Yuan, GuodongQin, KaihuaiBernd Froehlich and Roland Blach and Robert van Liere2014-01-272014-01-272007978-3-905674-02-61727-530Xhttps://doi.org/10.2312/EGVE/IPT_EGVE2007/077-083In this paper, we firstly propose a partial-sampling scheme to measure the intensity transfer function of a projector. Secondly, we implement the computation of luminance surface by rendering a texture rectangle on GPU. Thirdly, we generate an unified index data for all projectors. And then we compute the masks of photometric correction using a GPU based on topological consistency. Finally we integrate the masks with a GPU-based photometric correction pipeline to achieve the photometric uniformity for multi-projector display. The GPU-based framework is a combination of the GPU-based computation and the GPU-based photometric correction pipeline, which removes the bottleneck of the computation of luminance surface, attenuation mask and black offset mask. It is shown by experimental results prove that the GPU-based framework is effective and efficient.Categories and Subject Descriptors (according to ACM CCS): I.3.3 [Computer Graphics]: Multi-Projector Tiled Display