Decoding positional and color information from a coded pattern

Chang, Nelson L. ; Lim, Suk Hwan ; Tang, Feng

HP Development Company

关键词: single shot coded pattern;    camera calibration;    geometric calibration;    color consistency;    augmented reality;   

RP-ID  :  HPL-2010-131

学科分类：计算机科学（综合）

美国|英语

来源: HP Labs

【 摘 要 】

The paper presents decoding positional and color information using visual coded patterns for efficient geometric calibration and color consistency across multiple cameras. The patterns are generated from an alphabet of basis colors placed in unique spatial configurations called ChromaCodes. The imaged patterns are automatically decoded to derive the information needed for both geometric and color calibration. Previous 2-D structured patterns are typically designed for obtaining only geometry whereas the proposed patterns capture color information as well. Moreover, the unique decodability of the codes allows them to overcome the problem of pattern visibility in all views and enables effective geometric calibration and color consistency with even partial and/or occluded view of the pattern. Experimental results demonstrate that a single shot of the pattern is sufficient to encapsulate enough information to compute both geometry and color across one or more cameras, especially important for realtime or interactive applications.