Image Position Research Articles

Improved model of the Supernova Refsdal cluster MACS J1149.5+2223 thanks to VLT/MUSE

We present new \, observations of the Hubble Frontier Field (HFF) galaxy cluster MACS J1149.5+2223, lensing the well-known supernova ``Refsdal'' into multiple images, which has enabled the first cosmological applications with a strongly lensed supernova. Thanks to these data, targeting a northern region of the cluster and thus complementing our previous \, program on the cluster core, we have released a new catalog containing \, secure spectroscopic redshifts. We confirmed 22 cluster members, which had previously been only photometrically selected, and detected ten additional ones, resulting in a total of 308 secure members, of which 63<!PCT!> are spectroscopically confirmed. We further identified 17 new spectroscopic multiple images belonging to six different background sources. By exploiting these new and our previously published \, data, in combination with the deep HFF images, we developed an improved total mass model of MACS J1149.5+2223. This model includes 308 total mass components for the member galaxies and requires four additional mass profiles, one of which is associated with a cluster galaxy overdensity identified in the north, representing the dark matter mass distribution on larger scales. The values of the resulting 34 free parameters are optimized based on the observed positions of 106 multiple images from 34 different families, that cover an extended redshift range between 1.240 and 5.983. Our final model has a multiple image position root mean square value of which is in good agreement with other cluster lens models based on a similar number of multiple images. With this refined mass model, we have paved the way toward an improved strong-lensing analyses that will exploit the deep and high resolution observations with \, and \, on a pixel level in the region of the supernova Refsdal host. This will increase the number of observables by around two orders of magnitude, thus offering the opportunity to carry out more precise and accurate cosmographic measurements in the future.

Multi-detector fusion and Bayesian smoothing for tracking viral and chromatin structures

Automatic tracking of viral and intracellular structures displayed as spots with varying sizes in fluorescence microscopy images is an important task to quantify cellular processes. We propose a novel probabilistic tracking approach for multiple particle tracking based on multi-detector and multi-scale data fusion as well as Bayesian smoothing. The approach integrates results from multiple detectors using a novel intensity-based covariance intersection method which takes into account information about the image intensities, positions, and uncertainties. The method ensures a consistent estimate of multiple fused particle detections and does not require an optimization step. Our probabilistic tracking approach performs data fusion of detections from classical and deep learning methods as well as exploits single-scale and multi-scale detections. In addition, we use Bayesian smoothing to fuse information of predictions from both past and future time points. We evaluated our approach using image data of the Particle Tracking Challenge and achieved state-of-the-art results or outperformed previous methods. Our method was also assessed on challenging live cell fluorescence microscopy image data of viral and cellular proteins expressed in hepatitis C virus-infected cells and chromatin structures in non-infected cells, acquired at different spatial–temporal resolutions. We found that the proposed approach outperforms existing methods.

FLsM: Fuzzy Localization of Image Scenes Based on Large Models

This article primarily focuses on the study of image-based localization technology. While traditional methods have made significant advancements in technology and applications, the emerging field of visual image-based localization technology demonstrates tremendous potential for research. Deep learning has exhibited a strong performance in image processing, particularly in developing visual navigation and localization techniques using large-scale visual models. This paper introduces a sophisticated scene image localization technique based on large models in a vast spatial sample environment. The study involved training convolutional neural networks using millions of geographically labeled images, extracting image position information using large model algorithms, and collecting sample data under various conditions in elastic scene space. Through visual computation, the shooting position of photos was inferred to obtain the approximate position information of users. This method utilizes geographic location information to classify images and combines it with landmarks, natural features, and architectural styles to determine their locations. The experimental results show variations in positioning accuracy among different models, with the most optimal model obtained through training on a large-scale dataset. They also indicate that the positioning error in urban street-based images is relatively small, whereas the positioning effect in outdoor and local scenes, especially in large-scale spatial environments, is limited. This suggests that the location information of users can be effectively determined through the utilization of geographic data, to classify images and incorporate landmarks, natural features, and architectural styles. The study’s experimentation indicates the variation in positioning accuracy among different models, highlighting the significance of training on a large-scale dataset for optimal results. Furthermore, it highlights the contrasting impact on urban street-based images versus outdoor and local scenes in large-scale spatial environments.

The powerful lens galaxy cluster PLCK G287.0+32.9 (θE ∼ 43″)

Aims.We present a new high-precision strong-lensing model of PLCK G287.0+32.9, a massive lens galaxy cluster atz = 0.383, with the aim of obtaining an accurate estimation of its effective Einstein radius and total mass distribution. We also present a spectroscopic catalog containing accurate redshift measurements for close to 500 objects up to redshiftz = 6, including multiply lensed sources and cluster member galaxies.Methods.We exploited high-quality spectroscopic data from the Multi Unit Spectroscopic Explorer (MUSE), covering a central 3 arcmin2region of the cluster. We supplemented the spectroscopic catalog by including redshift measurements from VIsible MultiObject Spectrograph (VIMOS) and DEep Imaging Multi-Object Spectrograph (DEIMOS). We identified 129 spectroscopic cluster member galaxies with redshift values of 0.360 ≤ z ≤ 0.405, andmF160W ≤ 21. We complemented this galaxy cluster member sample with 24 photometric members identified with a convolutional neural network (CNN) approach. We also identified 114 multiple images from 28 background sources, of which 84 images from 16 sources are new and the remaining ones have already been identified in previous works. From these, we extracted “golden sample” of 47 secure multiple images and used them, together with the selected cluster member, to build and optimize several strong-lensing models with the software lenstool.Results.The best-fitting lens model shows a root mean square (RMS) separation value between the predicted and observed positions of the multiple images of 0.″75. Using its predictive power, we found three new multiple images and we confirm the configuration of three systems of multiple images that were not used for the optimization of the model. For a source at a redshift ofzs = 2, we found a cluster with an Einstein radius ofθE = 43.4″ ± 0.1″. This value is in agreement with previous estimates and corresponds to a total mass enclosed in the critical curve ofME = 3.33−0.07+0.02 × 1014 M⊙.Conclusions.The combined application of ancillaryHubbleSpace Telescope (HST) imaging, VIMOS and DEIMOS data, and the new MUSE spectroscopic observations allowed us to build a new lens model of the galaxy cluster PLCK G287.0+32.9, with an improvement in terms of reconstructing the observed positions of the multiple images of a factor of 2.5 with respect to previous models. The derived total mass distribution confirms this cluster to be a very prominent gravitational lens, with an effective Einstein radius ofθE ∼ 43″. We were also able to construct an extensive spectroscopic catalog containing 490 objects, of which 153 are bright cluster members withmF160W ≤ 21, and 114 are multiple images.

General multipoles and their implications for dark matter inference

ABSTRACT The flux ratios of strongly lensed quasars have previously been used to infer the properties of dark matter. In these analyses, it is crucial to separate the effect of the main lensing galaxy and the low-mass dark matter halo population. In this work, we investigate flux-ratio perturbations resulting from general third- and fourth-order multipole perturbations to the main lensing galaxy’s mass profile. We simulate four lens systems, each with a different lensing configuration, without multipoles. The simulated flux ratios are perturbed by 10–40 per cent by a population of low-mass haloes consistent with cold dark matter and, in one case, also a satellite galaxy. This level of perturbation is comparable to the magnitude of flux-ratio anomalies in real data that has been previously analysed. We then attempt to fit the simulated systems using multipoles instead of low-mass haloes. We find that multipoles with amplitudes of 0.01 or less can produce flux-ratio perturbations in excess of 40 per cent. In all cases, third- or fourth-order multipoles can individually reduce the magnitude of, if not eliminate, flux-ratio anomalies. When both multipole orders are jointly included, all simulated flux ratios can be fit to within the observational uncertainty. Our results indicate that low-mass haloes and multipoles are highly degenerate when modelling quadruply imaged quasars based just on image positions and flux ratios. In the presence of this degeneracy, flux-ratio anomalies in lensed quasars alone cannot be used to place strong constraints on the properties of dark matter without additional information that can inform our priors.

Georeferencing of Satellite Images with Geocoded Image Features

Abstract. Currently, using digital orthophoto map (DOM) and digital elevation model (DEM) as reference to achieve geometric positioning of newly acquired satellite images has become a popular photogrammetric approach. However, this method relies on DOM and DEM data which requires a lot of storage space in practical applications. In addition, for geometric positioning of satellite images, only sparse image feature points are needed as control points. Consequently, for the sake of convenience, the compression of control data emerges as a necessity with significant practical implications. This paper investigates a "cloud control" photogrammetry method based on geocoded image features. The method extracts SIFT feature points from DOMs, and obtains their ground coordinates, then constructs geocoded image feature library instead of DOM and DEM data as control, thus realizing the compression of control data. Experiments conducted on the Tianhui-1, Ziyuan-3 and Gaofen-2 satellite images demonstrate that the proposed method can achieve high-precision geometric positioning of satellite images and greatly reduce the size of the control data. Specifically, with the reduction of the reference data from 180~1248 MB 2 m DOM and 30 m DEM to 5~10 MB geocoded image features, the geopositional accuracies of the test Tianhui-1, Ziyuan-3 and Gaofen-2 images are improved from 3.12 pixels to 1.74 pixels, 3.69 pixels to 1.09 pixels, and 150.93 pixels to 2.67 pixels, respectively.

A method for detecting photovoltaic panel faults using a drone equipped with a multispectral camera

Abstract. Photovoltaic power stations utilizing solar energy, have grown in scale, resulting in an increase in operational maintenance requirements. Efficient inspection of components within these stations is crucial. However, the large area of photovoltaic power generation, coupled with a substantial number of photovoltaic panels and complex geographical environments, renders manual inspection methods highly inefficient and inadequate for modern photovoltaic power stations. To address this issue, this paper proposes a method and system for hot spot detection on photovoltaic panels using unmanned aerial vehicles (UAVs) equipped with multispectral cameras. The UAVs capture visible and infrared images of the photovoltaic power plant, which are then processed for photogrammetry to determine imaging position and attitude. The infrared images are stitched together using this information, forming a geographically referenced overall image. Hot spot detection is performed on the infrared images, enabling the identification of faulty photovoltaic panels and facilitating efficient inspection and maintenance. Experimental trials were conducted at a photovoltaic power station in Qingyuan, Guangdong Province China. The results demonstrate the effectiveness of the proposed method in accurately detecting panels with hot spot faults.

Bayesian inference of atomic column positions in scanning transmission electron microscopy images

Abstract Atomic-resolution scanning transmission electron microscopy combined with 2D Gaussian fitting enables the accurate and precise identification of atomic column positions within a few picometers. The measurement performance significantly depends on the signal-to-noise ratio of the atomic columns. In areas with low signal-to-noise ratios, such as near surfaces, the measurement performance was lower than that of the bulk. However, previous studies evaluated the accuracy and precision only in bulk areas, underscoring the need for a method that quantitatively evaluates the accuracy and precision of each atomic column position with various signal-to-noise ratios. This study introduced Bayesian inference to assess the accuracy and precision of determining individual atomic column positions under various signals. We applied this method to simulated and experimental images and demonstrated its effectiveness in identifying statistically significant displacements, particularly near surfaces with signal degradation. The use of vector maps and kernel density estimate plots obtained from Bayesian inference provided a probabilistic understanding of the atom displacement. Therefore, this study highlighted the potential benefits of Bayesian inference in high-resolution imaging to reveal material properties.

Abstract 3033: Registration And Segmentation Framework For 3d Instrument Visualization Relative To Preoperative Ct Imaging Via Spatially-tracked Ultrasound: A Phantom Study

Objectives: Surgical navigation guides surgery without radiation by mapping instrument locations onto preoperative CT (pCT). Adoption to vascular surgery remains novel. Feasibility of spatially-tracked ultrasound (US) for surgical registration and wire localization relative to pCT is demonstrated. Methods: pCT was taken (0.6mm) of a carotid artery phantom and a CT artery model (CTM) segmented. An US probe (Philips L17-5), was rigidly attached to an optical tracker and swept along the artery (1000 images). A navigation system (Medtronic StealthStation) acquired the tracker and 2D US image's positions via a spatial calibration (1.6mm accuracy). The artery in each image was segmented and tracking data combined to reconstruct a 3D US artery model (USM) (Fig. 1). The CTM and USM were registered via an iterative closest point algorithm. A Rosen wire (0.035in) was inserted into the artery. Acquiring an US sweep of the wire (500 images), it was segmented in each image via thresholding. Segmentations and tracking were combined into a 3D US wire-model (USWM). Its position was overlaid onto pCT via the USM-CTM registration. Another CT was acquired, registered via a fiducial point-based registration (0.86 mm accuracy), and the ground truth wire position segmented (Fig. 2). A tip-to-tip distance was measured between the USWM and CT wire positions. Results: CTM-to-USM registration error was 0.93 mm (RMSE) between artery point clouds. The tip-to-tip distance between CT and USWM wire models was 0.2 (longitudinal) and 1.4 (radial) mm. Conclusions: A vascular navigation framework capable of patient registration and highly accurate instrument localization is provided using tracked iUS.

Global pixel-value-ordering framework with dynamic sequence partition for reversible data hiding

Recently, pixel-value-ordering (PVO) based reversible data hiding (RDH) methods have become hotspots in spatial domain RDH research. In PVO-based methods, the pixel value correlation within a local region is exploited for data embedding. To embed secret data, the cover images are partitioned into non-overlapped rectangle blocks to be sorted locally, and a lower embedding distortion is thus obtained. However, their embedding capacities are often limited by the fixed local blocks from which pixels are sorted. To maintain the advantages and overcome disadvantages of the PVO-based methods, a new global pixel-value-ordering (GPVO) framework is proposed in this paper. By applying our GPVO framework, a PVO-based method can utilize pixels from any position in the cover image rather than from local pixel blocks. Then, through the dynamic sequence partition realized by the GPVO framework for the first time, the embedding capacity of PVO-based methods can be greatly improved while reducing embedding distortion further reduced. Finally, we propose a two-stage pairwise embedding scheme to be applied to the sequences, aiming to achieve an advanced embedding performance. Experimental results illustrate that the proposed GPVO works better than other related state-of-the-art RDH methods.

Experimental Determination of the Transverse Width and Position of the Virtual Images Produced by Thin Lenses and a Plane Mirror with Smartphone Photography

Experimental Determination of the Transverse Width and Position of the Virtual Images Produced by Thin Lenses and a Plane Mirror with Smartphone Photography

Gravitational lensing by a stable rotating regular black hole

Recent observational data from the Event Horizon Telescope (EHT) collaboration provide convincing realistic evidence for the existence of black hole rotation. From a phenomenological perspective, a recently proposed stable rotating regular (SRR) black hole circumvents the theoretical flaws of the Kerr solution. For the purpose of obtaining observational signatures of this black hole, we study its gravitational lensing effect. In the strong deflection limit, we calculate the deflection angle of light on the equatorial plane, the radius of the photon sphere, and other observables. The observables include the relativistic image position, separation, magnification, and time delays between different images. Then, by modeling M87* and Sgr A* as the SRR black hole, we compute their observables and evaluate the deviation of the observables from the Kerr case. In the weak deflection limit, we calculate the light deflection angle on the equatorial plane of M87* and Sgr A* via the Gauss-Bonnet theorem (GBT). With the growth of deviation parameter e, the gravitational lensing effect in the weak deflection limit intensifies monotonically, and the gravitational lensing effect in the strong deflection limit changes dramatically only at high spins. Our research may contribute to distinguish between SRR black holes from Kerr black holes under higher-precision astronomical observations.

Light refraction in the Earth’s atmosphere III. Inferior mirages: images locus

Our main goal is to find the locus of images formed as a result of an inferior mirage. To achieve our goal we show, at first, that, provided the beam entering a detection system has a small aperture, the image of a point object formed by that system in a vertical plane passing through the object is a point, regardless of whether the image formed by an optical system previously traversed by the beam has generated a non-point image of the object. Secondly, we show that the different images formed by the detection system, as its position relative to the object varies, are located on the caustic curve corresponding to the previously traversed optical system. Next, we have found the analytical expressions of the caustic curves corresponding to two particular cases, one of them being the inferior mirages. These expressions have been found by means of the Legendre antitransform of the asymptotic lines to the paths of light rays reaching the detector. For the case of inferior mirages, we have studied in detail the locus of the images in each vertical plane passing through the object as a function of the position of the object relative to the ground, its position relative to the detector, and the atmospheric conditions. Finally, we get somewhat into the matter of the images position as they would be seen by a casual observer of an inferior mirage.

Refractive index measurement using virtual—apparent image coincidence approach

Abstract In this study, we explore novel approaches to determining the index of refraction for various mediums by leveraging virtual images. We introduce a simplified yet effective method that utilizes apparent object positions resulting from light bending in a medium. Our focus lies in measuring the refractive index of liquids without the need for a microscope, emphasizing the measurement of real and apparent depths through the coincidence of virtual and apparent images. Comparative analysis with established techniques reveals the precision of our results, with accurate determinations of object and image positions. Through the collection and graphing of paired data, we calculate the refractive indices of water, vegetable oil, and ethanol. Our findings underscore the significance of this scientific methodology, offering time-efficient, implementable, and easily comprehensible procedures. We posit that this study holds promise for educational applications at various levels. Moreover, we propose extending our methodology to transparent solid materials, thereby broadening its potential applications.

Application of deep learning method for time reduction and precision improvement in displacement measurement during in-situ SEM tensile test

In this study, a novel material deformation measurement method is suggested to evaluate the mechanical properties of thin films. It is important to identify the mechanical properties of thin films in order to do a reliable design of these products. One of the methods to get the mechanical properties of thin films is to acquire images during an in-situ SEM (Scanning Electron Microscope) tensile test and analyze them. However, SEM image acquisition is mostly affected by the time varying motion of pixel positions in the consecutive images, a phenomenon called drift and distortion. In this study, some images of low resolution during in-situ SEM tensile test of the thin film are acquired in short time and restored to high resolution images using the deep learning model called Super Resolution Residual Network (SRResNet) to calculate the deformation of thin films. The proposed method increased the precision of measurement by 5 times, improved the accuracy of measurement by 20%, and reduced the measurement time by 19 times.

3D medical image encryption algorithm using biometric key and cubic S-box

Considering the scarcity of research on 3D medical image encryption, this paper proposes a novel 3D medical image encryption scheme based on biometric key and cubic S-box. To enhance the data security, biometric keys are utilized to overcome the limitations of traditional methods where secret keys with no practical meaning, fixed length, and finite key space, while cubic S-box is constructed to increase the nonlinearity of image cryptosystem. The proposed cryptosystem mainly consists of four phases: pseudo-random sequence generation, confusion, substitution, and diffusion. Firstly, the stepwise iterative algorithm based on coupled chaotic systems is utilized for generating pseudo-random sequences for confusion and diffusion. Secondly, the confusion algorithm based on multiple sorting can scramble pixel positions in 3D images. Thirdly, guided by the designed cubic S-box, pixel substitution is executed sequentially. Lastly, the diffusion algorithm based on ECA and finite field multiplication is capable of increasing the plaintext sensitivity of cryptosystem by concealing the statistical characteristics of plaintext. Simulation experiments performed on multiple 3D medical images demonstrate that the proposed encryption scheme exhibits favorable statistical performance, sufficiently large key space, strong system sensitivity and robustness, and can resist various typical cryptographic attacks.

Detecting oracle bone inscriptions via pseudo-category labels

Oracle bone inscriptions (OBIs) are the earliest Chinese characters and reserve abundant historical information. OBIs are detected by locating their positions in digital images. This has been a foundational task in modern archeological studies. Due to the development of deep neural networks(DNNs) in computer vision, detecting OBIs can be implemented by a more concise method instead of designing complicated hand-crafted features. However, existing models cannot perform well when noise areas are similar to some inscriptions. In this work, we present a simple but effective pseudo-label-based architecture for OBIs detection. Different from previous approaches, our method performs OBIs detection with the employment of information from multilabel annotations rather than single location information. We append a plug-and-play module that predicts the pseudo-label of an inscription after the backbone network for learning the particular structure prior to each inscription and brings this information to the backbone network by means of feature fusion. We make remarkable improvements on different backbone networks when using the proposed method on an OBIs detection dataset. The quantitative and qualitative results show that the proposed model can detect OBIs well and is an effective tool for assisting in the discovery and recognition of ancient writing.

ENHANCING MEDICAL DATA SECURITY IN E-HEALTH SYSTEMS USING BIOMETRIC-BASED WATERMARKING

In the field of Electronic Health (e-Health), Electronic Health Records (EHR) are transmitted between health professionals using e-Health systems for cooperative medical practice, medical monitoring, telemedical expertise, and telemedical imaging. Medical images are a crucial component of EHR and are used in various aspects of telemedicine systems such as expertise, consultation, teaching, and research. However, protecting the authenticity and copyrights of medical images is essential to prevent duplication, modification, or unauthorized distribution. This paper proposes a robust medical image copyright protection method that uses patient palm-print template as watermark and Lorenz chaotic map for template concealing and selecting the appropriate embedding positions in medical images. The novelty of the method lies in optimizing the expected number of modifications per pixel of the medical images after being watermarked. Experimental results indicate that this approach has a high performance with a genuine accept rate of 99.86% and can withstand various image processing attacks, including Gaussian noise, compression, and image rotations, while ensuring personal data security during telemedicine data exchange.

Manhole Cover Classification Based on Super-Resolution Reconstruction of Unmanned Aerial Vehicle Aerial Imagery

Urban underground pipeline networks are a key component of urban infrastructure, and a large number of older urban areas lack information about their underground pipelines. In addition, survey methods for underground pipelines are often time-consuming and labor-intensive. While the manhole cover serves as the hub connecting the underground pipe network with the ground, the generation of underground pipe network can be realized by obtaining the location and category information of the manhole cover. Therefore, this paper proposed a manhole cover detection method based on UAV aerial photography to obtain ground images, using image super-resolution reconstruction and image positioning and classification. Firstly, the urban image was obtained by UAV aerial photography, and then the YOLOv8 object detection technology was used to accurately locate the manhole cover. Next, the SRGAN network was used to perform super-resolution processing on the manhole cover text to improve the clarity of the recognition image. Finally, the clear manhole cover text image was input into the VGG16_BN network to realize the manhole cover classification. The experimental results showed that the manhole cover classification accuracy of this paper’s method reached 97.62%, which verified its effectiveness in manhole cover detection. The method significantly reduces the time and labor cost and provides a new method for manhole cover information acquisition.

HairNet2: deep learning to quantify cotton leaf hairiness, a complex genetic and environmental trait

BackgroundCotton accounts for 80% of the global natural fibre production. Its leaf hairiness affects insect resistance, fibre yield, and economic value. However, this phenotype is still qualitatively assessed by visually attributing a Genotype Hairiness Score (GHS) to a leaf/plant, or by using the HairNet deep-learning model which also outputs a GHS. Here, we introduce HairNet2, a quantitative deep-learning model which detects leaf hairs (trichomes) from images and outputs a segmentation mask and a Leaf Trichome Score (LTS).ResultsTrichomes of 1250 images were annotated (AnnCoT) and a combination of six Feature Extractor modules and five Segmentation modules were tested alongside a range of loss functions and data augmentation techniques. HairNet2 was further validated on the dataset used to build HairNet (CotLeaf-1), a similar dataset collected in two subsequent seasons (CotLeaf-2), and a dataset collected on two genetically diverse populations (CotLeaf-X). The main findings of this study are that (1) leaf number, environment and image position did not significantly affect results, (2) although GHS and LTS mostly correlated for individual GHS classes, results at the genotype level revealed a strong LTS heterogeneity within a given GHS class, (3) LTS correlated strongly with expert scoring of individual images.ConclusionsHairNet2 is the first quantitative and scalable deep-learning model able to measure leaf hairiness. Results obtained with HairNet2 concur with the qualitative values used by breeders at both extremes of the scale (GHS 1-2, and 5-5+), but interestingly suggest a reordering of genotypes with intermediate values (GHS 3-4+). Finely ranking mild phenotypes is a difficult task for humans. In addition to providing assistance with this task, HairNet2 opens the door to selecting plants with specific leaf hairiness characteristics which may be associated with other beneficial traits to deliver better varieties.