-blind deconvolution to find out their specialities. The simulated arc spectrum data is used in Comparative experiments. The results show that the blind deconvolution is better than the non-blind deconvolution not only at spectrum extraction result but also remarkably at signal noise ratio (SNR).

We propose a new method of blind deconvolution (BD) suitable for the recovery of truncated blurred images. In practical cases we often encounter truncated images, but they are difficult to treat properly with conventional BD algorithms, since nontruncated blurred images are assumed. The proposed method is based on simulated annealing (SA), to which we introduce an adaptive masking process. After a truncated segment is cut out from the blurred image, the sharp edges surrounding the segment are removed by multiplying the initial window, whose shape is determined with the initial estimate of the point spread function (PSF). A more appropriate estimate for the PSF is obtained from the masked segment based on the SA algorithm, and according to the new estimate of the PSF, the shape of the window is varied. These procedures are iterated until we finally obtain the best possible estimate of the PSF. In the last step, the entire blurred image is deconvolved with the final estimate of the PSF to obtain the recovered image. The effectiveness of the proposed method is confirmed by computer simulation.

Total variation minimizing blind deconvolution with shock filter reference

Single-image blind deconvolution is one of the most challenging fields in image processing which restores a sharp image from its blurred version. Nowadays blind deconvolution algorithms have made significant progress. However, the restoration of blurred images with little scale edges and periodic textures is still a hard work. To solve this problem, this paper proposes a new normalized sparse regularization blind deconvolution algorithm, which uses a gradient saliency map to prohibit the image small structures on image blurry kernel estimation. Firstly, salient detection is performed to select the important area which conforms with the human vision system and generates a binary mask to screen out useful gradients. Secondly, the normalized sparse regularization blind deconvolution method is applied to obtain accurate blur kernel and recover the sharp image. Finally, the experiment results show that the algorithm can effectively deblur the degraded image on different scenarios.

We use blind deconvolution methods in optical diffusion tomography to reconstruct images of objects imbedded in or located behind turbid media from continuous-wave measurements of the scattered light transmitted through the media. In particular, we use a blind deconvolution imaging algorithm to determine both a deblurred image of the object and the depth of the object inside the turbid medium. Preliminary results indicate that blind deconvolution produces better reconstructions than can be obtained using backpropagation techniques. Moreover, it does so without requiring prior knowledge of the characteristics of the turbid medium or of what the blur-free target should look like: important advances over backpropagation.

When imaging through layered media such as walls, the contents and thickness of the wall layers are generally not known a priori. Furthermore, compensating for their effects can be computationally intensive, as this generally requires modelling the transmission and reflection of complex fields through layered media. We propose a blind deconvolution method that does not require knowledge of the wall layers by directly estimating a circularly symmetric Green’s function that models the transmission through the wall layers, simultaneously addressing both problems. We experimentally demonstrate this technique by measuring the reflection through a multilayered structure of building materials at the K-band microwave frequencies, and using the blind deconvolution method to find the image of a reflective object behind the layers.

Detection of an incipient internal defect from a measurement point on the surface of a machine is often made difficult due to the corruption of the source signal by background noise usually through the transmission path. In order to detect or recover the source signal, it is essential to suppress or remove this background noise. With the advancement of digital signal processing techniques, this removal can be achieved using fixed or adaptive filters. The main problem with the application of filtering techniques is that one has to know the characteristics of the noise in advance or may require more than two channels of data if using an adaptive filter. This paper reports on the application of Blind Deconvolution method to recover fault information from the measured (observed) signals of a damaged bearing through the transmission channel and which was also corrupted by noise. The Blind Deconvolution method manages to recover the fault signal by estimating the inverse of the transmission (channel transfer function) path using the eigenvector algorithm (EVA). Modified crest factor (MCF) and Arithmetic Mean (AM) algorithms were used to optimise the equaliser parameters. Computer simulation and experimental studies were used to verify the applicability of the technique.

The present paper illustrates a geodesic-based learning algorithm over a curved parameter space for blind deconvolution application. The chosen deconvolving structure appears as a single neuron model whose learning rule arises from criterion-function minimization over a smooth manifold. In particular, we propose here a learning stepsize selection theory for the algorithm at hand. We consider the blind deconvolution performances of the algorithm as well as its computational burden. Also, a numerical comparison with seven blind-deconvolution algorithms known from the scientific literature is illustrated and discussed. Results of numerical tests conducted on a noiseless as well as a noisy system will confirm that the algorithm discussed in the present paper performs in a satisfactory way. Also, the performances of the presented algorithm will be compared with those exhibited by other blind deconvolution algorithms known from the literature.

Deconvolution is essential for high-resolution seismic data processing. Conventional deconvolution methods are either based on a stationary convolution model or under the assumption that Q factor and the source wavelet are known. However, in reality, seismic wavelet is usually unknown and time-varying during propagation due to attenuation. Thus, we propose a blind nonstationary deconvolution (BND) method which does not require advance Q factor and source wavelet as inputs and takes into account the lateral continuity of deconvolution results. Firstly, we develop an improved nonstationary convolution model consisting of the time-varying wavelet and reflectivity, which enables us to obtain reflectivity without attenuation estimation. To accommodate the changing frequency spectrum of seismic data, we present a time-varying wavelet estimation method using the frequency spectrum at every sample point and the generalised seismic wavelet function. By incorporating the extracted time-varying wavelet into the improved convolution model, we propose to formulate the objective function for reflectivity inversion as a joint low-rank and sparse inversion convex optimisation problem. It helps deconvolution results keep the sparsity in the vertical direction while maintaining the continuity in the horizontal direction. The performance of BND is evaluated through synthetic examples and a field data example.

An acoustic space is uniquely characterized by its reverberant behavior. Due to the complexity of the multiple reflections and diffraction of sound in an enclosure, currently available system identification algorithms cannot effectively estimate the impulse response of a concert hall simply from musical recordings in that space. This paper reports the use of blind image deconvolution methods to construct echo patterns from quadratic time-frequency representations of reverberant recordings of music. First, a quadratic time-frequency analysis is performed to decompose the musical signal into its constituent harmonic components. Quadratic time-frequency analysis methods are known to give enhanced resolution in the time-frequency plane in comparison to conventional Fourier analysis. Reverberant features then appear as blur in the time-frequency plane, which can be estimated by employing the methods of blind deconvolution of this “image” of the sound. The proposed algorithm retrieves both the blur pattern, which corresponds to the reverberation echogram, and the direct acoustic signal. By choosing the time-frequency frame scale and smoothing window, a multi-resolution analysis of the underlying acoustic impulse response is obtained. Various quadratic time-frequency analysis methods are evaluated and their relative performance is reported. The proposed methods are also compared to existing dereverberation algorithms.

A weighted total variation algorithm was proposed for solving some blind deconvolution problems,and then we got the existence of energy functional minimum and educed corresponding thermal flow.In computer simulations,the algorithm performs well even when we have no more accurate prior knowledge about point spread function and the original true image,besides some other constrains.Experiments show that the proposed algorithm is more effective and performs better than the total variation blind deconvolution method proposed.