Relative sky radiance from multi-exposure all-sky camera images

All-sky cameras are frequently used to detect cloud cover; however, this work explores the use of these instruments for the more complex purpose of extracting relative sky radiances. An all-sky camera (SONA202-NF model) with three colour filters narrower than usual for this kind of cameras is configured to capture raw images at seven exposure times. A detailed camera characterization of the black level, readout noise, hot pixels and linear response is carried out. A methodology is proposed to obtain a linear high dynamic range (HDR) image and its uncertainty, which represents the relative sky radiance (in arbitrary units) maps at three effective wavelengths. The relative sky radiances are extracted from these maps and normalized by dividing every radiance of one channel by the sum of all radiances at this channel. Then, the normalized radiances are compared with the sky radiance measured at different sky points by a sun and sky photometer belonging to the Aerosol Robotic Network (AERONET). The camera radiances correlate with photometer ones except for scattering angles below 10∘, which is probably due to some light reflections on the fisheye lens and camera dome. Camera and photometer wavelengths are not coincident; hence, camera radiances are also compared with sky radiances simulated by a radiative transfer model at the same camera effective wavelengths. This comparison reveals an uncertainty on the normalized camera radiances of about 3.3 %, 4.3 % and 5.3 % for 467, 536 and 605 nm, respectively, if specific quality criteria are applied.

Relative sky radiance from multi-exposure all-sky camera images

All-sky cameras are frequently used to detect cloud cover; however, this work explores the use of these instruments for the more complex purpose of extracting relative sky radiances. An all-sky camera (SONA202-NF model) with three colour filters, narrower than usual for this kind of cameras, is configured to capture raw images at seven exposure times. A detailed camera characterization of the black level, readout noise, hot pixels and linear response is carried out. A methodology is proposed to obtain a linear high dynamic range (HDR) image and its uncertainty, which represents the relative sky radiance map at three effective wavelengths. The relative sky radiance (normalized by the sum of all radiances) is extracted from these maps and compared with the sky radiance measured at different sky points by a sun/sky photometer belonging to the Aerosol Robotic Network (AERONET). The camera radiances are in line with photometer ones excepting for scattering angles below 10º, which is probably due to some light reflections on the fisheye lens and camera dome. Camera and photometer wavelengths are not coincident, hence camera radiances are also compared with sky radiances simulated by a radiative transfer model at the same camera effective wavelengths. This comparison reveals an uncertainty on the normalized camera radiances about 3.3 %, 4.3 % and 5.3 % for 467, 536 and 605 nm, respectively, if specific quality criteria are applied.

US racial inequality may be as deadly as COVID-19

The COVID-19 pandemic is causing a catastrophic increase in US mortality. How does the scale of this pandemic compare to another US catastrophe: racial inequality? Using demographic models, I estimate how many excess White deaths would raise US White mortality to the best-ever (lowest) US Black level under alternative, plausible assumptions about the age patterning of excess mortality in 2020. I find that 400,000 excess White deaths would be needed to equal the best mortality ever recorded among Blacks. For White mortality in 2020 to reach levels that Blacks experience outside of pandemics, current COVID-19 mortality levels would need to increase by a factor of nearly 6. Moreover, White life expectancy in 2020 will remain higher than Black life expectancy has ever been unless nearly 700,000 excess White deaths occur. Even amid COVID-19, US White mortality is likely to be less than what US Blacks have experienced every year. I argue that, if Black disadvantage operates every year on the scale of Whites’ experience of COVID-19, then so too should the tools we deploy to fight it. Our imagination should not be limited by how accustomed the United States is to profound racial inequality.

Relative Impact of Key Rendering Parameters on Perceived Quality of VR Imagery Captured by the Facebook Surround 360 Camera

High quality, 360 capture for Cinematic VR is a relatively new and rapidly evolving technology. The field demands very high quality, distortion- free 360 capture which is not possible with cameras that depend on fish- eye lenses for capturing a 360 field of view. The Facebook Surround 360 Camera, one of the few “players” in this space, is an open-source license design that Facebook has released for anyone that chooses to build it from off-the-shelf components and generate 8K stereo output using open-source licensed rendering software. However, the components are expensive and the system itself is extremely demanding in terms of computer hardware and software. Because of this, there have been very few implementations of this design and virtually no real deployment in the field. We have implemented the system, based on Facebook’s design, and have been testing and deploying it in various situations; even generating short video clips. We have discovered in our recent experience that high quality, 360 capture comes with its own set of new challenges. As an example, even the most fundamental tools of photography like “exposure” become difficult because one is always faced with ultra-high dynamic range scenes (one camera is pointing directly at the sun and the others may be pointing to a dark shadow). The conventional imaging pipeline is further complicated by the fact that the stitching software has different effects on various as- pects of the calibration or pipeline optimization. Most of our focus to date has been on optimizing the imaging pipeline and improving the qual- ity of the output for viewing in an Oculus Rift headset. We designed a controlled experiment to study 5 key parameters in the rendering pipeline– black level, neutral balance, color correction matrix (CCM), geometric calibration and vignetting. By varying all of these parameters in a combinatorial manner, we were able to assess the relative impact of these parameters on the perceived image quality of the output. Our results thus far indicate that the output image quality is greatly influenced by the black level of the individual cameras (the Facebook cam- era comprised of 17 cameras whose output need to be stitched to obtain a 360 view). Neutral balance is least sensitive. We are most confused about the results we obtain from accurately calculating and applying the CCM for each individual camera. We obtained improved results by using the average of the matrices for all cameras. Future work includes evaluating the effects of geometric calibration and vignetting on quality.

Effects of Black Luminance Level on Image Quality

The image quality is affected by the black luminance level of the image. This research aimed to investigate how low luminance levels are required to maintain image quality. The psychophysical experiment was carried out in a dark room using OLED display. Total of 6 different black luminance levels (0.003, 0.05, 0.1, 0.2, 0.3, and 1 cd/m2) were used in the experiment. Total of 20 participants was invited to evaluate the image quality. For the experiment, twelve test images are used and these test images categorized into three groups as dark, medium bright and bright image group by image histogram distribution. Each image is rendered by adjusting six different black luminance levels. Result found that the black level is higher than 0.1 cd/m2, the preference for the image is decreased. The best performance is achieved when the black level is 0.003 cd/m2, but there is no big difference from 0.1 cd/m2. The final result shows that a change in black level between about 0.003 cd/m2 and 0.1 cd/m2 does not significantly affect image quality.

Creating high contrast in virtual night driving

The reproduction of glare scenarios in driving simulators is restricted by the limitations of conventional projector and display technology. In existing solutions, light sources are usually added to the optical path by combining common simulator technology with, for example, grids based on light-emitting diodes (LED). In this article, we introduce a new way to simulate glare sources on a common driving simulator back-projection screen using an additional projector and an additional reflective screen. In a first attempt, a concept with retroreflective sheets is proposed, and the requirements and current limitations of this setup are shown. Further development of the concept leads to an attempt with holographic diffusers. With an experimental setup based on the second concept, we aim to assess its feasibility, evaluate current challenges, and outline future requirements. The results show that it is possible to simulate a static glare scenario with correct geometrical conditions and headlights with luminances of about 100 000 cd/m2. Current limitations are an unbalanced color efficiency, the size and cost of available diffusers, and the black-level values and in-picture contrast of the used projector that cause an outshining of the environmental simulation.