Abstract:Illumination consistency is important for photorealistic rendering of mixed reality. However, it is usually difficult to acquire illumination conditions of natural environments. In this paper, we propose a novel method for evaluating the light conditions of a static outdoor scene without knowing its geometry, material, or texture. In our method, we separate respectively the shading effects of the scene due to sunlight and skylight through learning a set of sample images which are captured with the same sun position. A fixed illumination map of the scene under sunlight or skylight is then derived reflecting the scene geometry, surface material properties and shadowing effects. These maps, one for sunlight and the other for skylight, are therefore referred to as basis images of the scene related to the specified sun position. We show that the illumination of the same scene under different weather conditions can be approximated as a linear combination of the two basis images. We further extend this model to estimate the lighting condition of scene images under deviated sun positions, enabling virtual objects to be seamlessly integrated into images of the scene at any time. Our approach can be applied for online video process and deal with both cloudy and sun shine situations. Experiment results successfully verify the effectiveness of our approach.

Research on Out-Door Scene Image Generation Approach under Different Weathers

Generation of Multi-Spectral Scene Images under Different Weather Conditions

Basis Image Decomposition for Outdoor Lighting Estimation

Image degradation and recovery based on multiple scattering in remote sensing and bad weather condition

Scene Simulation for LLL Night Vision System Based on Daytime Image

A New Interactive Enhanced Method for Image with Poor Visibility

Computer Graphics in China: A practical approach for real-time illumination estimation of outdoor videos

Recovering of Images Degraded by Atmosphere

Lighting Simulation of Augmented Outdoor Scene Based on a Legacy Photograph.

Basis Image Decomposition of Outdoor Time-Lapse Videos.

A Global Infrared Image Synthesis Model for Large-Scale Complex Urban Scene

Real-time Generation of Infrared Dynamic Sky Scene

Image visibility enhancement based on atmospheric transmission theory and weight analysis under bad weather condition

Generation of Dynamic Infrared Outer Space Scene

Light source estimation of outdoor scenes for mixed reality

Image-based Relighting from a Sparse Set of Outdoor Images.

An Algorithm for Outdoor Illumination Estimation Based on Partial Least Squares Method

Real-Time walkthrough system of large-scale infrared dynamic scene

An Infrared Image Synthesis Model For Large-Scale Complex Urban Scene

Online Illumination Estimation of Outdoor Scenes Based on Videos Containing No Shadow Area.

Robust and Real-Time Outdoor Illumination Estimation