城市景观三维重建及渲染算法研究
|
摘要
针对现有的三维重建及渲染算法无法满足城市景观三维重建的需求问题,本文提出基于Kal-man滤波修正的三维重建算法和基于改进LOD的地形渲染算法。首先利用Kalman滤波修正算法对三维重建光束平差算法进行优化,加速原算法的响应时间,然后对LOD地形渲染算法进行优化,使用节点评价函数进行判断,节点是否需要进行分裂,从而实现不同的细节层次,利用添加边以及删减边结合的方法来消除地形中出现的裂缝。仿真试验结果表明,本文提出的基于Kalman滤波修正的三维重建算法和基于改进LOD的地形渲染算法在城市景观三维重建的应用中效果良好。
According to the existing 3D reconstruction and rendering algorithms cannot meet the requirements of 3D reconstruction of urban landscape, 3D reconstruction algorithm based on Kalman filtering correction and terrain rendering algorithm based on the improved LOD are proposed. First, use the modified Kalman filtering algorithm to optimize 3D reconstruction beam adjustment algorithm, speed up the response time of the original algorithm, and then optimize the LOD terrain rendering algorithm;use node evaluation function to determine whether the nod need to split or not, to realize the different level of detail; combine the add and subtract edge to eliminate cracks of terrain. Simulation experiment results show that, the proposed 3D reconstruction algorithm based on Kalman filtering correction and terrain rendering algorithm based on the improved LOD have good effect in the application of 3D reconstruction of urban landscape.
According to the existing 3D reconstruction and rendering algorithms cannot meet the requirements of 3D reconstruction of urban landscape, 3D reconstruction algorithm based on Kalman filtering correction and terrain rendering algorithm based on the improved LOD are proposed. First, use the modified Kalman filtering algorithm to optimize 3D reconstruction beam adjustment algorithm, speed up the response time of the original algorithm, and then optimize the LOD terrain rendering algorithm;use node evaluation function to determine whether the nod need to split or not, to realize the different level of detail; combine the add and subtract edge to eliminate cracks of terrain. Simulation experiment results show that, the proposed 3D reconstruction algorithm based on Kalman filtering correction and terrain rendering algorithm based on the improved LOD have good effect in the application of 3D reconstruction of urban landscape.
引文
[1]邓世军,王永杰.数据分页技术的海量三维数据模型动态调度[J].测绘科学,2013,38(4):97-100.
[2]高山,陈思.城市三维建模技术与标准研究[J].测绘通报,2013,3:95-97.
[3]龙忠杰,王吉芳,左云波.一种改进的Harris与SUSAN相结合的角点检测算法[J].计算机应用与软件,2013,30(12):133-136.
[4]赵呈,吴锡生.基于Harris算法的图像角度计算方法[J].计算机工程,2013,39(12):167-170.
[5]孔珊珊,降爱莲.基于第二代模糊集阈值选择的图像Canny边缘检测[J].计算机应用与软件,2011,30(12):160-162.
[6]孙金彦,周绍光,陈超.基于改进的Hough变换提取影像边缘[J].计算机与数字工程,2013,41(9):1501-1504.
[7]花开胜,王林.基于平面模板的摄像机标定方法[J].计算机工程,2012,38(15):264-267.
[8]舒远,谈正,丁礼儒.利用空间正交约束的相机自标定和三维重建[J].西安交通大学学报,2005,39(2):138-141.
[9]彭声泽,郭兴吉.3D建筑重建中纹理自动提取与映射方法研究[J].科技通报,2014,30(1):72-75.
[10]陈慧蓉,付胜豪.三维场景相息图快速生成研究[J].重庆工商大学学报:自然科学版,2013,30(5):37-40.
[2]高山,陈思.城市三维建模技术与标准研究[J].测绘通报,2013,3:95-97.
[3]龙忠杰,王吉芳,左云波.一种改进的Harris与SUSAN相结合的角点检测算法[J].计算机应用与软件,2013,30(12):133-136.
[4]赵呈,吴锡生.基于Harris算法的图像角度计算方法[J].计算机工程,2013,39(12):167-170.
[5]孔珊珊,降爱莲.基于第二代模糊集阈值选择的图像Canny边缘检测[J].计算机应用与软件,2011,30(12):160-162.
[6]孙金彦,周绍光,陈超.基于改进的Hough变换提取影像边缘[J].计算机与数字工程,2013,41(9):1501-1504.
[7]花开胜,王林.基于平面模板的摄像机标定方法[J].计算机工程,2012,38(15):264-267.
[8]舒远,谈正,丁礼儒.利用空间正交约束的相机自标定和三维重建[J].西安交通大学学报,2005,39(2):138-141.
[9]彭声泽,郭兴吉.3D建筑重建中纹理自动提取与映射方法研究[J].科技通报,2014,30(1):72-75.
[10]陈慧蓉,付胜豪.三维场景相息图快速生成研究[J].重庆工商大学学报:自然科学版,2013,30(5):37-40.