80th Anniversary Prize
Sun Light Estimator
This app estimates the intensity of sun light from the 360-degree image of RICOH THETA.
It is often the case that images with proper HDR (High Dynamic Range) cannot be obtained even with varying shutter speed outdoors on a sunny day because the sun light is simply too strong. This poses a difficulty when one wants to use 360-degree image from RICOH THETA for the purpose of IBL (Image-Based Lighting). The intensity of direct sun light is said to be 100, 000 times as strong as the light from the other part of the sky. This application creates an HDR image using many images taken with different shutter speeds. Also, it estimates the intensity of the sun light, which was too strong to be captured correctly in the original images, by using the shadow of RICOH THETA cast in the images.
Once started, the application reads the consecutive images, and creates an HDR image, scans the saturated pixels, approximates the regions of the sky using multiple sources of directional light, and estimates the intensity of the sun light and approximates it with a directional light source. The data of the light sources of the estimated and approximated sun light and the sky is written in a format that can be read by freely available ray-tracing software, POV-Ray.
By using the output from this application, one can create a CG image that shows the similar brightness and shadow effect as the real objects captured in the photos.
We noticed that there are many submitted works that handle IBL, and it may reflect the current trends. This work proposes a method to implement realistic IBL of an outdoor scene under direct sun light, and was evaluated highly in terms of its usefulness.
Interview with the award winners
I’ve been conducting the research on CG that uses IBL (Image Based Lighting) by using RICOH THETA. After I presented our research results at an academic conference, I was recommended to participate in this contest by RICOH staff, so I did. I’m very pleased to receive recognition for my research. I’d like to thank my students, faculty and colleagues who have been involved in the research activity.