Wide-view Image Research Articles

Multiple Structural and Phase Transformations of MOF and Selective Hydrocarbon Gas Separation in its Amorphous, Glass Phase States.

The first wide-view image of multiple structural and phase transformations for MOFs, ranging from crystal state transformations to the extreme limit approaching liquid/glass phase, was presented. The process involves i) an initial crystalline transformation from square-layer framework [Co2(pybz)2(CH3COO)2] ⋅ DMF (Co2) to a 3-fold interpenetrated and ordered vacancies contained framework [Co(pybz)2(CH3OH)2] ⋅ 2CH3OH (CoM) due to in situ disassemble-reassemble, ii) thermal induced departure of a pair of cis-form coordinated methanol in CoM leads to amorphous framework a-dCoM, iii) glass transition to super-cooled liquid scl-dCoM, iv) obtaining MOF glass g-dCoM upon quenching the super-cooled liquid, and v) re-crystallization of super-cooled liquid generates 6-fold interpenetrated dia-net framework [Co(pybz)2]6n (rec-dCoM) under further heating. The access to glass from CoM, provides a new self-perturbation strategy to create MOF glasses without melting. The wider pore size distribution in amorphous/glassy MOFs than crystalline precursor achieved the first time selective hydrocarbon gas separation by breakthrough experiments, which bring efficient separation of 1 : 99 C2H2/C2H4 by either a-dCoM or g-dCoM and produce polymer grade C2H4 with purity≥99.5 % after a single adsorption process. Furthermore, the mixture of 50 : 50 C3H6/C3H8 can be separated by a-dCoM.

High-precision optical modeling method for galvanometer-driven dual-camera systems.

To build a galvanometer-driven dual-camera sensing system, it is important to accurately correlate the wide-view image coordinates with the pan-tilt mirror angles for adjusting the incident light path of the zoom-in camera. Existing optical modeling methods assume sufficient target distance and simplify dual-camera optical centers as coincident. However, this simplification is not valid in many practical cases and might cause severe system malfunctions, such as complete loss tracking of important targets. To address this problem, we propose a novel approach, to the best of our knowledge, to facilitate high-precision optical modeling and calibration of galvanometer-driven dual-camera systems. The proposed method takes into consideration the dual-camera optical center misalignment issue and builds a model for accurate estimation and rectification of target localization errors under various optical configurations. Qualitative and quantitative experimental results demonstrate the superiority of our method, improving the performance of galvanometer-driven dual-camera systems for high-precision optical sensing applications.

Large Field-of-View Nonlinear Holography in Lithium Niobate.

A nonlinear holographic technique is capable of processing optical information in the newly generated optical frequencies, enabling fascinating functions in laser display, security storage, and image recognition. One popular nonlinear hologram is based on a periodically poled lithium niobate (LN) crystal. However, due to the limitations of traditional fabrication techniques, the pixel size of the LN hologram is typically several micrometers, resulting in a limited field-of-voew (FOV) of several degrees. Here, we experimentally demonstrate an ultra-high-resolution LN hologram by using the laser poling technique. The minimal pixel size reaches 200 nm, and the FOV is extended above 120° in our experiments. The image distortions at large view angles are effectively suppressed through the Fourier transform. The FOV is further improved by combining multiple diffraction orders of SH fields. The ultimate FOV under our configuration is decided by a Fresnel transmission. Our results pave the way for expanding the applications of nonlinear holography to wide-view imaging and display.

Continuous digital zooming using generative adversarial networks for dual camera system

This paper presents a generative adversarial network (GAN) with patch match algorithm to realize a high-quality digital zooming using two camera modules with different focal lengths. In dual camera system, shorter focal length module produces the wide-view image with the low resolution. On the other hand, the longer focal length module produces the tele-view image via optical zooming. The long-focal image contains more details than short-focal image and can be used to guide short-focal image to reconstruct high frequency part. Firstly, a feature extraction block (FEB) is advanced to extract feature of long-focal image and short focal-image to reconstruct a wide-view image with different resolutions. Next, a patch match algorithm is integrated into convolution neural networks (CNN) to fuse information of long-focal with short-focal image and generate a new fused image. Finally, the fused image and short-focal image are merged with a feature fusion block (FFB) to predict high-resolution images. In addition, generative adversarial networks are used for filtering information integrated by previous network and output the zoomed image. Extensive experiments on benchmark datasets show that our algorithm achieves favorable performance against state-of-the-art methods.

Road pavement crack detection using deep learning with synthetic data

The improvement of road system quality is a critical task. The mechanism to address such important issue is close monitoring of road pavement condition. Traditional approach requires manual identification of damages. Taking into account considerable length of road system it is essential to create an effective automatic pavement defects detection tool. This approach will extremely reduce time for monitoring of current road state. In this paper global experience in solution of detection issues of road pavement’s distress is reviewed. The article includes information about the existing datasets of road defects, which are commonly used for detection and segmentation. The present work is based on deep learning approach with the use of synthetic generated training data for segmentation of cracks in driver-view image. The novelty of the approach lies in creating synthetic dataset for training state-of-the-art deep learning frameworks. The relevance of the research is emphasized by processing of wide-view images in which heterogeneous pixel intensity, complex crack topology, different illumination condition and complexity of background make the task challenging.

Pavement Image Datasets: A New Benchmark Dataset to Classify and Densify Pavement Distresses

Automated pavement distresses detection using road images remains a challenging topic in the computer vision research community. Recent developments in deep learning have led to considerable research activity directed towards improving the efficacy of automated pavement distress identification and rating. Deep learning models require a large ground truth data set, which is often not readily available in the case of pavements. In this study, a labeled dataset approach is introduced as a first step towards a more robust, easy-to-deploy pavement condition assessment system. The technique is termed herein as the pavement image dataset (PID) method. The dataset consists of images captured from two camera views of an identical pavement segment, that is, a wide view and a top-down view. The wide-view images were used to classify the distresses and to train the deep learning frameworks, while the top-down-view images allowed calculation of distress density, which will be used in future studies aimed at automated pavement rating. For the wide view group dataset, 7,237 images were manually annotated and distresses classified into nine categories. Images were extracted using the Google application programming interface (API), selecting street-view images using a python-based code developed for this project. The new dataset was evaluated using two mainstream deep learning frameworks: You Only Look Once (YOLO v2) and Faster Region Convolution Neural Network (Faster R-CNN). Accuracy scores using the F1 index were found to be 0.84 for YOLOv2 and 0.65 for the Faster R-CNN model runs; both quite acceptable considering the convenience of utilizing Google Maps images.

Image stitching by creating a virtual depth

A method for image stitching is presented. The approach focuses on images with parallax (depth variation) to create panoramic views with high fidelity. The approach creates the stitching seam at a virtual depth to convert hard stitching problems to simple ones. The virtual depth is created by applying local distortions to the input images at the stitching seam so that the contents visually appear to be located at the same depth. The presented approach targets a wide variety of applications that require generating high (or super) resolution, wide-view images. These applications include tele-presence (or tele-reality) applications such as shopping, touring, conferencing, planning or architecting, learning, inspection, and surveillance. Our results show that the proposed approach provides promising results compared to commercial products that rely on stitching solutions.

A novel 2D and 3D wide-view imaging approach by controlling optimal range for cytopathological analysis in light microscopic system

Abstract Due to limitations of acquired pixel size and small field of view in the light microscopic system, pathologists must move microscope platform along the X–Y–Z axes without losing focusing in order to see whole regions of the specimen. Wide-view imaging provides a more realistic cytopathological analysis on the 2D and 3D panoramic images with an ultra-high resolution by minimizing these limitations. In this study, a novel 2D and 3D wide-view imaging approach is proposed for cytopathological analysis in light microscopic systems. This approach is performed in four main stages; estimation of optimal range, generation of optimal focused 2D & 3D adjacent images, control of optimal range and 2D & 3D image stitching. In contrast to literature studies for 2D and 3D wide-view imaging, optimal range and the number of multi-focus images are selected according to types of specimen and magnification objectives, and re-arranged to prevent the difference of focusing between adjacent images during X–Y movements of the platform. In order evaluate the performance of our proposed 2D wide-view imaging approach, two different 2D wide-view imaging approaches suggested in literature are applied to our datasets. The generated 2D wide-view images are compared using the metrics without requiring reference image, and the success of the proposed 2D and 3D wide-view imaging approach is proven by both quantitative and visual results.

High-performance integral-imaging-based light field augmented reality display using freeform optics.

A new integral-imaging-based light field augmented-reality display is proposed and implemented for the first time, to our best knowledge, to achieve a wide see-through view and high image quality over a large depth range. By using custom-designed freeform optics and incorporating a tunable lens and an aperture array, we demonstrated a compact design of a light field head-mounted-display that offers a true 3D display view of 30° by 18°, maintains a spatial resolution of 3 arc minutes across a depth range of over 3 diopters, and provides a see-through field of view of 65° by 40°.

Exploring the Impact of Seasonality on Urban Land-Cover Mapping Using Multi-Season Sentinel-1A and GF-1 WFV Images in a Subtropical Monsoon-Climate Region

The objective of this research was to investigate the impact of seasonality on urban land-cover mapping and to explore better classification accuracy by using multi-season Sentinel-1A and GF-1 wide field view (WFV) images, and the combinations of both types of images in subtropical monsoon-climate regions in Southeast China. We obtained multi-season Sentinel-1A and GF-1 WFV images, as well as the combinations of both data, by using a support vector machine (SVM) and a random forest (RF) classifier. The backscatter intensity, texture, and interference-coherence images were extracted from Sentinel-1A images, and different combinations of these Sentinel-1A-derived images were used to evaluate their ability to map urban land cover. The results showed that the performance of winter images was better than that of any other season, while the summer images performed the worst. Higher classification accuracy was achieved by using multi-season images, and satisfactory classification results were obtained when using Sentinel-1A images from only three seasons. The best classification result was achieved using a combination of all Sentinel-1A data from all four seasons and GF-1 WFV data from winter, with an overall accuracy of up to 96.02% and a kappa coefficient reaching 0.9502. The performance of textures was slightly better than that of the backscatter-intensity images. Although the coherence data performed the worst, it was still able to distinguish urban impervious surfaces well. In addition, the overall classification accuracy of RF was better than that of SVM.

Wide-viewing integral imaging system using polarizers and light barriers array

BackgroundIntegral imaging is considered one of the most promising three-dimensional display technologies, while the limited viewing angle is regarded as a primary disadvantage of integral imaging display to reach a commercial level. This paper proposes a viewing angle enhancement method for both liquid crystal display (LCD) and the projection-type three-dimensional integral imaging system.MethodsThe proposed wide-viewing integral imaging system is established by using the lenslet array coupling with the polarizers, light barriers, and enlarged elemental images array. The size of light barrier and polarizer is equal to the pitch of each lenslet, the light barrier prevents light rays from passing through, and the polarizer controls the passage of light ray in the specific polarization direction. In the projection-type integral imaging system, two orthogonal elemental images arrays (EIA) are projected onto the projection screen simultaneously by the corresponding projectors. In LCD integral imaging system, two orthogonal EIAs are displayed by use of an LCD screen which can switch the polarization direction of the EIA by time-multiplexed technology within the time constant of the eyes’ response time.ResultsThe viewing angle can be enlarged dramatically by the improvement of the size of each elemental image according to the integral imaging principles. The experimental result shows that the proposed method exhibits approximately four times the viewing angle of conventional integral imaging with the same lens array.ConclusionsThe increment of viewing angle can be determined by the number of light barriers between two adjacently orthogonal polarizers, the more the light barrier, the larger the viewing angle.

5-4 8眼全周撮像システムを用いた動物体追跡(第5部門 メディア処理)

5-4 8眼全周撮像システムを用いた動物体追跡(第5部門 メディア処理)

Continuous digital zooming using local self-similarity-based super-resolution for an asymmetric dual camera system.

This paper presents a digital zooming method using a super-resolution (SR) algorithm based on the local self-similarity between the wide- and tele-view images acquired by an asymmetric dual camera system. The proposed SR algorithm consists of four steps: (i)registration of an optically zoomed image to the wide-view image, (ii)restoration of the central region of the zoomed wide-view image, (iii)restoration of the boundary region of the zoomed wide-view image, and (iv)fusion of the results from steps (ii) and (iii). Since an asymmetric dual camera system acquires different-resolution images on the same scene due to the different optical specifications, the proposed method can restore the low-resolution wide-view image using the ideal high-frequency component estimated from the optically zoomed image. Experimental results demonstrate that the proposed method can provide significantly improved high-resolution wide-view images compared to existing single-image-based SR methods.

Weak-beam scanning transmission electron microscopy for quantitative dislocation density measurement in steels.

To evaluate dislocations induced by neutron irradiation, we developed a weak-beam scanning transmission electron microscopy (WB-STEM) system by installing a novel beam selector, an annular detector, a high-speed CCD camera and an imaging filter in the camera chamber of a spherical aberration-corrected transmission electron microscope. The capabilities of the WB-STEM with respect to wide-view imaging, real-time diffraction monitoring and multi-contrast imaging are demonstrated using typical reactor pressure vessel steel that had been used in an European nuclear reactor for 30 years as a surveillance test piece with a fluence of 1.09 × 1020 neutrons cm-2. The quantitatively measured size distribution (average loop size = 3.6 ± 2.1 nm), number density of the dislocation loops (3.6 × 1022 m-3) and dislocation density (7.8 × 1013 m m-3) were carefully compared with the values obtained via conventional weak-beam transmission electron microscopy studies. In addition, cluster analysis using atom probe tomography (APT) further demonstrated the potential of the WB-STEM for correlative electron tomography/APT experiments.

OPTIMAL IMAGE STITCHING FOR CONCRETE BRIDGE BOTTOM SURFACES AIDED BY 3D STRUCTURE LINES

Crack detection for bridge bottom surfaces via remote sensing techniques is undergoing a revolution in the last few years. For such applications, a large amount of images, acquired with high-resolution industrial cameras close to the bottom surfaces with some mobile platform, are required to be stitched into a wide-view single composite image. The conventional idea of stitching a panorama with the affine model or the homographic model always suffers a series of serious problems due to poor texture and out-of-focus blurring introduced by depth of field. In this paper, we present a novel method to seamlessly stitch these images aided by 3D structure lines of bridge bottom surfaces, which are extracted from 3D camera data. First, we propose to initially align each image in geometry based on its rough position and orientation acquired with both a laser range finder (LRF) and a high-precision incremental encoder, and these images are divided into several groups with the rough position and orientation data. Secondly, the 3D structure lines of bridge bottom surfaces are extracted from the 3D cloud points acquired with 3D cameras, which impose additional strong constraints on geometrical alignment of structure lines in adjacent images to perform a position and orientation optimization in each group to increase the local consistency. Thirdly, a homographic refinement between groups is applied to increase the global consistency. Finally, we apply a multi-band blending algorithm to generate a large-view single composite image as seamlessly as possible, which greatly eliminates both the luminance differences and the color deviations between images and further conceals image parallax. Experimental results on a set of representative images acquired from real bridge bottom surfaces illustrate the superiority of our proposed approaches.

Non-dyadic fisheye lens correction model for image enhancement.

This paper presents a non-dyadic framework to improve example-based enhancement of radially distorted images acquired by a very wide-angle lens. In order to remove both jagging and blurring artifacts in the correction process of the fisheye lens' barrel distortion, the proposed method first performs non-dyadic or multiple-step geometric correction based on the parabolic equation-based lens distortion model. At each correction step, an example-based image enhancement method removes undesired geometric correction artifacts such as jagging and blurring. Experimental results demonstrate that the proposed method outperforms existing fisheye lens image enhancement methods in the sense of both subjective and objective measures. Based on both theoretical advancement and experimental results, the proposed method can be used for various wide-view imaging applications including vehicle front- and rear-view cameras and wide-angle video surveillance systems.

Visible wide angle view imaging system of KTM tokamak based on multielement image fiber bundle.

In the paper, new visible wide angle view imaging system of KTM tokamak is described. The system has been designed to observe processes inside of plasma and the processes occurring due to plasma-wall interactions through the long equatorial port. Imaging system is designed based on special image fiber bundle and entrance wide angle lens, which provide image of large section of the vacuum chamber, both poloidal half-section and divertor through the sufficiently long equatorial port. The system also consists of two video cameras: slow and fast with image intensifier. Commercial equipment had been used in design of the system that allowed reducing the cost and time for research and development. The paper also discusses advantages and disadvantages of the system in comparison with conventional endoscopes based on a lens system and considers its promising utilization in future tokamaks and future steady state fusion reactors.

2P1-D01 Proposal of a Wide View Image Sensor Configuration for Object Detection on the Water

2P1-D01 Proposal of a Wide View Image Sensor Configuration for Object Detection on the Water

Feature Descriptor Based on Normalized Corners and Moment Invariant for Panoramic Scene Generation

Panorama generation systems aim at creating a wide-view image by aligning and stitching a sequence of images. The technology is extensively used in many fields such as virtual reality, medical image analysis, and geological engineering. This research is concerned with combining multiple images with a region of overlap to produce a wide field of view by the detection of feature points for images with different camera motion in an efficient and fast way. Feature extraction and description are important and critical steps in panorama construction. This study presents techniques of corner detection, moment invariant and random sampling to locate the important features and built storing descriptors in the images under noise, transformation, lighting, little viewpoint changes, blurring and compression circumstances. Corner detection and normalization are used to extract features in the image, while the descriptors are built by moment invariant in an efficient way. Finally, the matching and motion estimation is implemented based on the random sampling method. The results of experiments conducted on images and video sequences taken by handheld camera and images taken from the internet. The results show that the proposed algorithm generates panoramic image and panoramic video of good quality in a fast and efficient way.

Wide-Field Laser Ophthalmoscopy for Mice: A Novel Evaluation System for Retinal/Choroidal Angiogenesis in Mice

The purpose of this study was to investigate the application of wide-field laser ophthalmoscopy (Optos) for the evaluation of established models of angiogenesis and the healthy retina in mice. To investigate whether angiogenesis and leakage in the retina and choroid can be evaluated with Optos, we used two models of angiogenesis: oxygen-induced retinopathy (OIR) and laser-induced choroidal neovascularization (CNV). Fundus imaging and fluorescein angiography (FAG) were performed with the Optos system without a contact lens. Furthermore, to evaluate in vivo leukocyte infiltration in these models, we injected acridine orange (AO) and performed imaging using Optos. In vivo fundus imaging with Optos did not require any additional optical device. Additionally, Optos enabled us to repeatedly obtain high-resolution color images and FAG images in the OIR model as well as in the CNV model in mice. Through a combination of Optos imaging and AO fluorography, the number and location of the infiltrating leukocytes could be identified in these models. Optos is a wide-viewing imaging tool for the noninvasive in vivo evaluation of common angiogenesis models, oxygen-induced retinopathy and laser-induced choroidal neovascularization, as well as the healthy retina in mice.