NUMA-aware image compositing on multi-GPU platform

NUMA-Aware Image Compositing on Multi-GPU Platform

Sort-last parallel rendering is widely used. Recent GPU developments mean that a PC equipped with multiple GPUs is a viable alternative to a high-cost supercomputer: the Fermi architecture supports uniform virtual addressing, providing a foundation for non-uniform memory access (NUMA) on multi-processor platforms. Such hardware changes require the user to reconsider the design of parallel rende...

Structure-Aware Image Compositing

Object-aware Gradient-Domain Image Compositing

We describe an approach to suppress bleeding artifacts without altering the boundary location in gradient-domain compositing, a technique to create seamless composites. While gradient-domain compositing has become a standard tool for many complex image editing tasks such as seamless cloning, panorama stitching or scene completion, its quality suffers from mismatches in the composited image regi...

Multi-Projector Image Correction on the GPU

Multi-projector displays require a variety of image correction techniques. Geometric distortions resulting from projecting onto non-planar display surfaces and non-linear properties of projector lenses must be corrected to produce usable images. Photometric issues such as varying luminance response between projectors and brightness discontinuities in projector overlap regions must also be corre...

Energy-Aware Data Management on NUMA Architectures