视觉感知驱动的三维城市场景数据组织与调度方法
|
摘要
针对大规模复杂三维城市场景三维可视化的数据调度效率不高、视觉一致性差等关键问题,提出了一种视觉感知驱动的复杂三维城市场景数据自适应组织管理与动态调度方法.该方法根据三维场景数据的空间分布特征,利用自适应四叉树对复杂三维城市场景进行不同层次粒度的划分,并自底向上遍历四叉树,为中间节点生成LOD(level of detail)和计算各个层级的几何误差,构建灵活的多粒度三维瓦片模型,最后根据屏幕误差评估建筑物模型的三维几何特征(形状、尺寸、高度)等视觉感知参数,约束不同细节层次模型的自适应调度.选择了柏林4个细节层次的三维城市模型数据,进行交互式三维漫游测试.实验结果表明,基于视点相关HLOD(hierarchical LOD)动态调度,动态可视化的渲染帧率始终保持在40 f/s左右,达到了网络环境下复杂三维城市场景数据动态调度的高效性和三维可视化视觉一致性的要求.
Focusing on the problems of inefficient 3D model scheduling and low visual consistency of complex 3D urban scene visualization,a self-adaptive organization and scheduling method for 3D urban scenes driven by visual perception is proposed. Based on the spatial distribution of 3D city models,the adaptive quadtree partition was used to divide the complex 3D scene into different levels of granularity.Subsequently,traversing the quadtree from bottom to top,levels of detail( LODs) were generated for the intermediate nodes,and the geometric error of each layer was calculated to construct a flexible multi-granularity 3D tile model. Finally,visual perception parameters such as the shape,size,andheight of the building model were evaluated by employing the screen error,which was used to constrain the self-adaptive scheduling of LODs. The 3D models of four LODs in Berlin were selected for an interactive 3D roaming test. The experimental results show that the rendering frame rate was more than40 f/s based on the view-related hierarchical LOD dynamic scheduling,which could ensure efficient dynamic scheduling and visual consistency of the complex 3D urban scene in a network environment.
Focusing on the problems of inefficient 3D model scheduling and low visual consistency of complex 3D urban scene visualization,a self-adaptive organization and scheduling method for 3D urban scenes driven by visual perception is proposed. Based on the spatial distribution of 3D city models,the adaptive quadtree partition was used to divide the complex 3D scene into different levels of granularity.Subsequently,traversing the quadtree from bottom to top,levels of detail( LODs) were generated for the intermediate nodes,and the geometric error of each layer was calculated to construct a flexible multi-granularity 3D tile model. Finally,visual perception parameters such as the shape,size,andheight of the building model were evaluated by employing the screen error,which was used to constrain the self-adaptive scheduling of LODs. The 3D models of four LODs in Berlin were selected for an interactive 3D roaming test. The experimental results show that the rendering frame rate was more than40 f/s based on the view-related hierarchical LOD dynamic scheduling,which could ensure efficient dynamic scheduling and visual consistency of the complex 3D urban scene in a network environment.
引文
[1]李德仁,朱欣焰,龚健雅.从数字地图到空间信息网格:空间信息多级网格理论思考[J].武汉大学学报:信息科学版,2003,28(6):642-650.LI Deren,ZHU Xinyan,GONG Jianya.From digital map to spatial information multi-grid:a thought of spatial information multi-grid theory[J].Geomatics and Information Science of Wuhan University,2003,28(6):642-650.
[2]戴晨光,邓雪清,张永生.海量地形数据实时可视化算法[J].计算机辅助设计与图形学学报,2004,16(11):1603-1607.DAI Chenguang,DENG Xueqing,ZHANG Yongsheng.An algorithm for real-time visualization of massive terrain dataset[J].Journal of Computer-Aided Design&Computer Graphics,2004,16(11):1603-1607.
[3]邬伦,刘瑜,张晶.地理信息系统:原理、方法和应用[M].北京:科学出版社,2001:156-158.
[4]FINKEL R A,BENTLEY J L.Quad trees a data structure for retrieval on composite keys[J].Acta Informatica,1974,4(1):1-9.
[5]仇林遥,王萌,朱庆,等.顾及时空语义的多主题瓦片数据优化检索方法[J].国防科技大学学报,2015,37(5):15-20.QIU Linyao,WANG Meng,ZHU Qing,et al.An optimal retrieval method of multi-theme image tiles considering the spatio-temporal semantics[J].Journal of National University of Defense Technology,2015,37(5):15-20.
[6]吴晨,朱庆,张叶廷,等.基于混合瓦片的海量DEM/DOM数据高效存储管理方法:以应急救灾数据库为例[J].地理信息世界,2014,21(3):69-72.WU Chen,ZHU Qing,ZHANG Yeting,et al.Efficient storage management of massive DEM/DOM data based on mixed tile range:a case study of emergency relief database[J].Geomatics World,2014,21(3):69-72.
[7]MAYER H.Automatic object extraction from aerial imagery:a survey focusing on buildings[J].Computer Vision and Image Understanding,1999,74(2):138-149.
[8]刘恒飞,刘纪平,王勇,等.格网划分与四叉树相结合的海量建筑物数据组织与调度[J].测绘通报,2010(11):4-6.LIU Hengfei,LIU Jiping,WANG Yong,et al.Building massive data organization and scheduling with the combination of grid partition and quadtree[J].Bulletin of Surveying and Mapping,2010(11):4-6.
[9]周圣川.大规模城市场景图形图像混合建模与视觉无损渲染技术[D].青岛:中国海洋大学,2014.
[10]胡金星,马照亭,吴焕萍,等.基于格网划分的海量地形数据三维可视化[J].计算机辅助设计与图形学学报,2004,16(8):1164-1168.HU Jinxing,MA Zhaoting,WU Huanping,et al.3D visualization of massive terrain data based on grid partition[J].Journal of Computer Aided Design&Computer Graphics,2004,16(8):1164-1168.
[11]张俊峰,姚志宏.基于四叉树孤立分割和屏幕误差的地形LOD算法[J].西南交通大学学报,2013,48(4):666-671.ZHANG Junfeng,YAO Zhihong.LOD algorithm of terrain based on conservative screen error and isolated division of quad-tree[J].Journal of Southwest Jiaotong University,2013,48(4):666-671.
[12]GUTHE M,BORODIN P,KLEIN R.Efficient viewdependent out-of-core visualization[C]∥Fourth Ionternational Conference on Virtual Reality and Its Applications in Industry.[S.l.]:International Society for Optics and Photonics,2004:428-438.
[13]HOPPE H.View-dependent refinement of progressive meshes[C]∥Proceedings of the 24th Annual Conference on Computer Graphics and Interactive Techniques.[S.l.]:ACM Press/Addison-Wesley Publishing Co.,1997:189-198.
[14]朱庆,龚俊.一种改进的真三维R树空间索引方法[J].武汉大学学报:信息科学版,2006,31(4):340-343.ZHU Qing,GONG Jun.An improved full 3D R-tree spatial index method[J].Geomatics and Information Science of Wuhan University,2006,31(4):340-343.
[15]龚俊,柯胜男,朱庆,等.一种八叉树和三维R树集成的激光点云数据管理方法[J].测绘学报,2012,41(4):597-604.GONG Jun,KE Shengnan,ZHU Qing.An efficient management method for point cloud data based on octree and 3D R-tree[J].Acta Geodaetica et Cartographica Sinica,2012,41(4):597-604.
[16]龚俊,朱庆,张叶廷,等.顾及多细节层次的三维R树索引扩展方法[J].测绘学报,2011,40(2):249-255.GONG Jun,ZHU Qing,ZHANG Yeting,et al.An efficient 3D R-tree Extension method concerned with levels of detail[J].Acta Geodaetica et Cartographica Sinica,2011,40(2):249-255.
[17]龚俊,朱庆,章汉武,等.基于R树索引的三维场景细节层次自适应控制方法[J].测绘学报,2011,40(4):531-534.GONG Jun,ZHU Qing,ZHANG Hanwu,et al.An adaptive control method of LODs for 3D scene based on Rtree index[J].Acta Geodaetica et Cartographica Sinica,2011,40(4):531-534.
[18]龚俊,谢潇.基于R树索引的三维可视化查询方法[J].武汉大学学报:信息科学版,2011,36(10):1140-1143,1153.GONG Jun,XIE Xiao.Three-dimension visualization query method based on R-tree[J].Geomatics and Information Science of Wuhan University,2011,36(10):1140-1143,1153.
[19]龚俊,柯胜男,朱庆,等.一种集成R树、哈希表和B*树的高效轨迹数据索引方法[J].测绘学报,2015,44(5):570-577.GONG Jun,KE Shengnan,ZHU Qing,et al.An efficient trajectory data index integrating R-tree,Hash and B*-tree[J].Acta Geodaetica et Cartographica Sinica,2015,44(5):570-577.
[20]高宇,邓宝松,陈鹏,等.基于连续HLOD的大规模场景快速绘制算法[J].小型微型计算机系统,2007,28(7):1267-1271.GAO Yu,DENG Baosong,CHEN Peng.Fast rendering algorithm for large-scale and complex scene based on continuous HLOD[J].Journal of Chinese Computer Systems,2007,28(7):1267-1271.
[21]GAO S,QI Y,SHEN X K,et al.A realtime rendering framework of large dataset environment based on precomputed hlod[C]∥Second Workshop on Digital Media and its Application in Museum&Heritages.[S.l.]:IEEE,2007:212-217.
[22]智明,郑姨婷.HLODs模型在场景快速绘制中的应用[J].计算机应用,2007,27(9):2126-2128.ZHI Ming,ZHENG Yiting.HLODs for fast rendering of large scale environments[J].Journal of Computer Applications,2007,27(9):2126-2128.
[23]曹巍.三维地形场景实时渲染技术相关算法的研究与改进[D].北京:中国地质大学,2009.
[24]张俊峰.大规模地形实时动态多分辨率显示关键算法研究[D].武汉:武汉大学,2011.
[2]戴晨光,邓雪清,张永生.海量地形数据实时可视化算法[J].计算机辅助设计与图形学学报,2004,16(11):1603-1607.DAI Chenguang,DENG Xueqing,ZHANG Yongsheng.An algorithm for real-time visualization of massive terrain dataset[J].Journal of Computer-Aided Design&Computer Graphics,2004,16(11):1603-1607.
[3]邬伦,刘瑜,张晶.地理信息系统:原理、方法和应用[M].北京:科学出版社,2001:156-158.
[4]FINKEL R A,BENTLEY J L.Quad trees a data structure for retrieval on composite keys[J].Acta Informatica,1974,4(1):1-9.
[5]仇林遥,王萌,朱庆,等.顾及时空语义的多主题瓦片数据优化检索方法[J].国防科技大学学报,2015,37(5):15-20.QIU Linyao,WANG Meng,ZHU Qing,et al.An optimal retrieval method of multi-theme image tiles considering the spatio-temporal semantics[J].Journal of National University of Defense Technology,2015,37(5):15-20.
[6]吴晨,朱庆,张叶廷,等.基于混合瓦片的海量DEM/DOM数据高效存储管理方法:以应急救灾数据库为例[J].地理信息世界,2014,21(3):69-72.WU Chen,ZHU Qing,ZHANG Yeting,et al.Efficient storage management of massive DEM/DOM data based on mixed tile range:a case study of emergency relief database[J].Geomatics World,2014,21(3):69-72.
[7]MAYER H.Automatic object extraction from aerial imagery:a survey focusing on buildings[J].Computer Vision and Image Understanding,1999,74(2):138-149.
[8]刘恒飞,刘纪平,王勇,等.格网划分与四叉树相结合的海量建筑物数据组织与调度[J].测绘通报,2010(11):4-6.LIU Hengfei,LIU Jiping,WANG Yong,et al.Building massive data organization and scheduling with the combination of grid partition and quadtree[J].Bulletin of Surveying and Mapping,2010(11):4-6.
[9]周圣川.大规模城市场景图形图像混合建模与视觉无损渲染技术[D].青岛:中国海洋大学,2014.
[10]胡金星,马照亭,吴焕萍,等.基于格网划分的海量地形数据三维可视化[J].计算机辅助设计与图形学学报,2004,16(8):1164-1168.HU Jinxing,MA Zhaoting,WU Huanping,et al.3D visualization of massive terrain data based on grid partition[J].Journal of Computer Aided Design&Computer Graphics,2004,16(8):1164-1168.
[11]张俊峰,姚志宏.基于四叉树孤立分割和屏幕误差的地形LOD算法[J].西南交通大学学报,2013,48(4):666-671.ZHANG Junfeng,YAO Zhihong.LOD algorithm of terrain based on conservative screen error and isolated division of quad-tree[J].Journal of Southwest Jiaotong University,2013,48(4):666-671.
[12]GUTHE M,BORODIN P,KLEIN R.Efficient viewdependent out-of-core visualization[C]∥Fourth Ionternational Conference on Virtual Reality and Its Applications in Industry.[S.l.]:International Society for Optics and Photonics,2004:428-438.
[13]HOPPE H.View-dependent refinement of progressive meshes[C]∥Proceedings of the 24th Annual Conference on Computer Graphics and Interactive Techniques.[S.l.]:ACM Press/Addison-Wesley Publishing Co.,1997:189-198.
[14]朱庆,龚俊.一种改进的真三维R树空间索引方法[J].武汉大学学报:信息科学版,2006,31(4):340-343.ZHU Qing,GONG Jun.An improved full 3D R-tree spatial index method[J].Geomatics and Information Science of Wuhan University,2006,31(4):340-343.
[15]龚俊,柯胜男,朱庆,等.一种八叉树和三维R树集成的激光点云数据管理方法[J].测绘学报,2012,41(4):597-604.GONG Jun,KE Shengnan,ZHU Qing.An efficient management method for point cloud data based on octree and 3D R-tree[J].Acta Geodaetica et Cartographica Sinica,2012,41(4):597-604.
[16]龚俊,朱庆,张叶廷,等.顾及多细节层次的三维R树索引扩展方法[J].测绘学报,2011,40(2):249-255.GONG Jun,ZHU Qing,ZHANG Yeting,et al.An efficient 3D R-tree Extension method concerned with levels of detail[J].Acta Geodaetica et Cartographica Sinica,2011,40(2):249-255.
[17]龚俊,朱庆,章汉武,等.基于R树索引的三维场景细节层次自适应控制方法[J].测绘学报,2011,40(4):531-534.GONG Jun,ZHU Qing,ZHANG Hanwu,et al.An adaptive control method of LODs for 3D scene based on Rtree index[J].Acta Geodaetica et Cartographica Sinica,2011,40(4):531-534.
[18]龚俊,谢潇.基于R树索引的三维可视化查询方法[J].武汉大学学报:信息科学版,2011,36(10):1140-1143,1153.GONG Jun,XIE Xiao.Three-dimension visualization query method based on R-tree[J].Geomatics and Information Science of Wuhan University,2011,36(10):1140-1143,1153.
[19]龚俊,柯胜男,朱庆,等.一种集成R树、哈希表和B*树的高效轨迹数据索引方法[J].测绘学报,2015,44(5):570-577.GONG Jun,KE Shengnan,ZHU Qing,et al.An efficient trajectory data index integrating R-tree,Hash and B*-tree[J].Acta Geodaetica et Cartographica Sinica,2015,44(5):570-577.
[20]高宇,邓宝松,陈鹏,等.基于连续HLOD的大规模场景快速绘制算法[J].小型微型计算机系统,2007,28(7):1267-1271.GAO Yu,DENG Baosong,CHEN Peng.Fast rendering algorithm for large-scale and complex scene based on continuous HLOD[J].Journal of Chinese Computer Systems,2007,28(7):1267-1271.
[21]GAO S,QI Y,SHEN X K,et al.A realtime rendering framework of large dataset environment based on precomputed hlod[C]∥Second Workshop on Digital Media and its Application in Museum&Heritages.[S.l.]:IEEE,2007:212-217.
[22]智明,郑姨婷.HLODs模型在场景快速绘制中的应用[J].计算机应用,2007,27(9):2126-2128.ZHI Ming,ZHENG Yiting.HLODs for fast rendering of large scale environments[J].Journal of Computer Applications,2007,27(9):2126-2128.
[23]曹巍.三维地形场景实时渲染技术相关算法的研究与改进[D].北京:中国地质大学,2009.
[24]张俊峰.大规模地形实时动态多分辨率显示关键算法研究[D].武汉:武汉大学,2011.