简单视觉标记增强现实系统建模方法研究

Augmented Reality System Modeling Based on Simple Visual Marker

研究一种基于简单视觉标记的增强现实融合配准策略,并以此为基础构建了增强现实系统模型.结合色彩和形状的特征融合检测方法,采用双重特征结构确保了标记检测的可靠性.基于弱透视模型,利用冗余特征实现标记物体的3D位姿还原.为使配准方案能够适应普通的视觉采集设备,提出根据摄像机标定时得到的扭曲参数,对渲染后的模型进行再扭曲的配准修正方法.最后,结合Open CV和Open GL中的空间变换的概念,讨论了虚拟和物理世界配准融合的解决方案,并结合视觉标记检测结果,提出了图像配准的实现方法.

An augmented reality image registration strategy is studied based on simple visual marker, which is furthered to build aug- mented reality system in this paper. Color and shape feature are integrated to ensure the reliability of marker detection. Redundant fea- ture are utilized to calculate pose of visual markers based on weak perspective model. To make image registration be suitable to com- mon vision capture device, calibration distortion parameters are used to redistorted rendered model. Finally, based on space transfor- marion in OpenCV and OpenGL, virtual and real registration is discussed and implemented with visual marker detection result.