Search

The models based on convolutional neural network have achieved excellent results in image super-resolution by acquiring prior knowledge from a large number of images, but such models still have problems such as the features between layers in the depth network cannot be effectively fused, the number of parameters is too large, and cross-channel feature learning is impossible. Based on this, a deep recursive residual channel attention network (DRRCAN) model was proposed in this paper. To solve the problem that the information between different layers in the deep network cannot be fused effectively, this paper constructs a channel feature fusion module, which can effectively fuse the feature information of different layers. To solve the problem that the parameters increase sharply due to the increase of network depth, recursive blocks are adopted in this paper, which greatly reduces the number of parameters in the deep network. The channel attention is integrated to enable the model to learn features across channels. In addition, to avoid gradient explosion or disappearance, residual modules, long skip connections are introduced to improve the stability and generalization ability of the model. Extensive benchmark evaluations validate the superiority of the proposed DRRCAN model compared with existing algorithms. [ABSTRACT FROM AUTHOR]

As a key tool in hydraulic fracturing, the packer plays a vital role in the process of unconventional oil and gas production. Its sealing performance has a significant impact on production. The packer is in high-temperature and high-pressure (HTHP) environment for a long time. It is easy to produce significant stress relaxation, which affects the sealing performance of the packer. In this paper, rubber foundation tests at 150 °C temperature were carried out. The hyperelastic–viscoelastic model of the rubber material was determined. The influence of the stress relaxation on its sealing performance was investigated. In addition, the effects of setting load, height, inner diameter, and outer diameter on the sealing performance were systematically investigated. The findings show that under HTHP, the stress relaxation phenomenon leads to a 6.9% decrease in peak shear stress, 10.3% decrease in peak von Mises stress, and 6.7% decrease in peak contact stress, reducing the risk of damage to the packing element and leading to a decrease in its sealing performance. In addition, it was found that increasing the setting load can significantly increase the contact stress, which also increases the risk of damage. When increasing the height, the peak shear stress and peak contact stress do not change much but can significantly reduce the peak von Mises stress. When increasing the inner diameter, the peak shear stress and peak contact stress show a trend of first increasing and then decreasing, and the peak von Mises stress has been decreasing. Increasing the outer diameter can better improve the force situation, increase the contact stress, and make the contact stress distribution more uniform. It is necessary to consider the influence of stress relaxation on the sealing performance of the packer under HTHP; meanwhile, this paper has guiding significance for the design of packer structures. [ABSTRACT FROM AUTHOR]

Many real-world applications, such as energy consumption alerts and long-term traffic planning, require the prediction of data changes over extended horizons. Long-term time series forecasting (LTSF) demands methods with robust prediction capabilities. Recently, transformer-based methods have shown immense potential for LTSF. However, these methods rely heavily on positional encoding to maintain temporal information, inevitably leading to the loss of temporal patterns. Moreover, various self-attention mechanisms often capture only intrasequence features, requiring a greater ability to capture intersequence features. The quadratic complexities of time and memory render training challenging when dealing with lengthy input sequences. We propose FL-Net to enhance the accuracy of LTSF. To capture the seasonal and trend components in the time series accurately, FL-Net segments the input sequences into coarse-grained trends and seasonal components using moving averages. Two sets of encoders extract the temporal features from these components. Our proposed frequency external attention method utilizes two external, compact, learnable, and shared memories to learn and store the time- and frequency-domain features of the entire training set, demonstrating high efficiency in terms of linear complexity. Simultaneously, we propose a multi-scale cross-decomposition method that further decomposes trends and seasonal elements into finer-grained components, thereby enhancing the capability of the model to extract temporal features. Experimental results on nine real-world benchmark datasets demonstrate that FL-Net achieves higher prediction accuracy in long-term forecasting than state-of-the-art methods. [ABSTRACT FROM AUTHOR]

This paper proposes a novel single-image super-resolution method based on local biquadratic spline with edge constraints and adaptive optimization in transform domain. The complex internal structure of the image makes the values of adjacent pixels often differ greatly. Using surface patches to interpolate image blocks can avoid large surface oscillation. Because the quadratic spline has better shape-preserving property, we construct the biquadratic spline surface on each image block to make the interpolation more flexible. The boundary conditions have great influence on the shape of local biquadratic spline surfaces and are the keys to constructing surfaces. Using edge information as a constraint to calculate them can reduce jagged and mosaic effects. To decrease the errors caused by surface fitting, we propose a new adaptive optimization model in transform domain. Compared with the traditional iterative back-projection, this model further improves the magnification accuracy by introducing SVD-based adaptive optimization. In the optimization, we convert similar block matrices to the transform domain by SVD. Then the contraction coefficients are calculated according to the non-local self-similarity, and the singular values are contracted. Experimental comparison with the other state-of-the-art methods shows that the proposed method has better performance in both visual effect and quantitative measurement. [ABSTRACT FROM AUTHOR]

Based on the image self‐similarity and singular value decomposition (SVD) techniques, the authors propose an iterative adaptive global denoising method. For the structural differences between image patches, they adaptively determine the size of the search window. In each window, a similar image patch matrix is constructed based on the multi‐scale similarity measure. In order to ensure the speed of the method, the adaptive step size and the number of image patches are introduced, and all image patches are denoised in different iterations. This not only ensures the speed of the method, suppresses residual noise, but also reduces the artefacts caused by the fixed step size and the number of image patches. Therefore, the problem of image denoising is converted to the estimation of low‐rank matrix. New singular values are estimated according to the noise level, and similar image patch matrices without noise are estimated using them and corresponding singular vectors. Experimental results show that compared with the state‐of‐the‐art denoising algorithms, this method has a higher PSNR and FSIM, and has a good visual effect. The new method can be applied to image and video restoration, target recognition and image classification. [ABSTRACT FROM AUTHOR]

Part of important structural edges in the image is smoothed due to the small gradients, while the others are preserved with greater gradients. Therefore, the authors propose a two‐stage image smoothing method based on edge‐patch histogram equalisation and patch decomposition. The authors' purpose is to increase the gradient of important structural edges while reducing the gradient of the texture region. Therefore, they divide the image into edge‐patches where the structural edges are concentrated or non‐edge‐patches where the texture details are concentrated by image segmentation. The edge‐patch needs to be equalised by the histograms for increasing the gradient of the edge pixels. All patches are decomposed to extract the smooth component for reducing the gradient of pixels. The smooth component of each patch is smoothed via L0 gradient minimisation. In order to ensure the continuity of the patch boundaries, the edge‐patch is inversely equalised. Finally, the whole image is smoothed via L0 gradient minimisation for removing residual textures and seams. Experimental results demonstrate that the proposed method is more competitive in maintaining important structural edges and removing texture details than the state‐of‐the‐art approaches. The proposed method can be applied to many areas of image processing. [ABSTRACT FROM AUTHOR]

Conditional proxy re-encryption allows a semi-trusted third-party proxy to convert Alice's ciphertext that meets a certain condition into another set of ciphertext. Meanwhile, the proxy cannot obtain any valuable information in the process. Nevertheless, a single re-encryption condition is difficult to meet the needs of practical applications, and the existing conditional proxy re-encryption scheme cannot flexibly control the conversion conditions. To solve this problem, this paper proposes a multi-conditional proxy broadcast re-encryption (MC-PBRE) scheme for file sharing systems. In our scheme, a single user can delegate the decryption rights of the encrypted file to a group of users. At the same time, users can set any number of conditions to control the user group decryption permissions. This paper demonstrates the security of the scheme for chosen-ciphertext attacks in the standard model. Furthermore, the program can resist collusion attacks, which means that the agent cannot collude with the user group to get the private key of the principal. [ABSTRACT FROM AUTHOR]

This article introduces customized screening ensemble with shape‐adaptive quantile regression (CseAQR), a novel probabilistic interval forecasting method built upon the quantile regression model. CseAQR utilizes ensemble learning to perform adaptive quantile regression prediction, which can handle the heteroscedasticity feature in time series data by using a weighted adaptive allocation loss function to enhance the adaptability of the basic quantile regression model on the dataset. The model performance predictor is used to select the optimal ensemble learner combination, assign reasonable adaptive weights to it, and obtain a preliminary prediction interval through weighted aggregation. Combining ensemble learners not only improves the accuracy and robustness of prediction intervals but also ensures the commutativity required for conformal prediction. Finally, the conformal prediction method is applied to locally adjust the prediction interval, constructing a more consistently aligned prediction interval with the actual data on a narrower basis. [ABSTRACT FROM AUTHOR]

In recent years, there has been a surge of research in the field of medical image segmentation using hybrid CNN‐Transformer network architectures. Most of these studies leverage the attention mechanism of ViT to overcome the limitations of CNN architectures in capturing long‐range dependencies. However, these hybrid model approaches also have some potential drawbacks. First, due to the heavy reliance of the Transformer's attention mechanism on global information, it can lead to a significant increase in computational cost when dealing with high‐resolution input images. Furthermore, the convolutional and attention mechanisms in hybrid models have different interpretability in information extraction and decision‐making, which poses a challenge for the interpretability of the convolutional part. Our proposed model, DWHA, addresses these limitations and outperforms state‐of‐the‐art models in a range of medical image segmentation tasks, such as abdominal multiorgan segmentation, automatic cardiac diagnosis, neurostructure segmentation, and skin lesion segmentation, achieving significantly superior performance. Specifically, on the abdominal multiorgan segmentation dataset, DWHA outperformed the previous state‐of‐the‐art baseline by 0.57%; on the neurostructure segmentation dataset, it achieved an improvement of 1.17%; and on the skin lesion segmentation dataset, it achieved an improvement of 0.91%. These significant improvements suggest that DWHA may become the preferred model in the field of medical image segmentation. [ABSTRACT FROM AUTHOR]

With the continuous development of deep learning, long sequence time-series forecasting (LSTF) has attracted more and more attention in power consumption prediction, traffic prediction and stock prediction. In recent studies, various improved models of Transformer are favored. While these models have made breakthroughs in reducing the time and space complexity of Transformer, there are still some problems, such as the predictive power of the improved model being slightly lower than that of Transformer. And these models ignore the importance of special values in the time series. To solve these problems, we designed a more concise network named Resformer, which has four significant characteristics: (1) The fully sparse self-attention mechanism achieves O ⁢ (퐿푙표푔퐿) time complexity. (2) The AMS module is used to process the special values of time series and has comparable performance on sequences dependency alignment. (3) Using quadratic linear transformation, a simple LT module is designed to replace the self-attention mechanism. It effectively reduces redundant information. (4) The DistPooling method based on data distribution is proposed to suppress redundant information and noise. A large number of experiments on real data sets show that the Resformer method is superior to the existing improved model and standard Transformer method. [ABSTRACT FROM AUTHOR]

• This paper proposes a new feature constrained polynomial interpolation, which not only considers the positional relationship between pixels, but also considers the edge information of the image. By incorporating the anisotropy of the image into the polynomial parameter solution, the polynomial we obtained can effectively preser ve the structure of the image. • We propose a new method for filtering in image structure residuals, which simultaneously utilizes the non local self similarity of the image and the sparsity of the global structure of the image. By decomposing the image into smooth components and structural residual components, we can effectively filter out the noise in the image without destroying the smooth components of the image. • We propose a new single image super resolution model that simultaneously utilizes non local self si milarity, cross resolution similarity, and global structural sparsity. The optimization steps can be selected independently according to the accuracy requirements, and it can also be combined with other image super resolution algorithms. Due to the important application value of image super-resolution, many image super-resolution algorithms have been proposed in recent years. However, many single-image super-resolution algorithms usually have their own limitations and cannot achieve ideal results. To this end, this paper proposes a new single-image super-resolution method that uses non-local self-similarity, cross-resolution similarity, and global structure sparsity without relying on external instances. First, we obtain the initial high-resolution image through feature-constrained polynomial interpolation. Then, we use a database built by the input image to perform cross-resolution learning to predict the missing high-frequency information in the image. Finally, we use the residual filtering proposed in this paper to remove the noise introduced during interpolation and cross-resolution learning. Our method can be combined with other image super-resolution algorithms. Through extensive comparison experiments to verify, our method achieves higher numerical accuracy and pleasing visual effects. [ABSTRACT FROM AUTHOR]

• This paper proposes a content-aware segmentation method. The intersection point and boundary information of the image patch and the smallest bounding box are stored in a quad-tree. • The histogram equalization is used to improve the gradient of edges with the smaller gradient in the edge-patches. To ensure color consistency, inverse equalization is required. • Extract the smoothing component by decomposing image for reducing gradients of texture details, so that structural edges and texture details can be more easily distinguished. • A two-stage image smoothing method with gradient direction constraints is proposed. Image patches are smoothed with the edge pixel ratio as the smoothing factor. For better removing textures with larger gradients while preserving structural edges with smaller gradients, we propose an image smoothing method based on histogram equalized content-aware patches and direction-constrained sparse gradients. In order to better smooth the boundary concentration regions while maintaining the continuity of boundary pixels, a content-aware patching technology with boundary constraints is proposed. The irregular patches are represented by the smallest rectangular bounding boxes to reduce the computational complexity. Based on edge information, patches are divided into edge-patches and non-edge-patches. Histogram equalization is used to improve the edge gradient of patches with structural edge concentration, while the image decomposition is used to reduce the gradient of texture details. Taking the edge information as the smoothing factor, each patch is smoothed via direction-constrained sparse gradients. The entire image needs to be further smoothed to remove residual texture details. Experimental results show that new method has better visual effects in retaining structural edges and removing texture details, and has many applications, including edge detection, image abstraction, detail enhancement, and texture mapping. [ABSTRACT FROM AUTHOR]

• This paper proposes a method for constructing the combined cubic surface. A quadratic polynomial surface patch is first constructed in the 3 × 3 (2 × 2) region centered on each pixel. Then the effect of each coefficient on low frequency and high frequency information is researched. First calculate the coefficient that has a large effect on the low-frequency information, and then calculate the coefficient that has a greater effect on the high- frequency information. All quadratic surface patches are weighted to form the cubic surface. • This paper proposes an efficient method for constructing the initial cubic polynomial surface fitting to the input image with the constraint of local features (such as image edges and textures) and distance. By sampling the initial surface, the initial high-resolution image with higher precision and better visual effect can be obtained. • This paper proposes an efficient method for constructing error high-resolution image to modify initial high-resolution, including calculating error high-resolution image and constructing the cubic surface with higher precision to smooth error high-resolution image. • This paper proposes an efficient method that combines non-local similarity iteration optimization with secondary filtering to optimize the error high-resolution image. The quality of image edges and textures plays a vital role in the visual effect of images. For keeping the features of image edges and textures better, we propose an image enlargement method based on cubic fitting surfaces with local features as constraints. The cubic fitting surface constructed with the edge and the distance as constraints effectively improves the approximation accuracy and shape retention, resulting in sharper edges in the initially enlarged image. The error between the cubic fitting surface and the input image surface leads to the error between the initially enlarged image and the input image. Using image self-similarity to filter the initially enlarged image can further improve its quality. In the iterative optimization process, the cubic surface with higher fitting accuracy is constructed on smaller regions of low resolution error images, which effectively improves the quality of the enlarged image and accelerates the convergence speed. Because the algorithm effectively utilizes the local feature information, compared with other state-of-the-art methods, the new method not only has higher PSNR/SSIM in numerical accuracy, but also has clear edges and textures visually. [ABSTRACT FROM AUTHOR]

Urban scene image segmentation is an important research area in high-resolution remote sensing image processing. However, due to its complex three-dimensional structure, interference factors such as occlusion, shadow, intra-class inconsistency, and inter-class indistinction affect segmentation performance. Many methods have combined local and global information using CNNs and Transformers to achieve high performance in remote sensing image segmentation tasks. However, these methods are not stable when dealing with these interference factors. Recent studies have found that semantic segmentation is highly sensitive to frequency information, so we introduced frequency information to make the model learn more comprehensively about different categories of targets from multiple dimensions. By modeling the target with local features, global information, and frequency information, the target features can be learned in multiple dimensions to reduce the impact of interference factors on the model and improve its robustness. In this paper, we consider frequency information in addition to combining CNNs and Transformers for modeling and propose a Multidimensional Information Fusion Network (MIFNet) for high-resolution remote sensing image segmentation of urban scenes. Specifically, we design an information fusion Transformer module that can adaptively associate local features, global semantic information, and frequency information and a relevant semantic aggregation module for aggregating features at different scales to construct the decoder. By aggregating image features at different depths, the specific representation of the target and the correlation between targets can be modeled in multiple dimensions, allowing the network to better recognize and understand the features of each class of targets to resist various interference factors that affect segmentation performance. We conducted extensive ablation experiments and comparative experiments on the ISPRS Vaihingen and ISPRS Potsdam benchmarks to verify our proposed method. In a large number of experiments, our method achieved the best results, with 84.53% and 87.3% mIoU scores on the Vaihingen and Potsdam datasets, respectively, proving the superiority of our method. The source code will be available at https://github.com/JunyuFan/MIFNet. [ABSTRACT FROM AUTHOR]

Effective wildfire evacuation planning requires understanding where evacuees are likely to travel and temporarily reside. Detailed information on evacuee destinations is a valuable input into critical decisions related to evacuation traffic control, shelter assignment, and return-entry planning. To improve our understanding of where evacuees stay, we analyzed GPS data generated by mobile devices from the 2019 Kincade Fire evacuation in Sonoma County, California. A new algorithm was developed to infer evacuee stops and destinations from mobile device GPS data. Evacuees were placed into distinct subgroups associated with their warning and departure context (e.g., self-evacuee, warned evacuee, ordered evacuee, shadow evacuee). The results reveal a surprising diversity of destinations not captured and mapped in prior studies. Analysis of the sub-groups showed marked differences in destination types and travel distances. The results of this study can aid in planning for future wildfires by providing emergency managers with detailed findings on empirical evacuee destination distributions. [ABSTRACT FROM AUTHOR]

Repeated positioning accuracy, axial stiffness, cutting vibration, and rotation noise are key indexes that affect the comprehensive performance of computer numerical control (CNC) servo turrets. In order to evaluate the comprehensive performance of CNC servo turrets accurately, a new evaluation method is proposed based on accuracy retentivity theory in this paper. The retentivity degree evaluation methodology of performance indexes of the turret based on error threshold and 3δprinciple are proposed. Analytic hierarchy process (AHP) is used to calculate the weight of the influence degree on the turret, and the comprehensive performance retentivity degree of the turret is obtained. A comprehensive performance degradation model is established by using least squares. Finally, the comprehensive performance tracking test and analysis are conducted by taking two turrets as an example. The degradation law of the comprehensive performance retentivity degree of turrets is obtained, and the remaining performance retentivity time is predicted. [ABSTRACT FROM AUTHOR]

Highlights • A pulsating plate type heat pipe radiator with multiple condensation ends is designed and manufactured. • The heat transfer performance and start-up performance of the vertical radiator are studied. • The advantages of this vertical radiator over the aluminum fin radiator are studied. Abstract A vertical radiator with multi pulse condensation ends and a plane evaporating end for vertical CPU cooling is developed. A 304 stainless steel test piece is processed and a detailed experimental study of heat transfer performance and start-up performance is performed. The influence of liquid filling rate, heating power and wind speed on the two performance of vertical heat pipe radiator is analyzed, and compared with aluminum fin computer CPU radiator. The results show that when the radiator works stably, the optimal filling ratio is 25%, the lowest thermal resistance is only 0.1 °C/W; the average temperature of heat source increases with the increase of heating power and decreases with the wind speed. The temperature distributions of the five cold ends exhibit a symmetry that attenuates from the middle to the both sides. In addition, when the radiator starts up, as the heating power increases, the initial start-up time becomes shorter, the start-up temperature becomes lower, and the stability is also better. But there is no symmetry in the start sequence of the condensation ends. Compared with the current aluminum fin radiator, the vertical radiator has superior heat dissipation performance, temperature uniformity and stability. Such as, under 80 W heating power, the average heat source temperature is 66 °C, the average temperature deviation of the heat source is 3.16 °C, compared to the aluminum fin radiator, they dropped by 17.3 °C and 0.61 °C, respectively. [ABSTRACT FROM AUTHOR]

An improved phase unwrapping method is proposed to reduce the projection fringes in three-dimensional (3D) surface measurement. Color fringe patterns are generated by encoding with sinusoidal fringe and stair phase fringe patterns in red and blue channels. These color fringe patterns are projected onto the tested objects and then captured by a color CCD camera. The recorded fringe patterns are separated into their RGB components. Two groups of four-step phase-shifting fringe patterns are obtained. One group of the stripes are four sinusoidal patterns, which are used to determine the wrapped phase. The other group of stripes are four sinusoidal patterns with the codeword embedded into stair phase, whose stair changes are perfectly aligned with the 2 π discontinuities of sinusoidal fringe phase, which are used to determine the fringe order for the phase unwrapping. The experimental results are analyzed and compared with those of the method in Zheng and Da (2012. Opt Express 20(22):24139–24150). The results show that the proposed method needs only four fringe patterns while having less error. It can effectively reduce the number of projection fringes and improve the measuring speed. [ABSTRACT FROM AUTHOR]