High Detailed Images Research Articles

High-Quality Text-to-Image Generation Using High-Detail Feature-Preserving Network

Multistage text-to-image generation algorithms have shown remarkable success. However, the images produced often lack detail and suffer from feature loss. This is because these methods mainly focus on extracting features from images and text, using only conventional residual blocks for post-extraction feature processing. This results in the loss of features, greatly reducing the quality of the generated images and necessitating more resources for feature calculation, which will severely limit the use and application of optical devices such as cameras and smartphones. To address these issues, the novel High-Detail Feature-Preserving Network (HDFpNet) is proposed to effectively generate high-quality, near-realistic images from text descriptions. The initial text-to-image generation (iT2IG) module is used to generate initial feature maps to avoid feature loss. Next, the fast excitation-and-squeeze feature extraction (FESFE) module is proposed to recursively generate high-detail and feature-preserving images with lower computational costs through three steps: channel excitation (CE), fast feature extraction (FFE), and channel squeeze (CS). Finally, the channel attention (CA) mechanism further enriches the feature details. Compared with the state of the art, experimental results obtained on the CUB-Bird and MS-COCO datasets demonstrate that the proposed HDFpNet achieves better performance and visual presentation, especially regarding high-detail images and feature preservation.

High definition 3D models of small legume seed: a close up application to support ancient plant genomics

Abstract. Legume seeds are essential for nutrition and play crucial roles in food security, climate change mitigation, biodiversity conservation, and sustainability. Among them, the common bean (Phaseolus vulgaris L.) is the most widely cultivated and vital to the food value chain. Originating in Mesoamerica and independently domesticated there and in the Andes about 8,000 years ago, common beans have adapted to diverse environments globally following the Columbian exchange. Ancient DNA (aDNA) sequencing offers insights into the bean’s adaptation and evolution, but traditional methods struggle with detailed 3D phenotypic analysis due to the beans’ small size and the destructive nature of aDNA extraction. Expensive high-detail optical instruments are not commonly available in genetic labs. This research proposes macro-photogrammetry as a low-cost technique for creating detailed 3D digital replicas of beans. By using samples from archaeological sites in northern Peru, this method preserves phenotypic information despite the destruction of physical samples during DNA extraction. The study details the process of capturing high-detail images with specialized equipment to build a digital phenotype library, preserving the morphological features of ancient beans for further study and future comparison with modern varieties.

Dense Convolutional Neural Network-Based Deep Learning Pipeline for Pre-Identification of Circular Leaf Spot Disease of Diospyros kaki Leaves Using Optical Coherence Tomography.

Circular leaf spot (CLS) disease poses a significant threat to persimmon cultivation, leading to substantial harvest reductions. Existing visual and destructive inspection methods suffer from subjectivity, limited accuracy, and considerable time consumption. This study presents an automated pre-identification method of the disease through a deep learning (DL) based pipeline integrated with optical coherence tomography (OCT), thereby addressing the highlighted issues with the existing methods. The investigation yielded promising outcomes by employing transfer learning with pre-trained DL models, specifically DenseNet-121 and VGG-16. The DenseNet-121 model excels in differentiating among three stages of CLS disease (healthy (H), apparently healthy (or healthy-infected (HI)), and infected (I)). The model achieved precision values of 0.7823 for class-H, 0.9005 for class-HI, and 0.7027 for class-I, supported by recall values of 0.8953 for class-HI and 0.8387 for class-I. Moreover, the performance of CLS detection was enhanced by a supplemental quality inspection model utilizing VGG-16, which attained an accuracy of 98.99% in discriminating between low-detail and high-detail images. Moreover, this study employed a combination of LAMP and A-scan for the dataset labeling process, significantly enhancing the accuracy of the models. Overall, this study underscores the potential of DL techniques integrated with OCT to enhance disease identification processes in agricultural settings, particularly in persimmon cultivation, by offering efficient and objective pre-identification of CLS and enabling early intervention and management strategies.

Optic coherence tomography-angiography – a new technique in the diagnosis and follow-up of the patients with age-related macular degeneration – an overview

Age-related macular degeneration (AMD) is a degenerative, progressive disease, which destroys the central retina. AMD has two main forms - „dry“ and „wet“ (exudative). In order to make the diagnosis of AMD with CNV, apart from ophthalmoscopy, specialized tests are needed - fluorescein angiography (FA), indocyanine angiography, OCT, OCT-angiography (OCT-A). OCTA is a new and non-invasive technology. Compared to the old, well-known techniques for retinal evaluation, it is very precious because of its high resolution and detailed images – which allow early diagnosis and objective monitoring of the lesions in the future. The aim of this review is to describe the principles of function, features of the device and its application in everyday clinical practice in diagnosing patients with macular pathology. A systematic search of the literature, published in the past decade was identified from PubMed and Ovid databases with reference to the Preferred Reporting items for Systematic Review. AMD is a disease that is increasingly socially important because of the aging population in developed countries. The increasing use of OCT-A nowadays suggests new classification systems. The non-invasiveness, the short examination time and the high informativeness that OCT-A brings in relation to AMD requires a deeper study of the methodology and the development of new classifications to facilitate clinical practice.

Conditional generative model with skip-connection structure for low-light image enhancement

Vision-guided coal mine robots often encounter challenges in low-light environments, capturing images marked by poor visibility and significant loss of detail, which complicates the advancement of smart mine safety production. Despite various advancements, achieving high visibility and detailed images through robotic vision sensors remains a formidable task. This paper introduces a conditional generative model incorporating a skip-connection structure to address these issues. We present a novel approach utilizing conditional Generative Adversarial Networks (GNAs) aimed at improving the visibility of image content. This involves an encoder-decoder framework with skip-connections functioning as the generator, and the discriminator, both tailored to enhance image detail. In addition, we implement loss functions integrated with boundary equilibrium constraints to counter the issue of model collapse in conditional GANs, thereby enhancing training stability and image fidelity. Our experimental findings demonstrate that our model delivers competitive results against other state-of-the-art low-light image enhancement models, alongside a standardized training regimen that ensures rapid and consistent convergence.

Dualization and Automatic Distributed Parameter Selection of Total Generalized Variation via Bilevel Optimization

Total Generalized Variation (TGV) regularization in image reconstruction relies on an infimal convolution type combination of generalized first- and second-order derivatives. This helps to avoid the staircasing effect of Total Variation (TV) regularization, while still preserving sharp contrasts in images. The associated regularization effect crucially hinges on two parameters whose proper adjustment represents a challenging task. In this work, a bilevel optimization framework with a suitable statistics-based upper level objective is proposed in order to automatically select these parameters. The framework allows for spatially varying parameters, thus enabling better recovery in high-detail image areas. A rigorous dualization framework is established, and for the numerical solution, a Newton type method for the solution of the lower level problem, i.e. the image reconstruction problem, and a bilevel TGV algorithm are introduced. Denoising tests confirm that automatically selected distributed regularization parameters lead in general to improved reconstructions when compared to results for scalar parameters.

Optimization of fuzzy c-means (FCM) clustering in cytology image segmentation using the gray wolf algorithm

BackgroundImage segmentation is considered an important step in image processing. Fuzzy c-means clustering is one of the common methods of image segmentation. However, this method suffers from drawbacks, such as sensitivity to initial values, entrapment in local optima, and the inability to distinguish objects with similar color intensity. This paper proposes the hybrid Fuzzy c-means clustering and Gray wolf optimization for image segmentation to overcome the shortcomings of Fuzzy c-means clustering. The Gray wolf optimization has a high exploration capability in finding the best solution to the problem, which prevents the entrapment of the algorithm in local optima. In this study, breast cytology images were used to validate the methods, and the results of the proposed method were compared to those of c-means clustering.ResultsFCMGWO has performed better than FCM in separating the nucleus from the other dark objects in the cell. The clustering was validated using Vpc, Vpe, Davies-Bouldin, and Calinski Harabasz criteria. The FCM and FCMGWO methods have a significant difference with respect to the Vpc and Vpe indexes. However, there is no significant difference between the performances of the two clustering methods with respect to the Calinski-Harabasz and Davies-Bouldin indices. The results indicate the better efficacy of the proposed method.ConclusionsThe hybrid FCMGWO algorithm distinguishes the cells better in images with less detail than in images with high detail. However, FCM exhibits unacceptable performance in both low- and high-detail images.

Fluorescent Staining of Arbuscular Mycorrhizal Structures Using Wheat Germ Agglutinin (WGA) and Propidium Iodide.

The colonization of a host plant root by arbuscular mycorrhizal (AM) fungi is a progressive process, characterized by asynchronous hyphal growth in intercellular and intracellular spaces, leading to the coexistence of diverse intraradical structures, such as hyphae, coils, arbuscules, and vesicles. In addition, the relative abundance of intercellular and intracellular fungal structures is highly dependent on root anatomy and the combination of plant and fungal species. Lastly, more than one fungal species may colonize the same root, adding a further level of complexity. For all these reasons, detailed imaging of a large number of samples is often necessary to fully assess the developmental processes and functionality of AM symbiosis. To this aim, the use of rapid and efficient staining methods that can be used routinely is crucial.We herein present a simple protocol to obtain high detail images of both overall intraradical fungal colonization pattern and fine morphology, in AM root sections of Lotus japonicus. The procedure is based on tissue clearing, fluorescent staining of fungal cell walls with fluorescein isothiocyanate-conjugated wheat germ agglutinin (FITC-WGA), and the combined counterstaining of plant cell walls with propidium iodide (PI). The resulting images can be acquired using traditional or confocal fluorescence microscopes and used for qualitative and quantitative analyses of fungal colonization, of particular interest for the comparison of mycorrhizal phenotypes between different experimental conditions or genetic backgrounds.

Design of a Mini Fluorescence Microscope

Microscopy is one of the most commonly used imaging methods in scientific research. The fluorescence microscopes which are used to determine the cell and tissue content at the molecular level, provides high-resolution images of the fluorochrome marked molecules. They are used in many areas with this feature. However, the use of fluorescence microscopy in biomedical field is limited due to various factors such as cost, size and difficulty of use. Also for some applications of fluorescence microscopy there are no need for very high quality and detailed images. Therefore, a small, practical and cost-effective mini fluorescent microscope is needed CITATION Mar17 \l 1055 (Sanz, Picazo-Bueno, Granero, & Mico, 2017) . In this study, a practical, portable and cost-effective fluorescent microscope was designed for specific proposes. Unlike the common fluorescent microscopes in this mini version one emission and one excitation wavelength is used. From top to bottom the design consists of a LED which is emitting the chosen wavelength of light, an excitation filter that can be used if necessary, a sample holder, a practical focus system developed in this project to be used for focus adjustment, a single-focus microscope objective lens, an emission filter and a camera. Plates holding these elements were formed using laser cutter, these plates are fixed in the form of layers with the appropriate materials. Filters and LED light source were determined and supplied according to the specific fluorescent dye to be used. These elements were tested by spectrophotometer and assembled together in the design. The LED power driver used in the design was supplied and the connections with the LED are completed. The focal point was calculated as 4 mm and the focal length as 44 mm so optical magnification 0.27 um/pixel was obtained. This means a 610 times magnification on a 15.6-inch screen. Using the mini fluorescent microscope a fluorochrome sensitive to 360nm wavelength was imaged. And the image from emitted light at 460nm wavelength was obtained. The entire design cost 2500 TL (434 USD). The dimensions of the fluorescence microscope are 120mm x 120mm x 170mm, 2448 c m 3 . Its weight is 1 kg. This mini fluorescence microscope can be used as an alternative to traditional fluorescent microscopes in developing countries and geographies where resources are scarce, because it is cost-effective and portable. As it is practical and user friendly, it can reduce the need for expert person. Also for point of care applications, this microscope can be an alternative for conventional fluorescence microscopy.

THE IMAGES OF WOMEN IN THE ART OF NORTHERN PONTIC REGION AS THE SOURCE FOR RECONSTRUCTION OF FEMALE SCYTHIAN COSTUMES

The costume is the category of culture which contains the information about material and spiritual spheres of life within the socium. The important sources for the costumes reconstruction are the archaeological artifacts. They combine findings — toreutics artifacts with antropomorhpys images as well as remains of headwearings, clothing, shoes, decorative elements of clothing. Antropomorphy was not inherent for the art of Scythian tribes. However in V century BC the images of «masks» without specific details occur. Later, in IV century BC we find toreutics products which focus on the «portraits». Among them we find classical ancient Greek images, and stylized, made in Greek-Scythian tradition. Such characteristics can be found in artifacts with plot compositions. They are distinguished by Greek iconography, as well as ethnographic characteristics of the characters. Generally they can be presented as an image of the goddess on the throne from different perspectives.
 The studying of the images showed that Scythian females had headwearings, generally inherent for many tribes. They are bedspreads, tapes, caps: cylindrical (poloses, modieses, kalafs) and cone-shaped. The ethnolocal specifics to such headwearings is given by decorative elements. Some of them have the specifics of Greek diadem — stephan. It is highlighted by archaeological findings — golden aprons for trimming poloses with arched frontal ledge.
 The analysis of female Scythian outer look allows to imagine the shape of shoulder and waist clothing: robes with sleeves longer than arm, wide dresses, draped skirt. Due to high-detailed images we can determine the way of shaping: observing details of the images showed two options for tunic-shaped cutting.
 Despite the schematics and the generalization in female imagery, studying toreutics works with female Scythian images presented is an important foundation point for reconstructing not only the general look, but also the ways of creating costumes, their aesthetic, ethnolocal and semantic specifics.

290 Nonlinear microscopy for the visualization of calcification and assessment of connective tissue fibers in pseudoxanthoma elasticum

Pseudoxanthoma elasticum (PXE, OMIM 264800) is a rare disorder caused by mutations of the ABCC6 gene in which decreased plasma levels of inorganic pyrophosphate (PPi) lead to ectopic mineralization resulting in severe dermatological, ophthalmologic and cardiovascular complications. Components of the dermal connective tissue can be differentiated with nonlinear optical (NLO) techniques, as collagen emits a strong second harmonic generation (SHG) signal given its non-centrosymmetric structure, while elastin as an endogenous chromophore can be visualized by two-photon excitation fluorescence (TPF). We carried out deep phenotyping, molecular genetic testing and plasma PPi measurement of five patients with PXE. Ex vivo SHG and TPF imaging and conventional histopathological evaluation of their skin was performed and compared to five age- and gender-matched healthy controls. Pathological mutations were found in the ABCC6 gene in all patients and their plasma PPi levels were significantly lower compared to controls by unpaired t test (p=0,0005). We visualized and quantified the degradation and mineralization of elastin fibers and extensive mid-dermal calcium salt deposition as well as the alterations of the collagen structure in the mid-dermis of patients with PXE utilizing TPF and SHG, respectively. Based on our data, NLO enables high-detail imaging and quantification of the specific features of PXE-affected skin. In vivo NLO may be a promising method in the assessment of PXE in the future, facilitating early diagnosis and monitoring.

Classification of bird species from video using appearance and motion features

The monitoring of bird populations can provide important information on the state of sensitive ecosystems; however, the manual collection of reliable population data is labour-intensive, time-consuming, and potentially error prone. Automated monitoring using computer vision is therefore an attractive proposition, which could facilitate the collection of detailed data on a much larger scale than is currently possible.A number of existing algorithms are able to classify bird species from individual high quality detailed images often using manual inputs (such as a priori parts labelling). However, deployment in the field necessitates fully automated in-flight classification, which remains an open challenge due to poor image quality, high and rapid variation in pose, and similar appearance of some species. We address this as a fine-grained classification problem, and have collected a video dataset of thirteen bird classes (ten species and another with three colour variants) for training and evaluation. We present our proposed algorithm, which selects effective features from a large pool of appearance and motion features. We compare our method to others which use appearance features only, including image classification using state-of-the-art Deep Convolutional Neural Networks (CNNs). Using our algorithm we achieved an 90% correct classification rate, and we also show that using effectively selected motion and appearance features together can produce results which outperform state-of-the-art single image classifiers. We also show that the most significant motion features improve correct classification rates by 7% compared to using appearance features alone.

Automated Link Generation for Sensor-Enriched Smartphone Images

The ubiquity of the smartphones makes them ideal platforms for generating in-situ content. In well-attended events, photos captured by attendees have diverse views that could be subjected to occlusion and abnormal lighting effects that could obscure the view. Such unstructured photo collections also have significant redundancy. Thus, a scene that is partially occluded or has bad contrast in one photo may be captured in another photo, possibly with higher details. We propose an application called Autolink that automatically establishes content-based links between sensor-annotated photos in unstructured photo collections captured using smartphones, such that users could navigate between high-context and high-detail images. This hierarchically structured image collection facilitates the design of applications for navigation and discovery, analytics about user photography patterns, user taste, and content/event popularity. Autolink includes a framework that constructs this hierarchy efficiently and with little content-specific training data by combining photo content processing with associated sensor logs obtained from multiple participants. We evaluated the performance of Autolink on two real-world sensor tagged photo datasets. The result shows that Autolink is able to efficiently cluster photos at 20 times faster than candidate algorithms, into the appropriate hierarchy with at least 70% precision and 37% better recall than candidate algorithms.

A New Dimension in Documenting New Species: High-Detail Imaging for Myriapod Taxonomy and First 3D Cybertype of a New Millipede Species (Diplopoda, Julida, Julidae).

We review the state-of-the-art approaches currently applied in myriapod taxonomy, and we describe, for the first time, a new species of millipede (Ommatoiulus avatar n. sp., family Julidae) using high-resolution X-ray microtomography (microCT) as a substantive adjunct to traditional morphological examination. We present 3D models of the holotype and paratype specimens and discuss the potential of this non-destructive technique in documenting new species of millipedes and other organisms. The microCT data have been uploaded to an open repository (Dryad) to serve as the first actual millipede cybertypes to be published.

A GA-BASED APPROACH FOR EPIPOLAR GEOMETRY ESTIMATION

In this paper, a genetic algorithm (GA)-based approach to estimate the fundamental matrix is presented. The aim of the proposed GA-based algorithm is to reduce the effect of noise and outliers in the corresponding points which affect the accuracy of the estimated fundamental matrix. Although in the proposed approach the GA is allowed to select the significant among all detected points, on the average half of the matched points have been determined to give optimum estimation of the fundamental matrix. Experiments with synthetic and real data show that the proposed approach is accurate especially in the presence of a high percentage of outliers. The proposed GA can always obtain good results in both high and low detailed images. Even for low detailed images which have a small number of matched points available to estimate the fundamental matrix, the proposed GA outperformed other methods.

Automated image mosaics by non‐automated light microscopes: theMicroMossoftware tool

Light widefield microscopes and digital imaging are the basis for most of the analyses performed in every biological laboratory. In particular, the microscope's user is typically interested in acquiring high-detailed images for analysing observed cells and tissues, meanwhile being representative of a wide area to have reliable statistics. The microscopist has to choose between higher magnification factor and extension of the observed area, due to the finite size of the camera's field of view. To overcome the need of arrangement, mosaicing techniques have been developed in the past decades for increasing the camera's field of view by stitching together more images. Nevertheless, these approaches typically work in batch mode and rely on motorized microscopes. Or alternatively, the methods are conceived just to provide visually pleasant mosaics not suitable for quantitative analyses. This work presents a tool for building mosaics of images acquired with nonautomated light microscopes. The method proposed is based on visual information only and the mosaics are built by incrementally stitching couples of images, making the approach available also for online applications. Seams in the stitching regions as well as tonal inhomogeneities are corrected by compensating the vignetting effect. In the experiments performed, we tested different registration approaches, confirming that the translation model is not always the best, despite the fact that the motion of the sample holder of the microscope is apparently translational and typically considered as such. The method's implementation is freely distributed as an open source tool called MicroMos. Its usability makes building mosaics of microscope images at subpixel accuracy easier. Furthermore, optional parameters for building mosaics according to different strategies make MicroMos an easy and reliable tool to compare different registration approaches, warping models and tonal corrections.

An Image Segmentation using Normalised Cuts in Multistage Approach

Normalised cuts algorithm is complex for image segmentation but it produces good segmentation result. At present, digital camera can produce high detail image. To avoid underutilising the high detail image, resizing image into smaller resolution is discouraged. This creates a constraint in resizing image to smaller resolution while preserving the important detail in the image. An image segmentation method using normalised cuts done in two-stage manner is proposed here to solve the issue of the image resolution is excessively reduced prior to image segmentation. In this work, an image is first separated into several regions(named as image cells). The locally produced segments from each of the image cells are then undergone for second stage segmentation to look for possibility of merging them up. This paper includes the experimental results using the mentioned approach and the experiment shows that it is capable to produce meaningful segments.

Fast encoding method for vector quantisation of images using subvector characteristics and Hadamard transform

In a framework of vector quantisation (VQ), the encoding speed is a key issue for its practical applications. To speed up the VQ encoding process, a fast encoding method in the Hadamard transform domain is presented. In the proposed method, a five-step test flow based on two characteristic values, the first element and variance of the transformed subvector, is introduced to reject a large number of unlikely codewords. In order to make full use of the energy–compaction property of an orthogonal transform, the partial distance search (PDS) method is used in its Steps 4 and 5. Experimental results show that the proposed algorithm outperforms most of the existing algorithms, especially in case of larger codebook size and high-detail images.

Narrow-linear and small-area forest disturbance detection and mapping from high spatial resolution imagery

Widespread disturbance has brought a large amount of narrow-linear and small-area disturbance features (e.g., trails, seismic lines, forest roads, well sites, and cut blocks) to forest areas throughout the past decade. This issue has prompted research into finding the appropriate data and methods for mapping these narrow-linear and small-area disturbance features in order to examine their impacts on wildlife habitat. In this paper, we first described the characteristics of small forest disturbances and presented the nature of problem. We then presented a framework for detecting and extracting narrow-linear and small-area forest disturbance features. Using a SPOT 5 high spatial detail image and existing GIS databases, we applied the framework to map narrow-linear and small-area forest disturbance features in a Bear Management area (BMA) in the eastern slopes of the Rocky Mountains in Alberta, Canada. The results indicated that the proposed framework produced accurate disturbance maps for cut blocks, and forest roads & trails. The high errors of omission in the cut lines map were attributed to inconsistent geometric and radiometric patterns in the 'rarely-used' or 'old' cut lines. The study confirmed the feasibility of rapidly updating incomplete GIS data with linear and small-area disturbance features extracted from high spatial detail SPOT imagery. Future work will be directed towards improvement of the framework and the extraction strategy to remove a large amount of spurious features and to increase accuracy for cut lines mapping.

Image denoising method based on bidimensional empirical mode decomposition and mean filtering

A new image-denoising method was proposed.Noised image was decomposed to a series of sub-band images by Bimensional Empirical Mode Decomposition(BEMD).High frequency sub-band images were denoised by mean filtering using different filter templates,and low frequency approximation image remained unchanged in this process.Then the denoised-image was obtained by composing the low frequency approximation images and the high frequency detailed images with mean filtering.Experimental result shows that the noise is effectively removed and the detail of the image is well preserved.This method has better denoising effect than single BEMD method and mean filtering method and the method combining wavelet transform with mean filtering.