数字景区三维建模关键技术研究
|
摘要
数字景区三维建模包括数据获取、图像分割、图像配准、几何建模、纹理建模等技术,其中数据获取、图像分割和图像配准是三维建模的基础,对建模的效果和质量有着重要的意义。本文主要针对数据获取、图像分割和图像配准这几个问题做了研究与探索。
当前对数据获取的研究都偏于具体的数字城市,具体的景物,而缺乏对数字景区的一般性研究。本文对景区三维建模所需的数据获取方法根据对象的不同进行了分类研究,并针对不同需求情况提出了具体的方法。而对于图形分割和配准当前研究又多侧重在精细和准确上,而数字景区面对的主要服务对象是广大的游客,他们关注的是虚拟景区的美丽景色和高效快捷的服务,这样数字景区中的分割和配准在针对游客中主要考虑的是速度和美观,对精确度要求不是太高。
本文在景区三维建模理论的基础上,对景区建模的关键技术图像分割进行了研究,如果图像前景物体的内部具有均匀一致的灰度值,并分布在另一个灰度值的均匀背景上,那么图像的灰度直方图将有明显的双峰,可以使用峰谷法进行阈值分割。本文根据景区的实际情况,提出了对具有二值倾向的雕塑图像采取阈值分割,并对其进行了优化,在理论上证明了理想状况下其误差率为极小,为三维建模的后期工作做了准备。最后本文在分割的基础上,对景区建模的关键技术图像配准进行了研究,结合数字景区建模实际,提出了数字景区图像单点配准方法,利用该技术对基于google地图的青岛某景区图像进行了配准,本文所给出的方法算法简单,易实现,因为原图是逐个像素得到的,每个像素的灰度是一步插值确定的,所以映射效率较高,并且由于不需要较多寻址而实现快速。但它的缺点也是显而易见的,那就是不够精确。这种方法可用在对速度有较高要求,但对精度要求不高的情况。
Data acquisition, image division, image matching, geometric modeling, texture modeling are the 3D modeling techniques of the digital scenic area . Data acquisition, image division and image matching are the key technologies of the 3D modeling of the digital scenic area , They have the vital significance to the quality and effect of the modelling. In this paper, a research and exploration will illustrate some solution about data acquisition, image division and image matching.
The current study of the data acquisition is based to obtain a specific data of the cities, specific features, and it is lack of a general study of the digital scenic area. In this paper, the data acquisition method of 3D modeling of the scenic requirements are in accordance with the different objects classified research, and the specific method is put forward for the different demand. As for the graphics division and matching, the current research is focused on more sophisticated and accurate, but the digital scenic areas are mainly serve the vast number of tourists, which are concerned about the beauty of virtual scenic and the fast and efficient of the service.The division and matching for the tourists in the main consideration is the speed and appearance, and the accuracy requirements are not very high.
In this paper, image division have been studied ,which based on the key technologies for modeling scenic. if the prospect-objects of the image have uniform gray value, which are distributed in another uniform gray background , then the image histogram peaks will be obvious, we can use the peak-valley segmentation method. In this paper, based on the actual situation of scenic spots, proposed that the sculpture image with two values tending may be taken threshold segmentation, and its optimization, in theory, is proved that the ideal situation have a very small error rate, which are prepared for 3D modeling of the late work. Finally, scenic-image-matching techniques have been studied, on the basis of the image division. Combined with modeling of the digital scenic areas, a single-point matching method have been proposed. Using the technology , a scenic image of Qingdao based on a google map have been matched. the method given in this paper is simple and easy to achieve, because the image is received by pixel, each pixel gray level is determined by one interpolation step. And as a result of not addressing more, this method is more efficient and fast. But it also have obvious shortcomings, that is not sufficiently precise. This method can be used in higher speed, but not require much on the accuracy of the situation.
当前对数据获取的研究都偏于具体的数字城市,具体的景物,而缺乏对数字景区的一般性研究。本文对景区三维建模所需的数据获取方法根据对象的不同进行了分类研究,并针对不同需求情况提出了具体的方法。而对于图形分割和配准当前研究又多侧重在精细和准确上,而数字景区面对的主要服务对象是广大的游客,他们关注的是虚拟景区的美丽景色和高效快捷的服务,这样数字景区中的分割和配准在针对游客中主要考虑的是速度和美观,对精确度要求不是太高。
本文在景区三维建模理论的基础上,对景区建模的关键技术图像分割进行了研究,如果图像前景物体的内部具有均匀一致的灰度值,并分布在另一个灰度值的均匀背景上,那么图像的灰度直方图将有明显的双峰,可以使用峰谷法进行阈值分割。本文根据景区的实际情况,提出了对具有二值倾向的雕塑图像采取阈值分割,并对其进行了优化,在理论上证明了理想状况下其误差率为极小,为三维建模的后期工作做了准备。最后本文在分割的基础上,对景区建模的关键技术图像配准进行了研究,结合数字景区建模实际,提出了数字景区图像单点配准方法,利用该技术对基于google地图的青岛某景区图像进行了配准,本文所给出的方法算法简单,易实现,因为原图是逐个像素得到的,每个像素的灰度是一步插值确定的,所以映射效率较高,并且由于不需要较多寻址而实现快速。但它的缺点也是显而易见的,那就是不够精确。这种方法可用在对速度有较高要求,但对精度要求不高的情况。
Data acquisition, image division, image matching, geometric modeling, texture modeling are the 3D modeling techniques of the digital scenic area . Data acquisition, image division and image matching are the key technologies of the 3D modeling of the digital scenic area , They have the vital significance to the quality and effect of the modelling. In this paper, a research and exploration will illustrate some solution about data acquisition, image division and image matching.
The current study of the data acquisition is based to obtain a specific data of the cities, specific features, and it is lack of a general study of the digital scenic area. In this paper, the data acquisition method of 3D modeling of the scenic requirements are in accordance with the different objects classified research, and the specific method is put forward for the different demand. As for the graphics division and matching, the current research is focused on more sophisticated and accurate, but the digital scenic areas are mainly serve the vast number of tourists, which are concerned about the beauty of virtual scenic and the fast and efficient of the service.The division and matching for the tourists in the main consideration is the speed and appearance, and the accuracy requirements are not very high.
In this paper, image division have been studied ,which based on the key technologies for modeling scenic. if the prospect-objects of the image have uniform gray value, which are distributed in another uniform gray background , then the image histogram peaks will be obvious, we can use the peak-valley segmentation method. In this paper, based on the actual situation of scenic spots, proposed that the sculpture image with two values tending may be taken threshold segmentation, and its optimization, in theory, is proved that the ideal situation have a very small error rate, which are prepared for 3D modeling of the late work. Finally, scenic-image-matching techniques have been studied, on the basis of the image division. Combined with modeling of the digital scenic areas, a single-point matching method have been proposed. Using the technology , a scenic image of Qingdao based on a google map have been matched. the method given in this paper is simple and easy to achieve, because the image is received by pixel, each pixel gray level is determined by one interpolation step. And as a result of not addressing more, this method is more efficient and fast. But it also have obvious shortcomings, that is not sufficiently precise. This method can be used in higher speed, but not require much on the accuracy of the situation.
引文
1.朱元嫣.旅游场景三维可视化方案[J].上海师范大学,2007,36(4):98-103
2.李莹,陈启祥.基于OEPENGL技术与3DS MAX的3D虚拟校园情景开发[J].武汉船舶职业技术学院学报,2007,(3):37-40.
3.崔杏园,钱桦.虚拟现实及其演变发展[J].工程地质计算机应用,2005,(4):56-57
4.宫本红.面向虚拟现实的古代与现代建筑优化建模技[D].山东青岛:中国海洋大学,2006:12-16
5.段学军,陈铭,王晓斌.虚拟城市场景建模方法与技术研究[J].系统仿真学报,2003,15(10):14-49
6.王秋玲,苗放,叶成名.基于数字地球平台的数字旅游应用研究[J].资源与产业,2008,10(3):108-111
7.徐则中,庄燕滨.三维建模系统的综述[J].测绘通报,2008,(2):16-19
8.DIAS P,SEQUEIRAV.Registration and Fusion of Intensity and Range Data for3DModelling of RealWorld Scenes[J].IEEE Conference on 3D Digital Imaging and Modeling,2003,5(2):418-425.
9.STAMOS I,ALLEN P E.3-DModelConstruction Using Range and Image Data[J].CVPR,2000,(1):531-537
10.DING Y,PING X.Range Image Segmentation Based on Randomized Hough Transform [J].PRL,2005,13(26):2033-2041
11.XIANG R,WANG R.Range Image Segmentation Based on Split-merge Clustering[J].ICPR04,2004,3(1):614-617
12.史文中,李必军,李清泉.基于投影点密度的车载激光扫描距离图像分割方法[J].测绘学报,2005,34(2):95-101
13.马大喜.关于三维城市模型数据获取的研究[J].科技情报开发与经济.2007,17(2):31-33
14.李志林,朱庆.数字高程模型[M].武汉:武汉大学出版社,2003:28-29
15.傅咏冬.三维城市建模中正射影像图的处理技巧[J].测绘技术装备,2004,6(3):52-55
16.马大喜.关于三维城市模型数据获取的研究[J].科技情报开发与经济,2007, 17(2):32-35
17.朱程程.城市战场三维可视化及应用研究[D].河南郑州:解放军信息工程大学,2007:28-32
18.刘立娜.虚拟城市建设中建模与可视化的研究与实现[D].河南郑州:解放军信息工程大学测绘学院,2006:32-35
19.张青蓉,王文永,付宏杰.虚拟植物的构建及在生物学科教学中的应用[J].系统仿真学报,2006,18(2):964-967
20.王剑,周国民.利用激光扫描仪获取植物三维模型的方法[J].湖北农业科学,2008,47(1):104-106
21.应骏,叶秀清,顾伟康.一个基于知识的边沿提取算法[J].中国图象图形学报,1999,4(3):239-24
22.罗希平,田捷,诸葛婴.图像分割方法综述[J].模式识别与人工智能,1999,12(3):300-312
23.Nalwa Vand BinfordT.On DetectingEdges.IEEETrans.PatternAnalysis and Machine Intelligence,1986,8(6):699-714
24.Bomana M.3D segmentation of MR images of the head for 3D display[J].IEEE Trans.PatternAnalysis andMachine Intelligence,1990,3(1):177-183
25.Bergholm F.Edge focusing.IEEETrans.PatternAnalysis andMachine Intelligence,1987,9(9):726-741
26.杨晖,曲秀杰.图像分割方法综述[J].电脑开发与应用,2005,18(3):21-23
27.王爱民,沈兰荪.图像分割研究综述[J].测绘技术,2000,19(5):1-3
28.刘文萍,吴立德.图像分割中阈值选取方法比较研究[J].模式识别与人工智能,1997,10(3):271-277
29.Brink A B.Gray level thresholding of images using a correlation criterion[J].Pattern Recognition Letters,1989,7(6):335-341
30.靳宏磊,朱蔚萍,李立源,陈维南.二维灰度直方图的最佳分割方法[J].模式识别与人工智能,1999,12(3):329-333
31.陈燕新,戚飞虎.基于竞争Hopfield网络的自动聚类图像分割方法[J].模式识别与人工智能,1998,11(2):215-221
32.胡世英,周源华.模糊选择多分辨率Kohonen聚类网络用于灰度图像分割[J].电子 学报,1999,27(10):34-37
33.尹平,王润生.基于边缘信息的分开合并图像分割算法[J].中国图象图形学报,1998,3(6):450-454
34.Wang J P.Stochastic Relaxation on Partitions with Connected Components and Its Application to Image Segmentation[J].IEEE Trans.Pattern Analysis and Machine Intelligence,1998,20(8):619-636
35.戴剑彬,张大力.图像分析中的松弛标记法[J].中国图象图形学报,1998,3(2):96-99
36.孙家广.计算机图形学[M].北京:清华大学出版社,1998:568-572
37.杨宇,张琦.半自动图像分割[J].中国传媒大学学报自然科学版,2008,16(3):26-31
38.章毓晋.图像分割[M].北京:北京科学出版社,2001:14-15
39.Brown L.G..A.survey of image registration techniques[J].ACM Computer,1992,24(4):325-376
40.章毓晋.图像处理[M].北京:清华大学出版社,2006:58-69
41.Rusinck H,Levy A.Performance of two methods for registering PET and MR brain scans[J].Conf Record 1991.IEEE Nuclear Science Symp Med Image Conf,1991,3(4):21-95
42.Besl PJ,Macky ND.A method for registration of 3-D shapes[J].IEEE Trans patt Anal Mach Intell,1992,14(4):239-240
43.Rosenfeld A.Kak A.C..Digital Picture Processing[J].Academic Press,Orlando,FL,1982,1(1):420-425
44.Svedlow M.,McGillem C.D.,Anuta P.E..Experimental examination of similarity measures and preprocessing methods used for image registration[J].Symposium on Machine Processing of Remotely Sensed Data,1976,9(12):4-9
45.Philippe Thevenaz,Michael Unser.Optimization of mutual information for multiresolution image registration[J].IEEE Trans.On Image Processing,2000,9(12):2083-2099
46.Reddy,Chatterji.An fft-based technique for translation and scale-invariant image registration[J].IEEE Trans.on Image Processing,1996,3(8):1266-1270
47.贾凤美.插值在图像(图形)处理中的使用[J].科学技术与工程,2007,(7):129-131
48.章毓晋.图像处理[M].北京:清华大学出版社,2006:71-73
49.陈宏盛.刘雨.计算方法[M].长沙:国防科技大学出版社,2001:247-246
50.刘勇奎.计算机图形学的基础算法[M].北京:科学出版社,2007:179-181
51.孙即祥.数字图像处理[M].石家庄:河北教育出版社,1993:157-159
52.罗笑南,王若梅.计算机图像学[M].广州:中山大学出版社,1996:136-182
2.李莹,陈启祥.基于OEPENGL技术与3DS MAX的3D虚拟校园情景开发[J].武汉船舶职业技术学院学报,2007,(3):37-40.
3.崔杏园,钱桦.虚拟现实及其演变发展[J].工程地质计算机应用,2005,(4):56-57
4.宫本红.面向虚拟现实的古代与现代建筑优化建模技[D].山东青岛:中国海洋大学,2006:12-16
5.段学军,陈铭,王晓斌.虚拟城市场景建模方法与技术研究[J].系统仿真学报,2003,15(10):14-49
6.王秋玲,苗放,叶成名.基于数字地球平台的数字旅游应用研究[J].资源与产业,2008,10(3):108-111
7.徐则中,庄燕滨.三维建模系统的综述[J].测绘通报,2008,(2):16-19
8.DIAS P,SEQUEIRAV.Registration and Fusion of Intensity and Range Data for3DModelling of RealWorld Scenes[J].IEEE Conference on 3D Digital Imaging and Modeling,2003,5(2):418-425.
9.STAMOS I,ALLEN P E.3-DModelConstruction Using Range and Image Data[J].CVPR,2000,(1):531-537
10.DING Y,PING X.Range Image Segmentation Based on Randomized Hough Transform [J].PRL,2005,13(26):2033-2041
11.XIANG R,WANG R.Range Image Segmentation Based on Split-merge Clustering[J].ICPR04,2004,3(1):614-617
12.史文中,李必军,李清泉.基于投影点密度的车载激光扫描距离图像分割方法[J].测绘学报,2005,34(2):95-101
13.马大喜.关于三维城市模型数据获取的研究[J].科技情报开发与经济.2007,17(2):31-33
14.李志林,朱庆.数字高程模型[M].武汉:武汉大学出版社,2003:28-29
15.傅咏冬.三维城市建模中正射影像图的处理技巧[J].测绘技术装备,2004,6(3):52-55
16.马大喜.关于三维城市模型数据获取的研究[J].科技情报开发与经济,2007, 17(2):32-35
17.朱程程.城市战场三维可视化及应用研究[D].河南郑州:解放军信息工程大学,2007:28-32
18.刘立娜.虚拟城市建设中建模与可视化的研究与实现[D].河南郑州:解放军信息工程大学测绘学院,2006:32-35
19.张青蓉,王文永,付宏杰.虚拟植物的构建及在生物学科教学中的应用[J].系统仿真学报,2006,18(2):964-967
20.王剑,周国民.利用激光扫描仪获取植物三维模型的方法[J].湖北农业科学,2008,47(1):104-106
21.应骏,叶秀清,顾伟康.一个基于知识的边沿提取算法[J].中国图象图形学报,1999,4(3):239-24
22.罗希平,田捷,诸葛婴.图像分割方法综述[J].模式识别与人工智能,1999,12(3):300-312
23.Nalwa Vand BinfordT.On DetectingEdges.IEEETrans.PatternAnalysis and Machine Intelligence,1986,8(6):699-714
24.Bomana M.3D segmentation of MR images of the head for 3D display[J].IEEE Trans.PatternAnalysis andMachine Intelligence,1990,3(1):177-183
25.Bergholm F.Edge focusing.IEEETrans.PatternAnalysis andMachine Intelligence,1987,9(9):726-741
26.杨晖,曲秀杰.图像分割方法综述[J].电脑开发与应用,2005,18(3):21-23
27.王爱民,沈兰荪.图像分割研究综述[J].测绘技术,2000,19(5):1-3
28.刘文萍,吴立德.图像分割中阈值选取方法比较研究[J].模式识别与人工智能,1997,10(3):271-277
29.Brink A B.Gray level thresholding of images using a correlation criterion[J].Pattern Recognition Letters,1989,7(6):335-341
30.靳宏磊,朱蔚萍,李立源,陈维南.二维灰度直方图的最佳分割方法[J].模式识别与人工智能,1999,12(3):329-333
31.陈燕新,戚飞虎.基于竞争Hopfield网络的自动聚类图像分割方法[J].模式识别与人工智能,1998,11(2):215-221
32.胡世英,周源华.模糊选择多分辨率Kohonen聚类网络用于灰度图像分割[J].电子 学报,1999,27(10):34-37
33.尹平,王润生.基于边缘信息的分开合并图像分割算法[J].中国图象图形学报,1998,3(6):450-454
34.Wang J P.Stochastic Relaxation on Partitions with Connected Components and Its Application to Image Segmentation[J].IEEE Trans.Pattern Analysis and Machine Intelligence,1998,20(8):619-636
35.戴剑彬,张大力.图像分析中的松弛标记法[J].中国图象图形学报,1998,3(2):96-99
36.孙家广.计算机图形学[M].北京:清华大学出版社,1998:568-572
37.杨宇,张琦.半自动图像分割[J].中国传媒大学学报自然科学版,2008,16(3):26-31
38.章毓晋.图像分割[M].北京:北京科学出版社,2001:14-15
39.Brown L.G..A.survey of image registration techniques[J].ACM Computer,1992,24(4):325-376
40.章毓晋.图像处理[M].北京:清华大学出版社,2006:58-69
41.Rusinck H,Levy A.Performance of two methods for registering PET and MR brain scans[J].Conf Record 1991.IEEE Nuclear Science Symp Med Image Conf,1991,3(4):21-95
42.Besl PJ,Macky ND.A method for registration of 3-D shapes[J].IEEE Trans patt Anal Mach Intell,1992,14(4):239-240
43.Rosenfeld A.Kak A.C..Digital Picture Processing[J].Academic Press,Orlando,FL,1982,1(1):420-425
44.Svedlow M.,McGillem C.D.,Anuta P.E..Experimental examination of similarity measures and preprocessing methods used for image registration[J].Symposium on Machine Processing of Remotely Sensed Data,1976,9(12):4-9
45.Philippe Thevenaz,Michael Unser.Optimization of mutual information for multiresolution image registration[J].IEEE Trans.On Image Processing,2000,9(12):2083-2099
46.Reddy,Chatterji.An fft-based technique for translation and scale-invariant image registration[J].IEEE Trans.on Image Processing,1996,3(8):1266-1270
47.贾凤美.插值在图像(图形)处理中的使用[J].科学技术与工程,2007,(7):129-131
48.章毓晋.图像处理[M].北京:清华大学出版社,2006:71-73
49.陈宏盛.刘雨.计算方法[M].长沙:国防科技大学出版社,2001:247-246
50.刘勇奎.计算机图形学的基础算法[M].北京:科学出版社,2007:179-181
51.孙即祥.数字图像处理[M].石家庄:河北教育出版社,1993:157-159
52.罗笑南,王若梅.计算机图像学[M].广州:中山大学出版社,1996:136-182