Low-resolution Images Research Articles

Tunnel Seismic Detection for Tunnel Boring Machine by Joint Active and Passive Source Method and Imaging Advanced Prediction

Abstract Tunnel seismic advanced prediction method enables the detection of anomalous bodies in front of the tunnel face and the reduction of tunnel construction risk. The active source and the passive source detection methods are two commonly used for imaging advanced prediction of tunnel boring machine (TBM). Although both methods achieve good results, each method still has room for improvement. The active detection method is not conducive to long-distance detection because the sources are generally far away from the anomalous bodies and the received signals are weak. The passive detection method usually produces results of low resolution and with limited forecast precision. Since the active and passive source methods are applied separately, the geophone utilization rate is low and the cost is high. Moreover, both methods have low signal-to-noise ratio and low imaging resolution for single detection data. To overcome the above drawbacks, this article proposes a joint TBM tunnel seismic detection method and an imaging advanced prediction method to detect anomalous bodies in medium and long distances. The advanced prediction employs multiple stacking schemes to enhance the reflection wave energy of the anomalous bodies and suppress the interference wave. The proposed method integrates the active source data into the passive source method to extract the P- and S-waves. The joint active and passive source method further incorporates multiple stack schemes to achieve much higher resolution imaging and provides more accurate detection results. Moreover, the proposed joint method allows multiple utilization of the geophone and reduces the cost. Simulation results are presented for performance verification.

Heuristic strategy using hybrid deep learning with transfer learning for oral cancer detection

Oral cancer becomes the most disastrous ailment that affects the oral cavity parts of the mouth. Oral cancer diagnosis is the main challenge in the medical field. It becomes expensive and less capable of classifying oral cancer. In some cases, it may cause unnecessary morbidity and mortality. Recently, the detection of malignant and premalignant oral lesions has been a critical process owing to their low image resolution and lower acquisition time. Thus, a novel hybrid deep learning with meta-heuristic-based optimization is proposed. The pre-processing occurs by median filtering and Contrast Limited Adaptive Histogram Equalization (CLAHE). The CLAHE method is utilized to reduce unwanted noise. Finally, the classification is done by a proposed hybrid-based deep learning model termed as Recurrent Deep Belief Network (RDBN), in which the Deep Belief Network (DBN) is incorporated with the Recurrent Neural Network (RNN). Here, the RDBN helps to increase the performance classification. Furthermore, the hyperparameters of the RDBN model, such as learning rate, epochs and hidden neurons, are tuned using the proposed Hybrid Beetle-Barnacle Swarm Optimization (HBBSO) algorithm, where the Barnacles Mating Optimizer (BMO) is superimposed with Beetle Swarm Optimization (BSO) algorithms. In the given proposed model, the selected features are extracted by the optimization algorithms. Here, the parameters are fine-tuned to get the better optimal solution. From the experimental outcome, the developed model has acquired 5.2% better than PSO-RDBN, 5.6% improved than GWO-RDBN, 3.2% enhanced than BSO-RDBN and 3.5% superior to BMO-RDBN regarding accuracy. Thus, the proposed model achieves higher results for detecting oral cancer with enhanced classification performance than the existing approaches.

NIMG-33. HIGH-RESOLUTION SODIUM MRI OF HUMAN GLIOMAS AT 3T USING PHYSICS-BASED GENERATIVE AI

Abstract BACKGROUND Sodium neuroimaging is a promising technique for diagnosing and monitoring brain tumors, providing insights into microenvironment and metabolism. However, at 3T, it is limited by low signal-to-noise ratio and resolution, resulting in long acquisition times and low-quality images. To address these limitations, we developed a physics-informed generative adversarial network (GAN) approach for high-resolution sodium neuroimaging of brain tumors at 3T. METHODS 5,078 anatomical sequences from 1,330 brain tumor patients undergoing routine proton MRI were used to create a synthetic dataset with physics-based simulated artifacts, including B0 magnetic susceptibility artifacts, chemical shift artifacts, Nyquist aliasing artifacts, Gibbs artifacts, and low-resolution images, with Rician noise added during training to augment the dataset. By using multiple proton MRI contrasts, we hypothesized that the GAN model would learn to correct artifacts while preserving the inherent contrast information. We employed a transfer learning approach with a modified Pix2PixGAN architecture and an Attention-R2UNet generator. Twenty glioma patients consented to participate in this IRB-approved prospective study. Sodium and proton MRIs were acquired in the same sessions using a 3T Siemens scanner using a dual-tuned head coil, with a 3D spoiled gradient-echo sequence optimized for short TE. Sodium-proton exchanger (NHE1) expression was tested with immunohistochemistry from image-guided surgical biopsies. RESULTS High-resolution synthetic-sodium MR images were generated using the GAN model. The proposed method successfully reconstructed high-resolution synthetic-sodium MR images with improved SNR compared to traditional sodium images. Synthetic and native measurements were strongly correlated (R2=0.8539, P<0.0001). A significant correlation was seen between synthetic-sodium MR measurements and relative NHE1 tissue expression (ρ=0.5817, P=0.0036). CONCLUSION Results suggest that high-resolution synthetic-sodium MRI retains much of the inherent information from native sodium MR while accurately representing NHE1 expression. This approach could make high-resolution sodium neuroimaging at 3T feasible in the clinical environment, potentially improving diagnosis, monitoring, and treatment of brain tumors.

MIGHTEE: The continuum survey data release 1

Abstract The MeerKAT International GHz Tiered Extragalactic Exploration Survey (MIGHTEE) is one of the large survey projects using the MeerKAT telescope, covering four fields that have a wealth of ancillary data available. We present Data Release 1 of the MIGHTEE continuum survey, releasing total intensity images and catalogues over ∼20 deg2, across three fields at ∼1.2-1.3 GHz. This includes 4.2 deg2 over the Cosmic Evolution Survey (COSMOS) field, 14.4 deg2 over the XMM Large-Scale Structure (XMM-LSS) field and deeper imaging over 1.5 deg2 of the Extended Chandra Deep Field South (CDFS). We release images at both a lower resolution (7–9 arcsec) and higher resolution (∼5 arcsec). These images have central rms sensitivities of ∼1.3 −2.7 μJy beam−1 (∼1.2 −3.6 μJy beam−1) in the lower (higher) resolution images respectively. We also release catalogues comprised of ∼144 000 (∼114 000) sources using the lower (higher) resolution images. We compare the astrometry and flux-density calibration with the Early Science data in the COSMOS and XMM-LSS fields and previous radio observations in the CDFS field, finding broad agreement. Furthermore, we extend the source counts at the ∼10 μJy level to these larger areas (∼20 deg2) and, using the areal coverage of MIGHTEE we measure the sample variance for differing areas of sky. We find a typical sample variance of 10-20percnt for 0.3 and 0.5 sq. deg. sub-regions at S1.4 ≤ 200 μJy, which increases at brighter flux densities, given the lower source density and expected higher galaxy bias for these sources.

Integrating Super-Resolution with Deep Learning for Enhanced Periodontal Bone Loss Segmentation in Panoramic Radiographs

Periodontal disease is a widespread global health concern that necessitates an accurate diagnosis for effective treatment. Traditional diagnostic methods based on panoramic radiographs are often limited by subjective evaluation and low-resolution imaging, leading to suboptimal precision. This study presents an approach that integrates Super-Resolution Generative Adversarial Networks (SRGANs) with deep learning-based segmentation models to enhance the segmentation of periodontal bone loss (PBL) areas on panoramic radiographs. By transforming low-resolution images into high-resolution versions, the proposed method reveals critical anatomical details that are essential for precise diagnostics. The effectiveness of this approach was validated using datasets from the Chungbuk National University Hospital and the Kaggle data portal, demonstrating significant improvements in both image resolution and segmentation accuracy. The SRGAN model, evaluated using the Peak Signal-to-Noise Ratio (PSNR) and Structural Similarity Index Measure (SSIM) metrics, achieved a PSNR of 30.10 dB and an SSIM of 0.878, indicating high fidelity in image reconstruction. When applied to semantic segmentation using a U-Net architecture, the enhanced images resulted in a dice similarity coefficient (DSC) of 0.91 and an intersection over union (IoU) of 84.9%, compared with 0.72 DSC and 65.4% IoU for native low-resolution images. These results underscore the potential of SRGAN-enhanced imaging to improve PBL area segmentation and suggest broader applications in medical imaging, where enhanced image clarity is crucial for diagnostic accuracy. This study also highlights the importance of further research to expand the dataset diversity and incorporate clinical validation to fully realize the benefits of super-resolution techniques in medical diagnostics.

Resolution-dependent MRI-to-CT translation for orthotopic breast cancer models using deep learning.


This study aims to investigate the feasibility of utilizing generative adversarial networks (GANs) to synthesize high-fidelity CT images from lower-resolution MR images. The goal is to reduce patient exposure to ionizing radiation while maintaining treatment accuracy and accelerating MR image acquisition. The primary focus is to determine the extent to which low-resolution MR images can be utilized to generate high-quality CT images through a systematic study of spatial resolution-dependent MRI-to-CT image conversion. 
Approach. 
Paired MRI-CT images were acquired from healthy control and tumor models, generated by injecting MDA-MB-231 and 4T1 tumor cells into the mammary fat pad of nude and BALB/c mice to ensure model diversification. To explore various MRI resolutions, we downscaled the highest-resolution MR image into three lower resolutions. Using a customized U-Net model, we automated region of interest masking for both MRI and CT modalities with precise alignment, achieved through three-dimensional affine paired MRI-CT registrations. Then our customized models, Nested U-Net Generative Adversarial Network (NUGAN) and Attention U-Net Generative Adversarial Network (AUGAN), were employed to translate low-resolution MR images into high-resolution CT images, followed by evaluation with separate testing datasets.
Main Results.
Our approach successfully generated high-quality CT images (0.142 mm²) from both lower-resolution (0.282 mm²) and higher-resolution (0.142 mm²) MR images, with no statistically significant differences between them, effectively doubling the speed of MR image acquisition. Our customized GANs successfully preserved anatomical details, addressing the typical loss issue seen in other MRI-CT translation techniques across all resolutions of MR image inputs.
 Significance. 
This study demonstrates the potential of using low-resolution MR images to generate high-quality CT images, thereby reducing radiation exposure and expediting MRI acquisition while maintaining accuracy for radiotherapy.&#xD.

Scalable multimodal approach for face generation and super-resolution using a conditional diffusion model

Multimodal Conditioned face image generation and face super-resolution are significant areas of research. To achieve optimal results, this paper utilizes diffusion models as the primary engine for these tasks. This paper presents two main contributions: (1) “Speaking the Language of Faces” (SLF): a flexible, modular, fusion-less and architecturally simple multimodal system. (2) A Scalability scheme and a sensitivity analysis which can assist practitioners in system parameter estimation and feature selection. SLF consists of two main components: a feature vector generator (encoder), and an image generator (decoder) utilizing a conditional diffusion model. SLF can accept various inputs, including low-resolution images, speech signals, person attributes (age, gender, ethnicity), or any combination of these. Moreover, Scalability based on conditional scale values is utilized. The implementation of SLF has confirmed its versatility (e.g., speech to face image generation, conditioned face super-resolution). We trained multiple system versions to conduct a sensitivity analysis and to determine the influence of each individual feature on the output image. Consequently, speaker embeddings have proven to be sufficient audio features for our task. It was also found that the effects of audio signals are profound and are more pronounced than those of the low resolution images (8 × 8), whose effects are still significant. The effect of gender, ethnicity and age were found to be moderate. On another note, conditional scale values significantly impact the system’s behavior and performance.

Lightweight Reference-Based Video Super-Resolution Using Deformable Convolution

Super-resolution is a technique for generating a high-resolution image or video from a low-resolution counterpart by predicting natural and realistic texture information. It has various applications such as medical image analysis, surveillance, remote sensing, etc. However, traditional single-image super-resolution methods can lead to a blurry visual effect. Reference-based super-resolution methods have been proposed to recover detailed information accurately. In reference-based methods, a high-resolution image is also used as a reference in addition to the low-resolution input image. Reference-based methods aim at transferring high-resolution textures from the reference image to produce visually pleasing results. However, it requires texture alignment between low-resolution and reference images, which generally requires a lot of time and memory. This paper proposes a lightweight reference-based video super-resolution method using deformable convolution. The proposed method makes the reference-based super-resolution a technology that can be easily used even in environments with limited computational resources. To verify the effectiveness of the proposed method, we conducted experiments to compare the proposed method with baseline methods in two aspects: runtime and memory usage, in addition to accuracy. The experimental results showed that the proposed method restored a high-quality super-resolved image from a very low-resolution level in 0.0138 s using two NVIDIA RTX 2080 GPUs, much faster than the representative method.

Optical system characterization in Fourier ptychographic microscopy

Fourier ptychographic microscopy (FPM) is a recent technique to overcome the diffraction limit of a low numerical aperture (NA) objective lens by algorithmic post-processing of several low-resolution images. It can increase the space-bandwidth product of an optical system by computationally combining images captured under different illumination conditions. Vignetting determines the spatial extent of the bright and dark regions in the captured images. State-of-the-art analyses treat vignetting as a nuisance that needs to be reduced or excluded from algorithmic consideration using ad hoc decision rules. In contrast, this work investigates vignetting effects as a tool to infer a range of properties of the optical system. Generally, the goal of the FPM reconstruction algorithm is to achieve results that closely resemble the actual specimen at the highest resolution possible. However, as the optimization process does not necessarily guarantee a unique solution, we identify system properties that support alignment between computational predictions and empirical observations, potentially leading to a more accurate and reliable analysis. To achieve this, we characterize the individual system components of the experimental setup and compare experimental data to both, geometrical and wave optical simulations. We demonstrate that using vignetting as an analytical tool enables the modeling of the geometric and coherence properties of the optical system as evidenced by the good agreement between our simulation and experiment.

Unsupervised Image Segmentation on 2D Echocardiogram

Echocardiography is a widely used, non-invasive imaging technique for diagnosing and monitoring heart conditions. However, accurate segmentation of cardiac structures, particularly the left ventricle, remains a complex task due to the inherent variability and noise in echocardiographic images. Current supervised models have achieved state-of-the-art results but are highly dependent on large, annotated datasets, which are costly and time-consuming to obtain and depend on the quality of the annotated data. These limitations motivate the need for unsupervised methods that can generalize across different image conditions without relying on annotated data. In this study, we propose an unsupervised approach for segmenting 2D echocardiographic images. By combining customized objective functions with convolutional neural networks (CNNs), our method effectively segments cardiac structures, addressing the challenges posed by low-resolution and gray-scale images. Our approach leverages techniques traditionally used outside of medical imaging, optimizing feature extraction through CNNs in a data-driven manner and with a new and smaller network design. Another key contribution of this work is the introduction of a post-processing algorithm that refines the segmentation to isolate the left ventricle in both diastolic and systolic positions, enabling the calculation of the ejection fraction (EF). This calculation serves as a benchmark for evaluating the performance of our unsupervised method. Our results demonstrate the potential of unsupervised learning to improve echocardiogram analysis by overcoming the limitations of supervised approaches, particularly in settings where labeled data are scarce or unavailable.

Learning super-resolution and pyramidal convolution residual network for vehicle re-identification

Vehicle re-identification (Vehicle Re-ID) aims at retrieving and tracking the specified target vehicle with multiple other cameras, which can provide help in checking violations and catching fugitives, but there are still the following problems that need to be solved urgently. First, the existing collected Vehicle Re-ID data often have low resolution and blur in local regions, so that the Vehicle Re-ID algorithm cannot accurately extract subtle feature representations. In addition, small features are easy to cause the disappearance of features under the operation of a large convolution kernel, which makes the model unable to capture and learn subtle features, resulting in inaccurate judgment of vehicles. In this study, we propose a Vehicle Re-ID method based on super resolution and pyramidal convolution residual network. Firstly, a super-resolution image generation network leveraging generative adversarial networks (GANs) is proposed. This network employs both content loss and adversarial loss as optimization criteria, ensuring an efficient transformation from a low-resolution image into a super-resolution counterpart, while meticulously preserving intricate high-frequency details. Then, multi levels of pyramidal convolution operations are designed to generate multi-scale features, which can capture information on different scales. Moreover, the concept of residual learning is applied between the multi levels of pyramidal convolution operations to expedite model optimization and enhance recognition capabilities. Ultimately, the double pyramidal convolutions are meticulously employed on both the original image and the super-resolution image, yielding low-noise feature representations and intricate semantic information respectively. By seamlessly fusing these two diverse sources of information, the resultant combined features exhibit heightened discrimination capabilities and significantly bolster the robustness of image features. In order to verify the effectiveness of the proposed method, extensive experiments are carried out on VeRi-776 and VehicleID datasets. The experimental results show that the method proposed in this paper effectively captures the detail information of vehicle images, accurately distinguishes the subtle differences between different vehicles of the same type, and is superior to state-of-the-art methods.

Differential Analysis of Modern Text Spotting Methods : A Systematic Review

Text spotting, the task of detecting and recognizing text within images, is vital in applications like document analysis, autonomous navigation, and surveillance. The motivation for this review arises from the growing need for accurate automated text extraction methods, driven by the surge of visual data and the complexity of real-world environments. Despite advances in deep learning and computer vision, current text spotting techniques face significant challenges, including handling complex backgrounds, curved or distorted text, varied font styles, and low-resolution images. These limitations restrict their effectiveness in diverse, real-world settings. This systematic review aims to conduct a differential analysis of modern text spotting methods, highlighting their strengths, weaknesses, and performance in addressing such challenges. The objectives are to evaluate state-of-the-art techniques, identify gaps in the field, and propose future research directions. By critically synthesizing recent literature, this review provides insights that can help enhance the robustness and accuracy of text spotting systems, making them more adaptable to real-world conditions.

Super resolution reconstruction of fluorescence microscopy images by a convolutional network with physical priors

Super-solution fluorescence microscopy, such as single-molecule localization microscopy (SMLM), is effective in observing subcellular structures and achieving excellent enhancement in spatial resolution in contrast to traditional fluorescence microscopy. Recently, deep learning has demonstrated excellent performance in SMLM in solving the trade-offs between spatiotemporal resolution, phototoxicity, and signal intensity. However, most of these researches rely on sufficient and high-quality datasets. Here, we propose a physical priors-based convolutional super-resolution network (PCSR), which incorporates a physical-based loss term and an initial optimization process based on the Wiener filter to create excellent super-resolution images directly using low-resolution images. The experimental results demonstrate that PCSR enables the achievement of a fast reconstruction time of 100 ms and a high spatial resolution of 10 nm by training on a limited dataset, allowing subcellular research with high spatiotemporal resolution, low cell phototoxic illumination, and high accessibility. In addition, the generalizability of PCSR to different live cell structures makes it a practical instrument for diverse cell research.

Joint feature selection and classification of low-resolution satellite images using the SAT-6 dataset

The modern industries of today demand the classification of satellite images, and to use the information obtained from it for their advantage and growth. The extracted information also plays a crucial role in national security and the mapping of geographical locations. The conventional methods often fail to handle the complexities of this process. so, an effective method is required with high accuracy and stability. In this paper, a new methodology named RankEnsembleFS is proposed that addresses the crucial issues of stability and feature aggregation in the context of the SAT-6 dataset. RankEnsembleFS makes use of a two-step process that consists of ranking the features and then selecting the optimal feature subset from the top-ranked features. RankEnsembleFS achieved comparable accuracy results to state-of-the-art models for the SAT-6 dataset while significantly reducing the feature space. This reduction in feature space is important because it reduces computational complexity and enhances the interpretability of the model. Moreover, the proposed method demonstrated good stability in handling changes in data characteristics, which is critical for reliable performance over time and surpasses existing ML ensemble methods in terms of stability, threshold setting, and feature aggregation. In summary, this paper provides compelling evidence that this RankEnsembleFS methodology presents excellent performance and overcomes key issues in feature selection and image classification for the SAT-6 dataset.

Self-learning based joint multi image super-resolution and sub-pixel registration

Multi Image Super-resolution (MISR) refers to the task of enhancing the spatial resolution of a stack of low-resolution (LR) images representing the same scene. Although many deep learning-based single image super-resolution (SISR) technologies have recently been developed, deep learning has not been widely exploited for MISR, even though it can achieve higher reconstruction accuracy because more information can be extracted from the stack of LR images. One of the primary obstacles encountered by deep networks when addressing the MISR problem is the variability in the number of LR images that act as input to the network. This impedes the feasibility of adopting an end-to-end learning approach, because the varying number of input images makes it difficult to construct a training dataset for the network. Another challenge arises from the requirement to align the LR input images to generate high-resolution (HR) image of high quality, which requires complex and sophisticated methods.In this paper, we propose a self-learning based method that can simultaneously perform super-resolution and sub-pixel registration of multiple LR images. The proposed method trains a neural network with only the LR images as input and without any true target HR images; i.e., the proposed method requires no extra training dataset. Therefore, it is easy to use the proposed method to deal with different numbers of input images. To our knowledge this is the first time that a neural network is trained using only LR images to perform a joint MISR and sub-pixel registration. Experimental results confirmed that the HR images generated by the proposed method achieved better results in both quantitative and qualitative evaluations than those generated by other deep learning-based methods.

Super-resolution imaging quality enhancement method for distributed array infrared camera

Abstract To address issues related to low resolution and high noise in infrared cameras, a distributed array infrared camera imaging system utilizing four cameras is proposed. The four cameras are arranged in an unconstrained array, and the combination algorithm of Projections onto Convex Sets (POCS) and Real-Enhanced Super-Resolution Generative Adversarial Networks (Real-ESRGAN) is applied to achieve high-quality super-resolution infrared imaging. The wavelet fusion algorithm is used to preprocess four low-resolution infrared images to reduce noise. Then, the POCS algorithm is used to reconstruct the preprocessed image. Finally, the Real-ESRGAN is employed for image reconstruction, resulting in an ultra-high-resolution infrared image. The results show that compared to single infrared camera imaging, the resolution of images reconstructed using the distributed infrared camera array is increased by 0.58 times, with the Modulation Transfer Function (MTF) increased by 1.2 times. Additionally, the entropy is increased by 18.87%, the standard deviation is increased by 13.51%, and the Naturalness Image Quality Evaluator (NIQE) is reduced by 18.87%. This demonstrates a significant enhancement in the super-resolution imaging quality of the distributed infrared camera array.

Correlative Multi-Scale Characterization of Nanoparticles Using Transmission Electron Microscopy

Chemical and physical properties of nanoparticles (NPs) are strongly influenced not only by the crystal structure of the respective material, including crystal structure defects but also by the NP size and shape. Contemporary transmission electron microscopy (TEM) can describe all these NP characteristics, however typically with a different statistical relevance. While the size and shape of NPs are frequently determined on a large ensemble of NPs and thus with good statistics, the characteristics on the atomic scale are usually quantified for a small number of individual NPs and thus with low statistical relevance. In this contribution, we present a TEM-based characterization technique, which can determine relevant characteristics of NPs in a scale-bridging way—from the crystal structure and crystal structure defects up to the NP size and morphology—with sufficient statistical relevance. This technique is based on a correlative multi-scale TEM approach that combines information on atomic scale obtained from the high-resolution imaging with the results of the low-resolution imaging assisted by a semi-automatic segmentation routine. The capability of the technique is illustrated in several examples, including Au NPs with different shapes, Au nanorods with different facet configurations, and multi-core iron oxide nanoparticles with a hierarchical structure.

Simultaneous single image super‐resolution and blind Gaussian denoising via slim ghost full‐frequency residual blocks

AbstractGiven that super‐resolution (SR) aims to recover lost information, and low‐resolution (LR) images in real‐world conditions might be corrupted with multiple degradations, considering basic bicubic down‐sampling as the sole degradation significantly limits the performance of most existing SR models. This paper presents a model for simultaneous super‐resolution and blind additive white Gaussian noise (AWGN) denoising with two components (netdeg and netSR) that is based on a generative adversarial network (GAN) to achieve detailed results. netdeg, featuring residual and innovative cost‐effective ghost residual blocks with a frequency separation module for obtaining long‐range information, blindly restores a clean version of the LR image. netSR leverages slim ghost full‐frequency residual blocks to process low‐frequency (LF) and high‐frequency (HF) information via static large convolutions and pixel‐wise highlighted input‐adaptive dynamic convolutions, respectively. To address the susceptibility of dynamic layers to noise and preserve feature diversity while reducing model’s costs, static and dynamic layer features are combined and highlighted. Diverse and non‐redundant features are then processed using ghost‐style blocks. The proposed model achieves comparable SR results in bicubic down‐sampling scenarios, outperform existing SR methods in the complex task of concurrent SR and AWGN denoising, and demonstrate robustness in handling images corrupted with varying levels of AWGN.

A Feature-Driven Inception Dilated Network for Infrared Image Super-Resolution Reconstruction

Image super-resolution (SR) algorithms based on deep learning yield good visual performances on visible images. Due to the blurred edges and low contrast of infrared (IR) images, methods transferred directly from visible images to IR images have a poor performance and ignore the demands of downstream detection tasks. Therefore, an Inception Dilated Super-Resolution (IDSR) network with multiple branches is proposed. A dilated convolutional branch captures high-frequency information to reconstruct edge details, while a non-local operation branch captures long-range dependencies between any two positions to maintain the global structure. Furthermore, deformable convolution is utilized to fuse features extracted from different branches, enabling adaptation to targets of various shapes. To enhance the detection performance of low-resolution (LR) images, we crop the images into patches based on target labels before feeding them to the network. This allows the network to focus on learning the reconstruction of the target areas only, reducing the interference of background areas in the target areas’ reconstruction. Additionally, a feature-driven module is cascaded at the end of the IDSR network to guide the high-resolution (HR) image reconstruction with feature prior information from a detection backbone. This method has been tested on the FLIR Thermal Dataset and the M3FD Dataset and compared with five mainstream SR algorithms. The final results demonstrate that our method effectively maintains image texture details. More importantly, our method achieves 80.55% mAP, outperforming other methods on FLIR Dataset detection accuracy, and with 74.7% mAP outperforms other methods on M3FD Dataset detection accuracy.

Multi-step reconstruction of radio-interferometric images

The large aperture arrays for the currently under construction SKA Observatory (SKAO) will allow for observations of the universe in the radio spectrum at unprecedented resolution and sensitivity. However, these telescopes will produce data on the scale of exabytes, introducing a slew of hardware and software design challenges. This paper proposes a multi-step image reconstruction framework that allows for partitioning of visibility data by baseline length. This enables more flexible data distribution and parallelization, aiding in processing radio-astronomical observations within given constraints. Additionally, as each step of the framework only relies on a subset of the total visibilities, one can perform reconstruction progressively, with the initial step performed on the SKAO Science Data Processors and the second on local clusters. The multi-step reconstruction is separated into two steps. First a low-resolution image is reconstructed with only short-baseline visibilities, and then using this image together with the long-baseline visibilities, the full-resolution image is reconstructed. The proposed method only operates in the minor cycle, and it can be easily integrated into existing imaging pipelines. We show that our proposed method allows for partitioning of visibilities by baseline without introducing significant additional drawbacks, reconstructing images of similar quality within similar numbers of major cycles compared to a single-step all-baselines approach that uses the same reconstruction method as well as compared to multi-scale CLEAN.