摘要
人工标志识别包括标志区域识别和标志图案识别,它是实现增强现实的重要步骤。通过对包含人工标志的图像进行二值化,采用连通域提取的算法实现了标志区域识别。对不同标志图案识别问题,提出了连通域数量判别法和模板匹配法。实验结果表明,所提方法能较好地识别人工标志,为实现虚实场景的实时融合打下了基础。
Artificial mark recognition, including mark area recognition and mark pattern recognition, is a significant process of augmented reality. A method to recognize mark area by connectivity area abstraction of the binary image which includes marks, is presented. According to different mark pattern recognition, a method based on the number of connectivity areas of the pattern and a method based on template match are presented. The result proves the methods presented can recognize the marks accurately, which gets prepared for the real time synthesis of the virtual and the real.
出处
《测控技术》
CSCD
2008年第10期7-9,12,共4页
Measurement & Control Technology
基金
江苏省社会发展计划基金资助(BS2007048)
南京市软件产业计划资助(2007软资119)
关键词
增强现实
标志识别
连通域提取
模板匹配
augmented reality
mark recognition
connectivity area abstraction
template match
