ABSTRACT To improve the resolution and clarity of low-light images, this paper presents a super-resolution reconstruction algorithm based on enhanced feature maps. The algorithm employs a generalized total variation model for image denoising and an improved Butterworth high-pass filter to extract high-frequency components. Weighted guided filtering is used to enhance illumination and preserve edge details, while nonlinear stretching improves saturation and contrast. A color recovery and edge-preservation mechanism based on the YCbCr color space and ER-ANR method is introduced to refine reflection components. Experimental results show that the proposed method effectively accomplishes super-resolution reconstruction with abundant details, clear visualization, and superior overall image quality.

High-quality infrared images are widely used in various vision tasks. However, infrared focal plane array (IRFPA) suffer from material and process limitations, resulting in non-uniform pixel response and severe spatial fixed pattern noise (FPN), which results in degraded images. In order to address the problem that existing infrared image nonuniformity correction (NUC) methods only focus on stripe noise removal but ignore the environmental noise, we propose a new multi-stage single-frame NUC processing pipeline. Firstly, for the low-rank property of infrared degraded images and the higher-order directional priori information of stripe noise, we construct an objective function which combines a higher-order gradient total variation model with a diagonal kernel paradigm as a constraint term. Multiple subproblems are solved by the alternating direction method of multipliers (ADMM) to obtain the recovered image after removal of stripe. Then, we incorporate the environmental noise into the pipeline by low-rank matrix approximationand employ singular-valued patch decomposition to efficiently separate the clean image from the noise. Extensive experiments are conducted on existing methods on real and simulated datasets, and the potential performance of the proposed method is verified in qualitative and quantitative evaluations. The code and datasets can be obtain at https://github.com/ImageVisioner/InfraredNUC.

A novel hybrid multi-regularization total variation model for edge-aware image smoothing

Maintaining the health and safety of the sun gears is critical because these provide an essential part of the transmission power in planetary gearboxes. Firstly, to reduce the influence of planetary motion that introduces phase delay into the fault vibration signals of sun gears, this paper applies the vibration separation method for preprocessing. Then, the higher-order nonconvex total variation model based on overlapping group sparsity theory is developed, and thus overcomes the recognition problem of information error caused by the traditional total variation model under noise interference. The studied model of this paper incorporates regularization terms from different perspectives; the overlapping group sparsity well characterizes the intercrossing relationship between adjacent elements, while the high-order nonconvex norm improves solution smoothness. Therefore, the above two regularization terms are introduced into the total variation model to effectively improve the noise reduction performance of the model and make the fault feature of sun gears more prominent than those of other components. In addition, the variable-splitting strategy is employed to iteratively solve each subproblem and obtain the optimal solution of the model. Finally, the cyclic autocorrelation function is adopted to further process the model output, leveraging the noise suppression capability of cyclic cumulants. The fault-related spectral lines of the sun gears are more pronounced than other interference components. The studied method analyzes analysis of fault test vibration data demonstrates that the studied method effectively addresses the fault diagnosis problem of sun gears.

In this paper, a novel variational model is proposed for image segmentation via joint restoration of images corrupted by blurring and Rician noise. The proposed model is built upon the piecewise constant Mumford–Shah framework and combines an appropriate data fidelity term with nonconvex total variation (NTV) regularization. The NTV regularization effectively denoises homogeneous regions while accurately preserving object boundaries to facilitate robust segmentation. To solve the resulting nonconvex optimization problem, a proximal alternating minimization algorithm is employed. In addition, an iteratively reweighted $ \ell_1 $ algorithm and the alternating direction method of multipliers are adopted to efficiently handle the corresponding subproblems. Numerical experiments demonstrate the effectiveness of the proposed model in achieving accurate and robust segmentation performance when compared with several state-of-the-art methods.

Planetary gearbox is a crucial variable-speed transmission component in wind turbines, and the solar wheel is prone to local faults under harsh environmental conditions. Firstly, to avoid the influence of time-varying transmission paths, this paper preprocesses the fault vibration signal of solar wheel using window function and further transforms it to the angular domain.. Then, a high-order Vese-Osher model based on partial differential equations is subsequently developed to effectively identify the fault features of solar wheel. By introducing the mathematical expression of second derivative, this approach avoids the over-smoothing problem associated with the first derivative. While the Euler equation is commonly used to solve the total variation model, it cannot handle the G-norm, therefore, the divergence norm is considered to reformulate the model. Additionally, two weighting parameters are introduced between different regularization terms to optimize the solution, enhancing the ability of model to suppress noise in the fault-windowed vibration signal of solar wheel in the angular domain and thereby making the fault spectral lines more distinguishable. The Split Bregman algorithm is adopted to avoid directly solving high-order nonlinear partial differential equations. Furthermore, the method leverages the fast Fourier transform and generalized soft-thresholding equation to decompose the global optimization problem into several subproblems. Since higher-order cumulant exhibit superior noise suppression, the diagonal slices of third-order cumulant for the model output are extracted, and further obtain the 1.5-dimensional demodulation spectrum. This makes the fault feature of solar wheel completely prominent. Finally, The robustness of the above method in fault diagnosis of solar wheel is verified by fault test data.

A novel weighted total variation model incorporating a fractional order variable exponent is proposed, designed to achieve adaptive smoothing in key image regions while effectively preserving edge information. Based on the properties of the log-exp function, an adaptive weighting function is introduced, assigning lower weights to edges and higher weights to smooth areas, thus enhancing structural detail retention and noise suppression. Through variational analysis, the corresponding Euler–Lagrange equation is derived, transforming the optimization problem into a gradient descent flow formulation amenable to efficient numerical solution. Extensive comparative experiments on three benchmark datasets demonstrate that the proposed approach yields substantial improvements over baseline methods and achieves performance competitive with state-of-the-art models.

The imprecise total variation model and its connections with game theory

Image denoising is a computer vision task that mainly aims to remove unwanted noise from a given image while preserving all necessary details and related information. Detection of edges and smooth image generation are required criteria for measuring the quality of the image denoiser. This paper introduces a new model for image denoising based on total variation (TV). The unprecedented novelty total variation (NTV) model combines norm-based total variation (TV) and norm-based TV regularization. This paper uses the implicit finite difference method to solve the NTV model numerically. The statistical measurements are used to compare the results obtained using the NTV model with those obtained using other models, which show the superiority of the proposed model in terms of its effectiveness and efficiency through removing different types of noise from images. This proposed model is effective in detail sharpening and texture preservation.

The advancement of high-quality camera technology has increased the demand for efficient video analysis methods. Current methods mostly rely on matrix-based approaches, which break data structures and lose some spatial information. This paper proposes a novel approach (TRLRTTV) that combines Low Rank Tensor Ring decomposition and Tensor Total Variation regularization for moving objects detection (MOD). For static background detection, the tensor ring (TR) decomposition is utilized to extract low rank information, and low rank assumption is placed on tensor factors instead of the original data. For moving objects, a tensor total variation model with l_{1/2} regularization is employed to ensure the representation of foreground information along the spatio-temporal direction. The results demonstrate that, compared to existing methods, the proposed algorithm achieves a 3%-8% performance improvement in background separation and the comprehensive performance metric f. Furthermore, the proposed method ensures robustness against noise interference, such as Gaussian and salt-and-pepper noise and more suitable for higher-dimensional video processing.

A non-monotone proximal point method for image reconstruction using non-convex total variation models

Image enhancement plays an important role in biometric systems, this paper presented automatic latent fingerprint segmentation and matching. While considerable progress has made in both rolled and plain fingerprint image enhancement, latent fingerprint enhancement is a challenging problem due to the poor image quality of latent fingerprint with unclear ridge structures and various overlapping patterns, along with the presence of structured noise. Prior to latent fingerprint segmentation and feature extraction, latent fingerprint image enhancement is important to suppress various types of noise and to clarify the ridge structure. This paper reviews the current techniques used for latent fingerprint enhancement and presents a hybrid model which combines the edge directional total variation model (EDTV) and quality image enhancement with lost minutia reconstruction. The NIST SD27 database is used to test the performance of the proposed techniques with RMSE and PSNR. The proposed technique is effectively clarify input latent fingerprint images and eliminate noise in good, bad and ugly latent fingerprint images. A statistically significant difference, which focused on the mean lengths of PSNR and RMSE for different categories of latent fingerprint, images (good, bad and ugly). The proposed technique performs well for the good latent fingerprint images compare to bad and ugly images. Enhancement respectively presents RMSE averages of 0.018373, 0.022287, and 0.023199 for the good, bad and ugly image SD27 image set, as opposed to 82.99068, 81.39749, and 81.07826 for PSNR. The proposed enhancement technique improved the matching accuracy of latent fingerprint images by about 30%.

Single-shot structured light illumination plays a crucial role in high-speed 3D reconstruction but remains a significant challenge, especially in the presence of complex surface textures. Fourier transform profilometry has been widely adopted for single-shot reconstruction, yet its performance is often degraded due to spectral leakage and frequency aliasing when applied to textured or irregular surfaces. To address these limitations, we propose a single-shot 3D reconstruction framework based on multiscale total generalized variation. The proposed approach integrates variational optimization with multi-frequency fringe modulation to enhance reconstruction accuracy. Specifically, a dual-frequency heterodyne fringe pattern is generated using a multifrequency modulation strategy. An improved adaptive orientation total generalized variation model with shared parameter constraints is then employed to accurately extract both high- and low-frequency fringe components. The final 3D shape is reconstructed through phase calculation using the dual-frequency heterodyne technique, which effectively suppresses spectral overlap. Experimental results conducted on highly textured surfaces demonstrate that our method achieves superior robustness and reconstruction quality compared to conventional approaches.

Planet-bearing plays a critical role in supporting planet gear and transmitting power, yet their fault-induced vibrations are inherently modulated by time-varying transmission paths, and thus it brings significant challenges to fault feature extraction. Therefore, this means that it is imperative to develop a robust fault analysis methodology. For one thing, this paper adopts minimum noise amplitude deconvolution (MNAD) to deal with the fault signal of planet-bearing inner ring. It is different from the traditional blind deconvolution method, mainly by suppressing the marked area noise components, which indirectly makes the fault feature prominent. For another, the first-order total variation model is prone to fall into the gradient effect problem. It takes into account the regular term of higher order total variational to achieve the purpose of minimizing approximation. In addition, two gradient weight parameters are introduced to balance the difference operator with different orders. For this reason, the generalized total variational denoising model (GTVDM) is constructed, and then the angular domain signal filtered by MNAD is input into this model. Meanwhile, it utilizes the Split Bregman algorithm to acquire the optimal solution of the constructed model. The core analysis process transforms an objective function optimization problem into several sub-problems, significantly improving computational efficiency. Finally, the experimental validation using fault test data of planet-bearing inner ring demonstrates that the combined MNAD algorithm and GTVDM framework effectively, and the fault feature order and modulation sideband caused by time-varying transmission path are observed in the squared envelope order spectrum.

ABSTRACT This letter introduces a novel approach for detecting oil films in multi-source remote sensing data acquisition experiments, which integrates the spectral-spatial joint difference index and deep learning architecture. The proposed method leverages the total variation model and normalized difference water index to quantitatively construct the normalized difference oil-water index. Subsequently, this index is combined with two-dimensional total variation image texture features through a designed oil film recognition network, enabling the extraction of spectral-spatial joint features from oil film remote sensing data. The proposed scheme demonstrates strong capability in effectively extracting contours for oil film detection, achieving impressive F1-scores of 85.93%, 89.16% and 91.51%, respectively, for oil film targets measuring 2 μ m , 4 μ m and 8 μ m . Importantly, these scores surpass the performance achieved by the current prevailing model.

Computed tomography (CT) is a widely used medical imaging modality, but its radiation exposure poses potential risks to human health. Sparse-view scanning has emerged as an effective approach to reduce radiation dose; however, images reconstructed using the filtered back-projection (FBP) algorithm from sparse-view projections often suffer from severe streak artifacts. Achieving high-quality CT image reconstructed from sparse-view projections remains a challenging task. Building on compressed sensing (CS), the total variation (TV) algorithm is applied for high-quality sparse-view reconstruction. We further propose a relative total variation (RTV) algorithm to enhance the accuracy of sparse-view reconstruction. Experimental results indicate that while the RTV algorithm improves accuracy, it has limitations in edge preservation. To address this, inspired by the success of directional TV (DTV) in limited-angle reconstruction, we develop a directional relative TV (DRTV) model. This model applies the RTV technique in both x and y directions independently, and we derive its adaptive steepest descent projection onto convex set (ASD-POCS) solution algorithm. Experiments conducted on simulated phantoms and real CT images demonstrate the correctness, convergence, and superior performance of the DRTV algorithm in sparse-view reconstruction. Compared with the TV, DTV, and RTV algorithm, the DRTV algorithm exhibits superior preservation of structural features and texture details. The DRTV algorithm represents an advanced method for high-precision sparse-view CT reconstruction, providing stable and accurate results. Moreover, the approach is applicable to other medical imaging modalities.

Adaptive optics (AO)-corrected retina flood illumination imaging technology is widely used for investigating both structural and functional aspects of the retina. Given the inherent low-contrast nature of original retinal images, it is necessary to perform image restoration. Total variation (TV) regularization is an efficient regularization technique for AO retinal image restoration. However, a main shortcoming of TV regularization is its potential to experience the staircase effects, particularly in smooth regions of the image. To overcome the drawback, a new image restoration model is proposed for AO retinal images. This model utilizes the overlapping group sparse total variation (OGSTV) as a regularization term. Due to the structural characteristics of AO retinal images, only partial information regarding the PSF is known. Consequently, we have to solve a more complicated myopic deconvolution problem. To address this computational challenge, we propose an ADMM-MM-LAP method to solve the proposed model. First, we apply the alternating direction method of multiplier (ADMM) as the outer-layer optimization method. Then, appropriate algorithms are employed to solve the ADMM subproblems based on their inherent structures. Specifically, the majorization–minimization (MM) method is applied to handle the asymmetry OGSTV regularization component, while a modified version of the linearize and project (LAP) method is adopted to address the tightly coupled subproblem. Theoretically, we establish the complexity analysis of the proposed method. Numerical results demonstrate that the proposed model outperforms the existing state-of-the-art TV model across several metrics.

BackgroundOrthogonal translation computed laminography (OTCL) has great potential for tiny fault detection in laminated structure thin-plate parts. It offers a larger magnification ratio but generates limited projection data, which would result in aliasing artifacts in the reconstructed image.ObjectiveOne way to minimize these artifacts is to use prior information, such as the piecewise constant property and prior image information. This work was inspired by the adaptive-weighted high order total variation (awHOTV) model, which is known for its ability to protect edge and detail information. Meanwhile, the laminated structure thin-plate parts are printed using computer-aided design (CAD) images, which provide structural information.MethodsTo create a reliable CAD information beforehand, we adopted a two-in-one estimation method. Therefore, combining the CAD information with the awHOTV model, we propose an improved adaptive weighted higher-order TV (Cpi-awHOTV) model based on the CAD prior and use the adaptive steepest descent projection onto convex set (ASD-POCS) algorithm to solve the imaging model.ResultsTo evaluate the performance of our algorithm, we compared it with existing filtered back projection (FBP), simultaneous algebraic reconstruction technique (SART), total variation (TV), adaptive-weighted TV (awTV), and high order TV (HOTV)algorithms on phantom1 and phantom2 with various scanning angle ranges. Additionally, we used the phantom2 as the CAD prior in real data experiments. The results show that, the Cpi-awHOTV algorithm can obtain high-quality reconstructed images and better quantitative evaluation indicators.ConclusionsVisual inspection and quantitative analysis of reconstructed images demonstrate that the Cpi-awHOTV algorithm effectively protects edge information, and reduces aliasing artifacts due to interference from adjacent slice structures.

In recent years, spectral computed tomography (CT) has attracted extensive attention. The purpose of this study is to achieve a low-cost and fast energy spectral CT reconstruction algorithm by implementing multi-limited-angle scans. General spectral CT projection data are collected over a full-angular range of 360 degrees. We simulate multi-source spectral CT by using a pair of X-ray source/detector. To speed up scanning, multi-limited-angle scanning was used in each energy channel. On this basis, an average image induced relative total variation (Aii-RTV) with multi-limited-angle spectral CT image reconstruction model is proposed. The iterative algorithm is used to solve Aii-RTV. Before iteration, the weighted average projection data of the multi-limited-angle energy spectral is carried out. In each step of the iterative algorithm flow is as follows: First, the relative total variation (RTV) reconstruction model is used to reconstruct the average image using average projection data. Then, the partial derivative of the average image is used to calculate the inherent variation in RTV model due to the integrity of the average image, and take its reciprocal as the weight coefficient of the windowing total variation of each energy channel reconstruction image. Finally, the average energy image is used to guide the multi-limited-angle projection data to reconstruct the image of each energy channel so as to suppress the limited-angle artifact of each energy channel image. In addition, we also discuss the influence of parameter selection on reconstructed image quality, which is important for regularization model. Through the reconstruction of multi-limited-angle spectral CT projection data, quantitative results and reconstructed images show that our algorithm has better performance than prior image constrained compressed sensing (PICCS) and RTV. The average PSNR of our reconstruction results in different channels was 35.6273, 4.533 and 2.301 higher than RTV (31.0943) and PICCS (33.3263), respectively.

The shapes and scales of agricultural and forestry plots are quite different, making it challenging to extract refined boundaries. In order to extract refined boundaries of agricultural and forestry plots of different shapes and scales, a refined extraction method of agricultural and forestry plot boundaries based on Yolov8n was proposed. By enhancing the high-order nonconvex total variation model to reduce noise and preprocess the initial agricultural and forestry plot image, the edge detail information of the image is enhanced; the Yolov8n model is used to extract and fuse the multi-scale agricultural and forestry plot features of the processed image to obtain a multi-scale feature map. Based on this, the bounding boxes of plots of different scales are detected; combined with the fast active contour model, the edge details of the obtained multi-scale plot bounding boxes are further refined to obtain more refined boundaries of agricultural and forestry plots. The results show that this method can enhance the edge details of agricultural and forestry plot images through noise reduction processing, and achieve refined extraction of agricultural and forestry plot boundaries of various scales and shapes. The extracted boundaries can almost match the actual boundaries, and the average accuracy can reach 98%. The above can provide a guarantee for the rational allocation of resources for precision agriculture.