Fisheye Lens Research Articles

Development of an eyeball integrated with a wide-angle lens camera and vision-sharing system for android avatars

Development of an eyeball integrated with a wide-angle lens camera and vision-sharing system for android avatars

Luminance Measurements using Smartphone Cameras

We evaluate the daylight performance visual comfort buildings by measuring light in luminance and illuminance using luminance and lux meter respectively. Detailed measurements of illuminance in a space using these methods on a grid can be time-consuming, while luminance meters to measure luminance values are quite expensive. Recent studies indicate that luminance-based metrics have the best fit to occupant responses in visual comfort surveys. But High Dynamic Range (HDR) techniques can quickly capture a large field of view at a high resolution and can be done with DSLR cameras that are more accessible than luminance meters. This research goes one step further to evaluate the HDR acquisition method using even more accessible smartphone cameras. The method includes calibration of the HDR images using lux meters which are affordable and easily available. A variety of smartphone cameras were evaluated in a wide range of brightness conditions in a scene. The error between the low-cost HDR approach and the conventional approach of HDR imaging is determined and a method for correction factors is proposed. Vignetting corrections for the smartphone camera lens have been demonstrated. A visual comfort survey of a primarily daylit workshop space is done to document as a test case to evaluate the feasibility and accuracy of the low-cost HDR approach. Lux meter for calibration of HDRI images is found to be reasonably accurate (average error of 7%). Of the 6 smartphones studied, the maximum luminance was captured by iPhone 6 and which could only capture 17% of the measured luminance. The fisheye lens used in this study increases the FOV from 68º (no fisheye) to 136° (with fisheye) but the FOV outside 100° was not useful due to large vignetting effect.

APPLICATION OF ACTION CAMERA VIDEO FOR FAST AND LOW-COST PHOTOGRAMMETRIC SURVEY OF CULTURAL HERITAGE

Abstract. The research work focuses on the growing of fast and low-cost data surveying methods (both in the acquisition and processing phases), to be used mainly as a support for the management and enhancement of architectural heritage. Attention is paid on the diffusion of action cameras, characterized by wide-angle lenses and 4K video shooting, with more and more affordable prices and on the elaboration of the data with photogrammetry software. The high-resolution videos captured by these tools, in fact, allow the selection of frames characterized by sufficient overlap, necessary for point cloud generation through Structure for Motion (SfM) processes. The focus is on indoor environment characterized by architectural value and on the data processing not only with professional photogrammetric application but also with cloud server-based system. Hence, the frames extracted from the video have been processed with different, both professional, economic, and user-friendly software tools, achieving several results. Finally, the output data have been compared with a survey method now well established in this field, such as laser scanning. The results of the analyses will be critically discussed in the paper to highlight the great potentialities and the possibility to apply this methodology in the future for heritage preservation.

A reflectionless compact elliptical half Maxwell fish-eye lens designed by transformation optics

A two-dimensional half Maxwell fish-eye lens is compressed using a linear transformation that maps a half circle to a half ellipse. The focusing property of the lens is preserved while making the device more compact. The boundary reflections, investigated for both TE and TM polarizations, were suppressed for beams directed toward the optical axis of the lens. A designed prototype provided a scanning range of ±20° with negligible reflections. The design’s functionality was verified using COMSOL multiphysics.

Synthesis of a Calibration-Free Visual Feedback Controller for an Inverted Pendulum Using a Fisheye Lens

This article considers the synthesis of feedback controllers for a class in which strong robust stability is required against sensing errors. The motivation dealing with this class arises from the stabilization problem of a cart-driven inverted pendulum by visual feedback control using a fisheye lens. While fisheye lenses guarantee a wide-angle of view, a large image distortion at off-center area is inevitable. Therefore this study first derives a geometric measurement model of a pendulum angle using a general fisheye lens model. A new design method for robust controllers for our problem is then proposed. Finally, a robust feedback controller that allows lens distortion is designed. The effectiveness of the controller based on the proposed model and the proposed design method is verified through numerical simulations and experiments.

Precision Calibration of Omnidirectional Camera Using a Statistical Approach

Omnidirectional optoelectronic systems (OOES) find applications in many areas where a wide viewing angle is crucial. The disadvantage of these systems is the large distortion of the images, which makes it difficult to make wide use of them. The purpose of this study is the development an algorithm for the precision calibration of an omnidirectional camera using a statistical approach. The calibration approach comprises three basic stages. The first stage is the formation of a cloud of points characterizing the view field of the virtual perspective camera. In the second stage, a calibration procedure that provides the projection function for the camera calibration is performed. The projection functions of traditional perspective lenses and omnidirectional wide-angle fisheye lenses with a viewing angle of no less than 180° are compared. The construction of the corrected image is performed in the third stage. The developed algorithm makes it possible to obtain an image for part of the field of view of an OOES by correcting the distortion from the original omnidirectional image.Using the developed algorithm, a non-mechanical pivoting camera based on an omnidirectional camera is implemented. The achieved mean squared error of the reproducing points from the original omnidirectional image onto the image with corrected distortion is less than the size of a very few pixels.

Abnormalities of thyroid function tests in patients receiving Immune Checkpoint Inhibitor Treatment for Cancer; importance of a wide-angled lens

Searchable abstracts of presentations at key conferences in endocrinology ISSN 1470-3947 (print) | ISSN 1479-6848 (online)

Multi-Thread AI Cameras Using High-Speed Active Vision System

In this study, we propose a multi-thread artificial intelligence (AI) camera system that can simultaneously recognize remote objects in desired multiple areas of interest (AOIs), which are distributed in a wide field of view (FOV) by using single image sensor. The proposed multi-thread AI camera consists of an ultrafast active vision system and a convolutional neural network (CNN)-based ultrafast object recognition system. The ultrafast active vision system can function as multiple virtual cameras with high spatial resolution by synchronizing exposure of a high-speed camera and movement of an ultrafast two-axis mirror device at hundreds of hertz, and the CNN-based ultrafast object recognition system simultaneously recognizes the acquired high-frame-rate images in real time. The desired AOIs for monitoring can be automatically determined after rapidly scanning pre-placed visual anchors in the wide FOV at hundreds of fps with object recognition. The effectiveness of the proposed multi-thread AI camera system was demonstrated by conducting several wide area monitoring experiments on quick response (QR) codes and persons in nature spacious scene such as meeting room, which was formerly too wide for a single still camera with wide angle lens to simultaneously acquire clear images.

The Duality of Ray-Based and Pinhole-Camera Modeling and 3D Measurement Improvements Using the Ray-Based Model.

Geometrical camera modeling is the precondition for 3D-reconstruction tasks using photogrammetric sensor systems. The purpose of this study is to describe an approach for possible accuracy improvements by using the ray-based-camera model. The relations between the common pinhole and the generally valid ray-based-camera model are shown. A new approach to the implementation and calibration of the ray-based-camera model is introduced. Using a simple laboratory setup consisting of two cameras and a projector, experimental measurements were performed. The experiments and results showed the possibility of easily transforming the common pinhole model into a ray-based model and of performing calibration using the ray-based model. These initial results show the model’s potential for considerable accuracy improvements, especially for sensor systems using wide-angle lenses or with deep 3D measurements. This study presents several approaches for further improvements to and the practical usage of high-precision optical 3D measurements.

Pixel-Wise Fisheye Image Correction Method With Single-View Phase Target

Fisheye lens is widely employed in many practices due to its wide field-of-view (FOV) and portability. There are however serious image distortions followed up with the fisheye lens. To address the problem, this letter proposed a pixel-wise fisheye image correction method based on single-view phase target. Compared with traditional correction methods which rely on a specific fisheye model to calibrate the parameters of lens distortion, the proposed method can establish the pixel-to-pixel mapping between lens distortion and ideal undistortion without fisheye lens modeling, which is more flexible and efficient to use in practice. Given the correspondence above, the distorted image can be further mapped into the corrected image by means of interpolation. Experiments on different scenes demonstrate the feasibility and accuracy of the proposed method to correct fisheye images for lens distortion.

딥러닝 기반의 2차원 이미지 왜곡 분류 및 이미지 각도 보정 시스템 개발

When taking a picture with a camera, the distortion that is different from reality occurs due to wide-angle lenses and camera angles. In this paper, we propose an image distortion classification and calibration program that provides users with standard images before distortion by classification and calibrating distortion. The program automatically predicts camera parameters from a single input image and proceeds with calibration. Inputting the image, distortion image classification using deep learning (CNN) determines whether Wide-angle lens distortion and Camera-angle distortion exist. When it is determined that distortion exists, deep learning and OpenCV are used to calibrate the distortion state according to each image characteristic. As a result of the program operation, it was confirmed that the output image was calibrated similarly to the actual one, and more fine distortion calibration can be expected by finding distortions that were difficult to judge only with human eyes.

Characterization of Cubic Zirconia as a Lens Material Suitable for Autonomous Driving

As the development of autonomous driving technology is now in full swing, the demand for miniaturized optical modules mounted on various sensors has increased. Particularly, the optical lens used for such autonomous driving must demonstrate stable performance and durability despite rapid changes in the external environment. In this regard, cubic zirconia (CZ) can be used as an optical lens due to its high refractive index, which is above 2.1 in visible and near-infrared wavelengths, along with its chemical and mechanical durability. Thus, in this paper, we investigated the temperature-dependent physical properties of CZ fabricated by the skull melting method. The temperature coefficient of the refractive index (dn/dT) of the fabricated CZ plate in the temperature range of 25–100 °C decreased from 9.76 × 10−5/K to 7.00 × 10−5/K as the wavelength increased from 447.0 nm to 785.0 nm. The estimated Abbe number decreased from 33.98 at 25 °C to 33.12 at 100 °C, while the measured coefficient of thermal expansion (CTE) was 9.91 × 10−6/K, which revealed that the dn/dT value of the CZ plate was more affected by the dispersion than by the CTE. In addition, the CZ samples with a high refractive index, coated with a dielectric multilayer showed a high average transmittance of 98.2% at the investigated wavelengths, making it suitable for miniaturization or wide-angle optical lens modules. To secure the durability required for automobile lenses, the variation in the surface profile of the CZ before and after the external impact was also analyzed, revealing much better performance than TAF glass. Therefore, the observed results demonstrate that the CZ material is suitable for use as an optical lens for autonomous vehicles.

Sponge City Planning and Information System Development Based on Geographic Information Fuzzy Processing.

During the development of the urban water system from 1.0 to 3.0, the impervious surface area gradually increased, hindering the natural infiltration and self-purification process of urban rainwater, resulting in serious urban water pollution, urban waterlogging in the rainy season, and groundwater problems. Since water will seep into the ground, serious water pollution will cause damage to the ground. Therefore, in dealing with urban rainwater problems, we want to use the sponge's ability to absorb and store water to build our cities into sponge cities. In this paper, we have constructed a sponge city planning and information system development based on geographic information fuzzy processing. We use differentiated fuzzy processing methods to eliminate classified information to achieve a perfect combination with nearby images and use ordinary fuzzy processing methods to solve the problem of nonconfidential information. This paper discusses the impact of natural topography on the planning and construction of sponge cities, including whether natural topography will affect rainwater, whether it will affect the distribution of different strata, and whether it will affect the utilization of groundwater resources. The basic functions of this platform are provided by a series of functions of GIS, and multiple modules are developed according to management requirements. The initial state of street view data is a lot of fisheye lens photos and corresponding point location information, which are displayed online after data preprocessing, detection information, editing information, and blurring processing.

Three-Dimensional Foot Position Estimation Based on Footprint Shadow Image Processing and Deep Learning for Smart Trampoline Fitness System

In the wake of COVID-19, the digital fitness market combining health equipment and ICT technologies is experiencing unexpected high growth. A smart trampoline fitness system is a new representative home exercise equipment for muscle strengthening and rehabilitation exercises. Recognizing the motions of the user and evaluating user activity is critical for implementing its self-guided exercising system. This study aimed to estimate the three-dimensional positions of the user’s foot using deep learning-based image processing algorithms for footprint shadow images acquired from the system. The proposed system comprises a jumping fitness trampoline; an upward-looking camera with a wide-angle and fish-eye lens; and an embedded board to process deep learning algorithms. Compared with our previous approach, which suffered from a geometric calibration process, a camera calibration method for highly distorted images, and algorithmic sensitivity to environmental changes such as illumination conditions, the proposed deep learning algorithm utilizes end-to-end learning without calibration. The network is configured with a modified Fast-RCNN based on ResNet-50, where the region proposal network is modified to process location regression different from box regression. To verify the effectiveness and accuracy of the proposed algorithm, a series of experiments are performed using a prototype system with a robotic manipulator to handle a foot mockup. The three root mean square errors corresponding to X, Y, and Z directions were revealed to be 8.32, 15.14, and 4.05 mm, respectively. Thus, the system can be utilized for motion recognition and performance evaluation of jumping exercises.

Broadband and wide-angle plane focal surface Luneburg lens

The energy crisis has aroused widespread concern, and the reform of energy structure is imminent. In the future, the energy structure will be dominated by the solar energy and other renewable energy sources. The solar concentrating technology as a promising method has been widely studied for collecting solar energy. However, the previous solar concentrating technologies suffer from some drawbacks, such as low focusing efficiency and large concentrating size. The Luneburg lens with highly efficient aberration-free focusing provides a new route for solar/energy concentrator. In this work, we designed a plane focal surface Luneburg lens (PFSLL) by transformation optics (TO). The PFSLL provides a relatively high focusing efficiency and concentration ratio of collection of energy. At the same time, it circumvents the disadvantage of curve surface of the classical Luneburg lens in device integration. Based on the reciprocity of electromagnetic waves, the PFSLL can also be applied to the antenna field to achieve broadband wide-angle scanning and highly directional radiation.

Super-resolution imaging of Maxwell’s fish-eye lens based on surface polaritons

Super-resolution imaging plays a crucial role in the fields of nanolithography, high volume transmission and sensing. Relentless efforts have been made to realize super-resolution imaging in the past decades. In this work, inspired by the mechanism of surface plasmon polaritons (SPPs), we find that Maxwell’s fish-eye lens (MFEL) coated with cylindrical layer of negative permeability can achieve super-resolution imaging. The amplification of evanescent waves in the negative permeability layer facilitates the transmittance of high spatial frequency information from object point to imaging point in the form of magnetic surface polaritons (MSPs). Both analytical calculations and numerical simulations are employed to prove the super-resolution imaging performance of MSPs-assisted MFEL. Our results may pave a new way for super-resolution imaging in metallic systems.

A technique for constructing depth map from images of orthogonally-divergent fisheye stereo

The issues of using a pair of orthogonally-divergent cameras with fisheye lenses in computer vision systems are considered. An algorithm has been developed for processing frames received from cameras to obtain a pair of rectified images. For a virtual model, the accuracy of constructing a scene depth map in the area of intersection of the cameras' fields of view was estimated.

Barrel Distortion Correction for Digital Optic-Based Motion Tracking in Ship Hydrodynamic Model Experiments

In hydrodynamic ship-model experiments, motion tracking of ship model trajectory is employed to predict the maneuverability of the ship when operating in its actual environment. There are some systems for motion tracking, such as inertial measurement unit, radar, laser, and infrared sensors. The system has proven reliable, however, the disadvantage is that they are all relatively expensive. This paper proposes a low-cost digital optic-based motion tracking system using consumer-grade video camera. In order to examine the accuracy of the system, some experiments to apply the method were carried out in maneuvering offshore engineering basin at the Indonesian Hydrodynamics Laboratory, The National Agency for Research and Innovation. To cover a broader area of the trajectory, a wide-angle camera lens was utilized, consequently, the image recorded by the camera experienced a barrel distortion. As a reference for the measurements, gridlines spaced 2 meters apart were drawn on the floor of the basin. Using the grid pattern, the distortion was corrected by finding the relationship between the grid pattern in the object space and in the distorted image. A fourth-order polynomial regression was applied to solve the problem. The results showed that the method was effective for correcting distorted images with a relative standard error of 0.42%.

Correction of radial distortion for fish eye lens camera model using code V

This work proposes a design of compensation for radial distortion of the barrel and pincushion distortion on picture geometry, both of which damage the image by utilizing the fisheye design model in the code v tool as the basis for the compensation. In order to prevent the picture's form from degrading, this model was constructed using image angle height mapping. As a result, there are two stages to this paper. The first step is to create a fish eye design model that has a field of vision (FOV) of 180 degrees by selecting the most appropriate optical design specification and design parameters. The second step is an examination of the design model utilizing several Code V choices, such as 2D image simulation (ISM), modulation transfer function, spot diagram, and distortion grid. Radial distortion compensation suggests the projection mapping model as a solution to this problem; this model also incorporates the stereographic approach. The compensation outcome is displayed by the 2D picture simulation using the most advanced quantitative measurements and state-of-the-art methods.

Design of a Miniaturized Wide-Angle Fisheye Lens Based on Deep Learning and Optimization Techniques.

This paper presents the optimization design of a miniaturized five-element wide-angle fisheye lens using a deep learning algorithm. Zemax optical design software was used to simulate and optimize the wide-angle fisheye lens. A deep learning algorithm helped to find the best combination of different lens materials. We first used six lens elements as an initial configuration to design miniaturized wide-angle fisheye lenses using the optimization process. The optical system components were gradually decreased to five lens elements. Both OKP4HT and polymethyl methacrylate (PMMA) plastic aspheric lenses were selected to replace the second spherical glass lens in the original design. We propose two types of wide-angle fisheye lens designs with four spherical lenses and one aspheric lens. The results for these designs indicated a viewing angle of 174°, a total length of less than 15 mm, a spot size of less than 6 μm, lateral color within ±1 μm, field curvature within ±0.02 mm, and F-θ distortion of ±3.5%. In addition, the MTF value was larger than 0.4 at the spatial frequency of 100 cycles/mm.