The irradiance Jacobian for partially occluded polyhedral sources
SIGGRAPH '94 Proceedings of the 21st annual conference on Computer graphics and interactive techniques
Monte Carlo techniques for direct lighting calculations
ACM Transactions on Graphics (TOG)
Lightcuts: a scalable approach to illumination
ACM SIGGRAPH 2005 Papers
Accurate Direct Illumination Using Iterative Adaptive Sampling
IEEE Transactions on Visualization and Computer Graphics
ACM SIGGRAPH 2006 Papers
Matrix row-column sampling for the many-light problem
ACM SIGGRAPH 2007 papers
An efficient GPU-based approach for interactive global illumination
ACM SIGGRAPH 2009 papers
Real-time soft shadow mapping by backprojection
EGSR'06 Proceedings of the 17th Eurographics conference on Rendering Techniques
| Hi-index | 0.00 |
This paper presents a scalable, adaptive algorithm for direct illumination in scenes with many area light sources. Instead of presampling light sources into massive point lights like previous multilight algorithms, the algorithm builds a tree structure for all area lights using agglomerative hierarchical clustering algorithm and evaluates direct illumination by a collection of tree nodes. Each node represents a light cluster and its output radiance is computed using two estimation methods. For each pixel point, a node set is found adaptively through tree traversal guided by error estimation and overall radiance variation. As the algorithm works on area lights directly, the error incurred by converting area lights to point lights is avoided. The experimental results show that our algorithm can generate images with higher fidelity and is more efficient compared with existing point light based multi-light algorithms.