基于视点的三维点云自适应多细节层次模型动态绘制
|
摘要
在计算机几何图像设计、医疗研究诊断、物体判定识别、自动化测量等领域应用的迫切需求下,需要对三维点云动态绘制方法进行研究。目前存在的基于图形处理单元(GPU)的三维点云动态绘制方法。首先根据高度场的形式获得三维点云数据;然后利用高通滤波法对获得的三维点云进行消噪处理;再根据Scene Graph的方法表示三维点云之间的逻辑关系;最后利用GPU完成三维点云的动态绘制。为了降低噪声对三维点云绘制的影响,绘制更加流畅清晰的画面,提出一种基于视点的三维点云自适应多细节层次(LOD)模型动态绘制方法。首先利用激光扫描仪器或结构光扫描仪器等获取三维点云数据,并利用双边滤波器进行平滑去噪;然后根据特征配对的方法对三维点云进行特征对的提取与配准;最后以三维点云特征对配准结果为依据,利用三维点云的内外存调度完成三维点云的自适应LOD动态绘制。实验结果分析证明,所提方法能有效地去除噪声,可以更好地描述目标物体或背景,使绘制画面更加流畅。
In the urgent need of computer geometric image design,medical research and diagnosis,object identification,automatic measurement and other fields,it is necessary to study the dynamic rendering method of threedimensional point clouds. The existing 3D point cloud rendering method based on graphics processing unit(GPU)obtained 3D point cloud data firstly according to the form of height field,and then high-pass filtering method was used to de-noise the obtained 3D point cloud. Then the logic relationship between 3D point clouds was represented by SceneGraph method,and the dynamic rendering of 3D point clouds is completed by GPU. In order to reduce the influence of noise on 3D point cloud rendering and draw a more fluent and clear picture,an adaptive 3D point cloud LOD dynamic rendering method based on viewpoint was proposed. Firstly,3D point cloud data were obtained by using laser scanning instrument or structured light scanning instrument. Then the feature pairs of 3D point clouds were extracted and registered according to the method of feature pairing. Finally,based on the registration results of 3D point cloud features,the adaptive LOD dynamic rendering of 3D point cloud was accomplished by using the internal and external memory scheduling of 3D point cloud. The experimental results show that the proposed method can remove noise effectively,describe the target object or background better,and make the picture more smooth.
In the urgent need of computer geometric image design,medical research and diagnosis,object identification,automatic measurement and other fields,it is necessary to study the dynamic rendering method of threedimensional point clouds. The existing 3D point cloud rendering method based on graphics processing unit(GPU)obtained 3D point cloud data firstly according to the form of height field,and then high-pass filtering method was used to de-noise the obtained 3D point cloud. Then the logic relationship between 3D point clouds was represented by SceneGraph method,and the dynamic rendering of 3D point clouds is completed by GPU. In order to reduce the influence of noise on 3D point cloud rendering and draw a more fluent and clear picture,an adaptive 3D point cloud LOD dynamic rendering method based on viewpoint was proposed. Firstly,3D point cloud data were obtained by using laser scanning instrument or structured light scanning instrument. Then the feature pairs of 3D point clouds were extracted and registered according to the method of feature pairing. Finally,based on the registration results of 3D point cloud features,the adaptive LOD dynamic rendering of 3D point cloud was accomplished by using the internal and external memory scheduling of 3D point cloud. The experimental results show that the proposed method can remove noise effectively,describe the target object or background better,and make the picture more smooth.
引文
1朱利,岳爱珍,周聪.面向城市建筑物3D点云模型的快速绘制方法[J].计算机辅助设计与图形学学报,2015,27(8):1442-1450Zhu Li,Yue Aizhen,Zhou Cong.A fast rendering method for 3D city building point cloud models[J].Journal of Computer-aided Design&Computer Graphics,2015,27(8):1442-1450
2陈国栋,叶东文,党琪琪.混合场景中阴影绘制方法研究[J].计算机工程,2016,42(2):242-248Chen Guodong,Ye Dongwen,Dang Qiqi.Research on shadow rendering method in hybrid scene[J].Computer Engineering,2016,42(2):242-248
3孙发军,刘建国.Word文档中二叉树的自动绘制方法研究[J].计算机应用与软件,2016,33(12):230-233Sun Fajun,Liu Jianguo.On automatic binary tree drawing method in word documents[J].Computer Applications and Software,2016,33(12):230-233
4丁莉.基于VS2008的实时曲线绘制方法研究[J].自动化技术与应用,2016,35(6):110-112Ding Li.Real time curve drawing method based on the VS2008[J].Techniques of Automation and Applications,2016,35(6):110-112
5吕梦春,张晓林.一种去除虚假边缘的任意视点绘制方法[J].电子设计工程,2016,24(15):187-190LüMengchun,Zhang Xiaolin.A method of ghost contour removal for free-viewpoint[J].Electronic Design Engineering,2016,24(15):187-190
6王鹏,叶冰.基于NI-DAQ的高采样实时数据绘制方法与实现[J].计算机测量与控制,2015,23(5):1660-1663Wang Peng,Ye Bing.Curve plotting method of high sampling realtime data based on NI-DAQ[J].Computer Measurement&Control,2015,23(5):1660-1663
7芦文峰,王凯,吴亚非,等.图形绘制引擎的效率和跨平台研究[J].计算机工程与设计,2016,37(5):1400-1404Lu Wenfeng,Wang Kai,Wu Yafei,et al.Research on efficiency and cross-platform of graphics rendering engine[J].Computer Engineering and Design,2016,37(5):1400-1404
8潘云峰,李福翠,邵枫.一种面向视点绘制的深度图像后处理方法[J].计算机工程与应用,2015,51(11):158-161Pan Yunfeng,Li Fucui,Shao Feng.Depth map post-processing method for view synthesis[J].Computer Engineering and Applications,2015,51(11):158-161
9娄达平,王晓东,富显祖,等.基于深度图像分割的虚拟视点绘制[J].计算机工程,2016,42(10):12-19Lou Daping,Wang Xiaodong,Fu Xianzu,et al.Virtual view point rendering based on range image segmentation[J].Computer Engineering,2016,42(10):12-19
10邓博文,刘春松,吴凡贤,等.基于GPU的地震剖面图形快速绘制算法[J].计算机科学,2016,43(4):294-298Deng Bowen,Liu Chunsong,Wu Fanxian,et al.Fast graphical rendering of seismic profile algorithm based on GPU[J].Computer Science,2016,43(4):294-298
2陈国栋,叶东文,党琪琪.混合场景中阴影绘制方法研究[J].计算机工程,2016,42(2):242-248Chen Guodong,Ye Dongwen,Dang Qiqi.Research on shadow rendering method in hybrid scene[J].Computer Engineering,2016,42(2):242-248
3孙发军,刘建国.Word文档中二叉树的自动绘制方法研究[J].计算机应用与软件,2016,33(12):230-233Sun Fajun,Liu Jianguo.On automatic binary tree drawing method in word documents[J].Computer Applications and Software,2016,33(12):230-233
4丁莉.基于VS2008的实时曲线绘制方法研究[J].自动化技术与应用,2016,35(6):110-112Ding Li.Real time curve drawing method based on the VS2008[J].Techniques of Automation and Applications,2016,35(6):110-112
5吕梦春,张晓林.一种去除虚假边缘的任意视点绘制方法[J].电子设计工程,2016,24(15):187-190LüMengchun,Zhang Xiaolin.A method of ghost contour removal for free-viewpoint[J].Electronic Design Engineering,2016,24(15):187-190
6王鹏,叶冰.基于NI-DAQ的高采样实时数据绘制方法与实现[J].计算机测量与控制,2015,23(5):1660-1663Wang Peng,Ye Bing.Curve plotting method of high sampling realtime data based on NI-DAQ[J].Computer Measurement&Control,2015,23(5):1660-1663
7芦文峰,王凯,吴亚非,等.图形绘制引擎的效率和跨平台研究[J].计算机工程与设计,2016,37(5):1400-1404Lu Wenfeng,Wang Kai,Wu Yafei,et al.Research on efficiency and cross-platform of graphics rendering engine[J].Computer Engineering and Design,2016,37(5):1400-1404
8潘云峰,李福翠,邵枫.一种面向视点绘制的深度图像后处理方法[J].计算机工程与应用,2015,51(11):158-161Pan Yunfeng,Li Fucui,Shao Feng.Depth map post-processing method for view synthesis[J].Computer Engineering and Applications,2015,51(11):158-161
9娄达平,王晓东,富显祖,等.基于深度图像分割的虚拟视点绘制[J].计算机工程,2016,42(10):12-19Lou Daping,Wang Xiaodong,Fu Xianzu,et al.Virtual view point rendering based on range image segmentation[J].Computer Engineering,2016,42(10):12-19
10邓博文,刘春松,吴凡贤,等.基于GPU的地震剖面图形快速绘制算法[J].计算机科学,2016,43(4):294-298Deng Bowen,Liu Chunsong,Wu Fanxian,et al.Fast graphical rendering of seismic profile algorithm based on GPU[J].Computer Science,2016,43(4):294-298