Image Quality Enhancement Research Articles

Enhanced three-dimensional imaging with large depth of field using pre-processing in optical scanning holography

Optical scanning holography (OSH) emerges as a groundbreaking single-pixel real-time holographic recording technique, charting new territory beyond conventional digital holography. A notable challenge in OSH, as with all three-dimensional (3D) imaging methods, is the susceptibility of reconstructed images corrupted by defocus noise. The noise is an unintended result of diffraction on the focused plane from out-of-focus planes of the three-dimensional object, which can degrade image quality upon holographic reconstruction. Current methods, however, all use some algorithms to post-process holographic data to achieve the purpose of eliminating out-of-focus noise, which is difficult to meet the requirements of real-time image processing. We propose an innovative pre-processing approach in optical scanning holography that not only eliminates defocus noise but also significantly enhances image quality. Preprocessing is accomplished by the utilization of a focus-tunable lens (FTL) to refocus the optical scanning beam according to the depth of the occlusion-free object. Complex holograms are then captured without the artifacts associated with defocus through real-time holographic data pre-processing. Numerical simulations and experimental results demonstrate that our proposed method can directly obtain high-quality reconstructed images with all the 3D information being in focus simultaneously.

A Hybrid Transfer Learning Framework for Brain Tumor Diagnosis

Brain tumors are among the most severe health challenges, necessitating early and precise diagnosis for effective treatment planning. This study introduces an optimized hybrid transfer learning (TL) framework for brain tumor classification using magnetic resonance imaging images. The proposed system integrates advanced preprocessing techniques, an ensemble of pretrained deep learning models, and explainable artificial intelligence (XAI) methods to achieve high accuracy and reliability. The methodology enhances image quality through noise reduction and contrast enhancement, facilitating robust feature extraction. The ensemble model combines VGG16 and ResNet152V2 architectures, achieving a classification accuracy of 99.47% on a challenging four‐class dataset. Additionally, gradient‐weighted class activation mapping and SHapley Additive exPlanations (SHAP)‐based XAI techniques provide visual and quantitative insights into model predictions, improving interpretability and clinical trust. This comprehensive framework demonstrates the potential of hybrid TL and XAI in advancing diagnostic accuracy and supporting clinical decision‐making for brain tumor detection. The results underscore its applicability in clinical settings, particularly in resource‐constrained environments.

A comparison of carbon ions versus protons for integrated mode ion imaging.

Incorporating image guidance into ion beam therapy is critical for minimizing beam range uncertainties and realizing the modality's potential. One promising avenue for image guidance is to capture transmission ion radiographs (iRads) before and/or during treatment. iRad image quality is typically maximized using a single-event imaging system, which involves tracking individual ions, albeit the approach is generally not suited to clinical beam settings. An alternative faster and clinically compatible method is integrated mode imaging, where individual pencil beam data is acquired, rather than single ion data. To evaluate the usefulness of transmission ion imaging for image guidance, it is crucial to evaluate the image quality of integrated mode iRad systems. We report extensive image quality metrics of integrated mode carbon ion radiographs (cRads) and compare them with proton radiographs (pRads). iRads were obtained at the Marburg Ion Beam Therapy Center using a plastic volumetric scintillator equipped with CCD cameras. The detector captures orthogonal views of the 3D energy deposition in the scintillator from individual pencil beams. Four phantoms were scanned using a field of view and a beam spacing of 1 mm. First, 9 tissue-substitute inserts were used to evaluate water equivalent thickness (WET) accuracy. Radiographs of those inserts were reconstructed for beam spacings ranging from 1 to 7 mm to evaluate the impact of spacing on quantitative accuracy. For spatial resolution, custom 3D printed line pair (lp) modules ranging from 0.5 to 10 lp/cm were scanned. To evaluate low contrast detectability, a custom 3D printed low contrast module consisting of 20 holes with depths ranging from 1 to 8 mm and diameters from 1 to 10 mm was scanned. iRads of an anthropomorphic head phantom were alsoobtained. Spatial resolution and low contrast detection are systematically improved for cRads compared to pRads. Image resolution was 3.7 lp/cm for cRads and 1.7 lp/cm for pRads in the center of the field of view. Spatial resolution was found to vary with the object's location in the field of view. While pRads could mostly resolve low contrast holes of 10 mm in diameter, cRads could resolve holes of up in 4 mm diameter. WET accuracy is similar for both ion species, with a root mean squared error of approximately 1 mm. WET accuracy was stable (maximum of 0.1 mm increase) across beam spacings, although important under-sampling artifacts were observed for iRads reconstructed using large beam spacings, especially for cRads. iRads of the anthropomorphic head phantom showed improved apparent contrast using cRads, especially to identify bonystructures. This work is the first investigation of image quality metrics such as spatial resolution and low contrast detectability for integrated mode cRads, with a full comparison with pRads. Enhanced image quality is obtained with cRads compared to pRads, although pRads still maintain high WET accuracy and deliver image quality within acceptablebounds.

Prosedur Pemeriksaan MRI Cruris dengan Kontras pada Kasus Tumor

Background: This study examines the MRI Cruris contrast examination procedure for tumor cases at Dr. Moewardi Hospital. The hospital's MRI protocol includes sequences such as 3-plane localizer, Sag T1, Sag T2, Ax T1, Ax T2 STIR, Coronal T1, Coronal PD FS, Coronal T2* MERGE, 3D Coronal TRICKS, and post-contrast sequences Sag T1+C, Coronal T1+C, and Ax T1+C. The study aims to analyze the MRI Cruris contrast procedure, the rationale for using 5 ml contrast, and the importance of the Coronal 3D TRICKS sequence.Methods: A qualitative research method with a case study approach was used, involving observation, interviews, and documentation. The data was analyzed and presented in quotations to conclude.Result and Discussion: The MRI procedure begins with patient preparation, including laboratory tests for urea and creatinine levels, fasting for 4–6 hours, and filling out informed consent. Patients are positioned supine with feet first, and an air coil is placed on the leg. The use of 5 ml contrast enhances image quality, improves tumor detection, evaluates blood vessels, shows lesions, and determines tumor location and spread. The Coronal 3D TRICKS sequence provides clear MRA angiography, enabling better visualization of feeding arteries and rapid imaging of major blood vessels within 10 seconds.Conclusion: In conclusion, the MRI Cruris contrast protocol at Dr. Moewardi Hospital ensures high-quality imaging for tumor detection and diagnosis. The 5 ml contrast dose effectively enhances MRI images, and the Coronal 3D TRICKS sequence plays a crucial role in evaluating vascular structures.

Data-driven signal-to-noise enhancement in scattering near-field infrared microscopy.

This study introduces a machine-learning approach to enhance signal-to-noise ratios in scattering-type scanning near-field optical microscopy (s-SNOM). While s-SNOM offers a high spatial resolution, its effectiveness is often hindered by low signal levels, particularly in weakly absorbing samples. To address these challenges, we utilize a data-driven "patch-based" machine learning reconstruction method, incorporating modern generative adversarial neural networks (CycleGANs) for denoising s-SNOM images. This method allows for flexible reconstruction of images of arbitrary sizes, a critical capability given the variable nature of scanned sample areas in point-scanning probe-based microscopies. The CycleGAN model is trained on unpaired sets of images captured at both rapid and extended acquisition times, thereby modeling instrument noise while preserving essential topographical and molecular information. The results show significant improvements in image quality, as indicated by higher structural similarity index and peak signal-to-noise ratio values, comparable to those obtained from images captured with four times the integration time. This method not only enhances image quality but also has the potential to reduce the overall data acquisition time, making high-resolution s-SNOM imaging more feasible for a wide range of biological and materials science applications.

Optimizing AR Waveguide Structures to Enhance Imaging Quality

Optimizing AR Waveguide Structures to Enhance Imaging Quality

Deep learning CT image restoration using system blur and noise models.

The restoration of images affected by blur and noise has been widely studied and has broad potential for applications including in medical imaging modalities such as computed tomography. Recently, deep learning approaches have demonstrated the potential to enhance image quality beyond classic limits; however, most deep learning models attempt a blind restoration problem and base their restoration on image inputs alone without direct knowledge of the image noise and blur properties. We present a method that leverages both degraded image inputs and a characterization of the system's blur and noise to combine modeling and deep learning approaches. Different methods to integrate these auxiliary inputs are presented, namely, an input-variant and a weight-variant approach wherein the auxiliary inputs are incorporated as a parameter vector before and after the convolutional block, respectively, allowing easy integration into any convolutional neural network architecture. The proposed model shows superior performance compared with baseline models lacking auxiliary inputs. Evaluations are based on the average peak signal-to-noise ratio and structural similarity index measure, selected examples of top and bottom 10% performance for varying approaches, and an input space analysis to assess the effect of different noise and blur on performance. In addition, the proposed model exhibits a degree of robustness when the blur and noise parameters deviate from their true values. Results demonstrate the efficacy of providing a deep learning model with auxiliary inputs, representing system blur and noise characteristics, to enhance the performance of the model in image restoration tasks.

Efficient Liver Segmentation using Advanced 3D-DCNN Algorithm on CT Images

According to the latest global cancer statistics for 2022, liver cancer ranks as the ninth most common disease in women. Segmenting the liver and distinguishing it from tumors within it pose a significant challenge due to the complex nature of liver imaging. Common imaging methods such as Magnetic Resonance Imaging (MRI), Computer Tomography (CT), and Ultrasound (US) are employed to distinguish liver tissue from liver tumors after collecting a sample. Attempting to partition the liver and tumor based on grayscale shades or shapes in abdominal CT images is not ideal because of the overlapping intensity levels and the variability in the location and shape of soft tissues. To address this issue, this study introduces an effective method for liver image segmentation using a 3D deep Convolutional Neural Network (3D-DCNN). The process involves several stages. First, liver images undergo preprocessing to enhance image quality, including median filtering, adaptive filtering, and converting them to grayscale. The feature extraction phase focuses on extracting four sets of features, such as the Local Binary Pattern (LBP) and the Gray-Level Co-occurrence Matrix (GLCM). Additionally, an Iterative Region Growing (IRG) technique is developed to improve the Dice Similarity Coefficient (DSC) prediction by enhancing the quality of the input images obtained from segmented images. This method enables the segmentation of the liver in abdominal CT image volumes and can subsequently be used to segment liver tumor images to evaluate the performance of the proposed 3D-DLNN approach. This method was implemented in MATLAB, and its performance was evaluated using various metrics. In experimental analysis, the proposed technique outperformed other methods, including Jaccard with JISTS-FCM, Fuzzy C-Means (FCM), and FCM with Cluster Size Adjustment (FCM-CSA).

Image synthesis with class-aware semantic diffusion models for surgical scene segmentation.

Surgical scene segmentation is essential for enhancing surgical precision, yet it is frequently compromised by the scarcity and imbalance of available data. To address these challenges, semantic image synthesis methods based on generative adversarial networks and diffusion models have been developed. However, these models often yield non-diverse images and fail to capture small, critical tissue classes, limiting their effectiveness. In response, a class-aware semantic diffusion model (CASDM), a novel approach which utilizes segmentation maps as conditions for image synthesis to tackle data scarcity and imbalance is proposed. Novel class-aware mean squared error and class-aware self-perceptual loss functions have been defined to prioritize critical, less visible classes, thereby enhancing image quality and relevance. Furthermore, to the authors' knowledge, they are the first to generate multi-class segmentation maps using text prompts in a novel fashion to specify their contents. These maps are then used by CASDM to generate surgical scene images, enhancing datasets for training and validating segmentation models. This evaluation assesses both image quality and downstream segmentation performance, demonstrates the strong effectiveness and generalisability of CASDM in producing realistic image-map pairs, significantly advancing surgical scene segmentation across diverse and challengingdatasets.

Evaluation of Image Quality and Scan Time Efficiency in Accelerated 3D T1-Weighted Pediatric Brain MRI Using Deep Learning-Based Reconstruction.

This study evaluated the effect of an accelerated three-dimensional (3D) T1-weighted pediatric brain MRI protocol using a deep learning (DL)-based reconstruction algorithm on scan time and image quality. This retrospective study included 46 pediatric patients who underwent conventional and accelerated, pre- and post-contrast, 3D T1-weighted brain MRI using a 3T scanner (SIGNA Premier; GE HealthCare) at a single tertiary referral center between March 1, 2023, and April 30, 2023. Conventional scans were reconstructed using intensity Filter A (Conv), whereas accelerated scans were reconstructed using intensity Filter A (Fast_A) and a DL-based algorithm (Fast_DL). Image quality was assessed quantitatively based on the coefficient of variation, relative contrast, apparent signal-to-noise ratio (aSNR), and apparent contrast-to-noise ratio (aCNR) and qualitatively according to radiologists' ratings of overall image quality, artifacts, noisiness, gray-white matter differentiation, and lesion conspicuity. The acquisition times for the pre- and post-contrast scans were 191 and 135 seconds, respectively, for the conventional scan. With the accelerated protocol, these were reduced to 135 and 80 seconds, achieving time reductions of 29.3% and 40.7%, respectively. DL-based reconstruction significantly reduced the coefficient of variation, improved the aSNR, aCNR, and overall image quality, and reduced the number of artifacts compared with the conventional acquisition method (all P < 0.05). However, the lesion conspicuity remained similar between the two protocols. Utilizing a DL-based reconstruction algorithm in accelerated 3D T1-weighted pediatric brain MRI can significantly shorten the acquisition time, enhance image quality, and reduce artifacts, making it a viable option for pediatric imaging.

Enhancing single image dehazing with self-supervised convolutional neural network and dark channel prior integration

The removal of noise from images holds great significance as clear and denoised images are vital for various applications. Recent research efforts have been concentrated on the dehazing of single images. While conventional methods and deep learning approaches have been employed for daytime images, learning-based techniques have shown impressive dehazing results, albeit often with increased complexity. This has led to the persistence of prior-based methods, despite their slightly lower performance. To address this issue, we propose a novel deep learning-based dehazing method utilizing a self-supervised convolutional neural network (CNN). This approach incorporates both the input hazy image and the dark channel prior. By leveraging an encoder, the combined information of the dark channel prior and haze image is encoded into a condensed latent representation. Subsequently, a decoder is employed to reconstruct the clean image using these latent features. Our experimental results demonstrate that our proposed algorithm significantly enhances image quality, as indicated by improved peak signal-to-noise ratio (PSNR) and structural similarity index (SSIM) values. We perform both quantitative and qualitative comparisons with recently published techniques, showcasing the efficacy of our approach.

Thermographic images for screening oral health problems in older adults: A pilot study

ObjectiveThis pilot study assessed the feasibility of using facial thermography to detect intra-oral problems in older adults with cognitive decline and care-resistant behaviors, who are often unable to communicate pain or early symptoms. MethodsTwenty-three older adults (mean age 73.7 ± 13.2 years) with cognitive decline were enrolled. Thermal images of four facial views were taken using a smartphone-connected FLIR thermal camera. Intra-oral examinations were conducted, and the thermographic data were analyzed to extract temperature values in the regions of interest (ROI). Point-biserial correlations and ROC curve analyses were performed to evaluate associations between temperature data and clinical findings, with a significance level of p<0.05. ResultsIntra-oral issues requiring treatment were found in 12 participants, with six reporting clinical pain. The overall mean temperature in the ROI was 33.5 ± 3.9°C, and significant temperature differences were found between the body and ROI temperatures. Correlations were observed between clinical problems and various temperature metrics, including the minimum and maximum ROI temperatures (rpb=-0.327, p=0.002 and rpb=-0.309, p=0.003). ROC analysis indicated that ROI temperature could predict the presence or absence of clinical problems, with AUC values ranging from 0.651 to 0.796 for different metrics. ConclusionThermographic facial imaging shows significant potential as a non-invasive tool for detecting oral health problems in vulnerable older adults. While promising, further research is essential to enhance image quality, streamline the technique, and incorporate AI for improved diagnostic accuracy and ease of use. Clinical SignificanceThis non-invasive, inexpensive technique is easy to perform, independent of patient compliance and, is promising to detect early oral problems in noncommunicative patients.

Enhancing Face Recognition Dataset using Generative Artificial Intelligence: Review Study

Face recognition is a fast growing and crucial area of study due to its applicability in different life applications. Face recognition systems often face significant challenges due to poor-quality datasets, which can result from factors such as low illumination, low resolution, facial distortions, and varying environmental conditions. These limitations hinder the performance and reliability of face recognition technologies. Generative Adversarial Networks (GANs) have emerged as a powerful tool to address these issues by enhancing image quality and generating realistic facial representations under challenging conditions. This review paper explores various GAN-based models each designed to tackle specific face recognition challenges. These models demonstrate significant improvements in enhancement ratios, ranging from 75% to 97%, by leveraging advanced generative techniques to reconstruct and augment facial images. Beyond face recognition, the methodologies discussed in this research have broader implications for enhancing image datasets in other domains, such as medical imaging, surveillance, and autonomous driving, where poor data quality is a common issue. By improving dataset quality through generative AI, this research paves the way for more robust and accurate machine learning applications across diverse fields.

A three-dimensional marine plastic litter real-time detection embedded system based on deep learning.

Marine plastic pollution has emerged as a significant ecological and biological issue impacting global marine ecosystems. To develop real-time cleaning systems for marine plastic litter, we implemented a three-dimensional marine plastic litter real-time detection (3D-MPLRD) system that incorporates deep learning techniques. Specifically, image quality assessment and enhancement are applied to mitigate the adverse effects of harsh underwater conditions on image quality, thereby enhancing the training efficacy of models beyond that achieved by conventional approaches. To facilitate the deployment of the 3D-MPLRD system on embedded devices, the YOLOv5 model is compressed and quantified, capitalizing on its robustness to precision loss and model simplification. Practical experimental results demonstrated that the 3D-MPLRD system outperformed the model trained on original datasets in terms of precision, recall, F1-score, and mean average precision. This innovative 3D-MPLRD system serves as a foundational reference for marine environmental protection based on intelligence systems, as opposed to traditional manual methods.

Multiscale local sparsity and prior learning algorithm for Cherenkov-excited luminescence scanned tomography reconstruction

Cherenkov-excited luminescence scanned tomography (CELST) is an emerging imaging technique and its potential applications during radiation therapy have just recently been explored. The aim of CELST is to recover the distribution of luminescent probes from emission photons. However, CELST images tend to suffer from low resolution and degraded image quality due to light multiple scattering and limited boundary measurements. Therefore, inaccurate information about the status of the luminescent probe is provided. To accurately capture the sparsity characterization of a luminescent probe and achieve the high-quality image, a novel reconstruction method, to our knowledge, is proposed for CELST by combining a sparse prior with an attention network, termed LKSVD-Net. A multiscale learned KSVD is first incorporated to obtain the local sparsity information of a luminescent probe. Subsequently, a prior attention network is designed to leverage the prior features related to the measurements. The multiscale sparsity and prior features are finally combined to complete the image reconstruction. Experimental results demonstrate that the LKSVD-Net can notably enhance image quality even in a 20 dB signal-to-noise ratio (SNR). Furthermore, the proposed LKSVD-Net yields improved quantitative accuracy for 4 mm diameter probes with an edge-to-edge distance of 2 mm. The results demonstrate that LKSVD-Net improves the peak signal-to-noise ratio (PSNR) by approximately 15.1%, structural similarity index measure (SSIM) by about 95.8%, and Pearson correlation (PC) by around 3% compared to Tikhonov regularization.

Deep learning reconstruction of zero-echo time sequences to improve visualization of osseous structures and associated pathologies in MRI of cervical spine

ObjectivesTo determine whether deep learning-based reconstructions of zero-echo-time (ZTE-DL) sequences enhance image quality and bone visualization in cervical spine MRI compared to traditional zero-echo-time (ZTE) techniques, and to assess the added value of ZTE-DL sequences alongside standard cervical spine MRI for comprehensive pathology evaluation.MethodsIn this retrospective study, 52 patients underwent cervical spine MRI using ZTE, ZTE-DL, and T2-weighted 3D sequences on a 1.5-Tesla scanner. ZTE-DL sequences were reconstructed from raw data using the AirReconDL algorithm. Three blinded readers independently evaluated image quality, artifacts, and bone delineation on a 5-point Likert scale. Cervical structures and pathologies, including soft tissue and bone components in spinal canal and neural foraminal stenosis, were analyzed. Image quality was quantitatively assessed by signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR).ResultsMean image quality scores were 2.0 ± 0.7 for ZTE and 3.2 ± 0.6 for ZTE-DL, with ZTE-DL exhibiting fewer artifacts and superior bone delineation. Significant differences were observed between T2-weighted and ZTE-DL sequences for evaluating intervertebral space, anterior osteophytes, spinal canal, and neural foraminal stenosis (p < 0.05), with ZTE-DL providing more accurate assessments. ZTE-DL also showed improved evaluation of the osseous components of neural foraminal stenosis compared to ZTE (p < 0.05).ConclusionsZTE-DL sequences offer superior image quality and bone visualization compared to ZTE sequences and enhance standard cervical spine MRI in assessing bone involvement in spinal canal and neural foraminal stenosis.Critical relevance statementDeep learning-based reconstructions improve zero-echo-time sequences in cervical spine MRI by enhancing image quality and bone visualization. This advancement offers additional insights for assessing bone involvement in spinal canal and neural foraminal stenosis, advancing clinical radiology practice.Key PointsConventional MRI encounters challenges with osseous structures due to low signal-to-noise ratio.Zero-echo-time (ZET) sequences offer CT-like images of the C-spine but with lower quality.Deep learning reconstructions improve image quality of zero-echo-time sequences.ZTE sequences with deep learning reconstructions refine cervical spine osseous pathology assessment.These sequences aid assessment of bone involvement in spinal and foraminal stenosis.Graphical

Patch-based dual-domain photon-counting CT data correction with residual-based WGAN-ViT.

X-ray photon-counting detectors (PCDs) have recently gained popularity due to their capabilities in energy discrimination power, noise suppression, and resolution refinement. The latest extremity photon-counting computed tomography (PCCT) scanner leverages these advantages for tissue characterization, material decomposition, beam hardening correction, and metal artifact reduction. However, technical challenges such as charge splitting and pulse pileup can distort the energy spectrum and compromise image quality. Also, there is a clinical need to balance radiation dose and imaging speed for contrast-enhancement and other studies. This paper aims to address these challenges by developing a dual-domain correction approach to enhance PCCT reconstruction quality quantitatively and qualitatively. We propose a novel correction method that operates in both projection and image domains. In the projection domain, we employ a residual-based Wasserstein generative adversarial network (R-WGAN) to capture local and global features, suppressing pulse pileup, charge splitting, and data noise. This is facilitated with traditional filtering methods in the image domain to enhance signal-to-noise ratio while preserving texture across each energy channel. To address GPU memory constraints, our approach utilizes a patch-based volumetric refinement network. Our dual-domain correction approach demonstrates significant fidelity improvements across both projection and image domains. Experiments on simulated and real datasets reveal that the proposed model effectively suppresses noise and preserves intricate details, outperforming the state-of-the-art methods. This approach highlights the potential of dual-domain PCCT data correction to enhance image quality for clinical applications, showing promise for advancing PCCT image fidelity and applicability in preclinical/clinical environments.

Post-migration seismic data conditioning workflow on a merged dataset

Analyzing amplitude anomalies in seismic data requires a comprehensive understanding of the geological context and the accuracy of the seismic image. Over the past four decades, numerous surveys in mature basins like the US Gulf of Mexico have undergone reprocessing and merging to enhance imaging quality. The merging of seismic data volumes demands careful attention during processing, as the different volumes are often acquired at different times with different hardware, acquisition geometries, and exploration objectives. If insufficient care is taken, significant differences in the amplitude and spectra of the merged survey components can pose challenges when used as input for machine learning techniques or seismic attribute interpretation and studies. We implemented spectral balancing followed by structure-oriented filtering to address the abovementioned discrepancies in a merged survey. Spectral balancing equalizes high and low frequencies, creating a more uniform frequency spectrum. Structure-oriented filtering eliminates random and cross-cutting coherent noise while preserving structural and stratigraphic features. This workflow ameliorates the discrepancies between the areas covered by the individual surveys, resulting in a more consistent interpretation across the seam between the two surveys. This study found that these two processes improved vertical resolution and lateral delineation of fault and stratigraphic edges. However, we also observed in our data that increasing the spectrum runs the risk of increasing the effect of both low and high-frequency noise, including acquisition footprint. Surprisingly, we found that spectral balancing can diminish the appearance of stratigraphic edges that were fortuitously imaged by the original narrowband data volume. Finally, because both spectral balancing and structure-oriented filtering are set to not damage relative amplitudes, we need to apply amplitude gain control to balance them on the shallower parts of the merged survey.

Improving NIR single-pixel imaging: using deep image prior and GANs

We introduce a hybrid approach that combines deep image prior (DIP) with generative adversarial networks (GANs) to improve the resolution of single-pixel imaging (SPI). SPI excels in challenging conditions such as low light or limited spectral camera availability, particularly in the near-infrared (NIR) range from 850 to 1550 nm. By employing an unsupervised image super-resolution technique based on DIP, we reduce the need for extensive direct SPI image datasets. This innovation simplifies enhancing image quality in specific NIR bands. We provide numerical and experimental evidence to support our method and detail the enhancements in UNet and GAN architectures across four neural network configurations.

Enhancing Image Quality in Crop Disease Diagnosis: A Comparative Evaluation of Laplacian and Average Filtering with Established Techniques

The agricultural sector is vital to ensuring global food security, yet plant diseases caused by various environmental factors lead to significant reductions in crop yields. Early detection of such diseases is critical, as crop health directly affects both yield and quality. This research introduces a novel pre-processing algorithm designed to enhance the accuracy of crop disease detection from leaf images. The proposed algorithm is compared against three widely used filtering techniques: Median filtering, Wiener filtering, and Gaussian filtering. Before implementing the proposed approach, a detailed assessment of these conventional methods is conducted.The algorithm integrates Laplacian filtering and average filtering to optimize image pre-processing. The effectiveness of this approach is evaluated using performance metrics such as Peak Signal-to-Noise Ratio (PSNR), Mean Squared Error (MSE), and Mutual Information (MI). Results indicate that the proposed method achieves a 0.34% increase in PSNR, a 0.92% reduction in MSE, and a 0.48% improvement in MI compared to existing techniques. These improvements underscore the algorithm's ability to enhance image quality and outperform traditional methods.The findings suggest that incorporating Laplacian and average filtering into the pre-processing pipeline introduces a more efficient methodology for image enhancement. This approach holds potential for advancing image analysis and interpretation in agriculture and other fields, providing a foundation for further innovation in image processing technologies.