高级检索

基于射线投影的多视场深度像融合

Integration of Multiple Range Images Based on Ray Casting

  • 摘要: 首先确定匹配好深度像的最小轴对齐包围盒.通过盒面,沿着x,y,z 3个轴方向投射3个均匀分布的射线组,计算得到与所有深度像的交点,并通过Dexel结构来存储射线组与交点.加入距离和法向夹角2个判断依据来去掉重叠点,从而将多个视场深度像融合为完整的、无冗余的三维模型.实验结果表明,该方法误差小、速度快,而且简单有效.

     

    Abstract: An axis-aligned bounding box for all the registered range images is set up first.Then three sets of uniform-distributed rays are cast and passed through the faces of the box along three orthogonal axes of x,y,z,respectively. The intersections between the rays and the range images are then calculated and stored in Dexels data structure.The overlapped intersections are identified by one specific restriction based on their locations and normals.After removing the overlapped points,a complete non-redundant three-dimension model is obtained.The experimental results show that the proposed algorithm is accurate,and efficient.

     

/

返回文章
返回