摘要
针对点云数据的Delaunay三角网格纹理映射速度慢、映射效果不够细腻及不适合大规模点云数据纹理映射等问题,提出一种基于球面纹理映射的点云数据重建改进方法,并在Qsplat算法的基础上进行实现。采用Qsplat算法对大规模点云数据进行模型重建,利用球面等比约束纹理映射算法建立纹理坐标、球面、点云重建模型三者之间的数学关系,实现大规模点云数据的球面纹理映射。实验结果表明,与传统的三角网格纹理映射相比,该方法可明显提高纹理映射的速度和质量,拓宽球面等比约束纹理映射方法的应用范围,适用于大规模点云数据的纹理映射。
Data texture mapping methods based on the Delaunay triangular mesh are usually computationally slow and the quality of mapping is low,making them unsuitable for large-scale point cloud data. In this paper,an improved spherical texture mapping method is developed for the reconstruction of point cloud data and it can be implemented based on the Qsplat algorithm. The Qsplat algorithm is used to re-establish the model of large-scale point cloud data. The spherical equal-ratio constraint texture mapping is used to obtain the mathematical relationship among texture coordinates,sphere and the reconstructed model,realizing the spherical texture mapping of large scale point data. Experimental results show that the speed and the quality of mapping of the proposed method are much improved compared with those traditional triangular texture mapping methods.
出处
《计算机工程》
CAS
CSCD
北大核心
2015年第1期218-222,共5页
Computer Engineering
基金
国家"863"计划基金资助项目(2013AA013804)
国家自然科学基金资助项目(61175072
51165033
61163023)
国家"973"计划基金资助项目(2011CB302400)
江西省科技支撑计划基金项目(20121BBE50023)
