Blending Luminance Uniformity Matrix: A New Robust Photometric Calibration in Multi-Projector Systems
Zhecui Chen,Xubo Yang,Shuangjiu Xiao,Pengyan Hou
DOI: https://doi.org/10.1109/icise.2009.363
2009-01-01
Abstract:Large-scale displays as multi-projector systems are used in scientific visualization, virtual reality and other visually intensive applications. It always needs complicated photometric calibration in multi-projector systems. The most popular methods of photometric calibration can be generalized into two types: edge blending and LAM (luminance attenuation maps). Unfortunately, both these two methods have their drawbacks. In this paper, we provide a new method which combines the blending with luminance attenuation by slightly reducing the area of luminance attenuation in overlapped region, which efficiently solves the problem of "luminance inconsistency in the edge of overlapped region" in LAM and the problem of "un-global luminance uniformity" in blending. In addition, instead of finding the minimum common luminance value in the photometric calibration, we just find a more suitable common luminance value which can well solve the dim problem in camera-based photometric uniformity of prior approaches. Moreover, the process of photometric calibration on multi-projecting system will be no more complicated than before. The method is robust and accurate, and can be implemented with commercial off-the-shelf components.