According to the characteristics of bore data,a model of 3D geologic body with generalized tri-prism as the primitive modeling element is constructed while the modeling process and key algorithms of modeling are prese...According to the characteristics of bore data,a model of 3D geologic body with generalized tri-prism as the primitive modeling element is constructed while the modeling process and key algorithms of modeling are presented here in detail.Using this method,the original bore data go through Delaunay triangulation to generate irregular triangular network on the surface,and then links stratum segments on the adjoining bores in session to form tri-prisms which would be pinched out.Finally stratified 3D geologic body model is built by an iterated search which searches for consecutive layer of the same property.The result shows that this method can effectively simulate stratified stratum modeling.展开更多
Data acquisition and modeling are the two important, difficult and costful aspects in a Cybercity project. 2D-GIS is mature and can manage a lot of spatial data. Thus 3D-GIS should make the best of data and technology...Data acquisition and modeling are the two important, difficult and costful aspects in a Cybercity project. 2D-GIS is mature and can manage a lot of spatial data. Thus 3D-GIS should make the best of data and technology of 2D-GIS. Construction of a useful synthetic environment requires integration of multiple types of information like DEM, texture images and 3D representation of objects such as buildings. In this paper, the method for 3D city landscape data model and visualization based on integrated databases is presented. Since the data volume of raster are very huge, special strategies(for example, pyramid gridded method) must be adopted in order to manage raster data efficiently. Three different methods of data acquisition, the proper data structure and a simple modeling method are presented as well. At last, a pilot project of Shanghai Cybercity is illustrated.展开更多
The 3D visualization model of slop with structural plane can displayed the characters of structural plane in slop directly, and illustrated the spatial combination. It is a modem and critical question in the field of ...The 3D visualization model of slop with structural plane can displayed the characters of structural plane in slop directly, and illustrated the spatial combination. It is a modem and critical question in the field of geotechnical engineering. Based on the peculiarity of the reconnaissance and the research of the visualization by formers, systemized the method fit for building 3D visualization model of slop with structural plane. Write the special program with Visual C^-＋ computer language and illustrated it by OpenGL, the program can displayed and captured the random section plane. The program has a satisfied result by proving with the real projects.展开更多
具身3D感知任务需通过自然语言指令,从第一人称视角理解动态环境.然而,当前基于大语言模型(large language models,LLMs)的具身3D感知方法仍存在局限:一些方法受限于语义输出范围与定位精度,另一些LLM方案缺乏能够有效聚合多视角语义与...具身3D感知任务需通过自然语言指令,从第一人称视角理解动态环境.然而,当前基于大语言模型(large language models,LLMs)的具身3D感知方法仍存在局限:一些方法受限于语义输出范围与定位精度,另一些LLM方案缺乏能够有效聚合多视角语义与几何特征的统一编码器,而这些特征对实现精准语言对齐与精确定位至关重要.为了解决这个问题,我们提出M3Fusion,首个面向具身3D感知的端到端多视角多模态统一融合框架.该框架紧密整合来自多个第一人称视角的2D视觉语义与3D几何特征,将其投影至共享3D空间,形成统一的M3-tokens.这些token不仅能够无缝对齐语言指令以理解复杂任务,还能同步解码生成精确的3D边界框.我们提出一种专门的三阶段训练策略以实现跨模态对齐.在3D视觉定位(3D visual grounding,3DVG)与3D视觉问答(3DQA)数据集上的评估表明:我们的框架在定位精度上显著提升,同时保持问答能力,充分证明了统一表征与框架设计的优越性.展开更多
3D stereoscopic visualization technology is coming into more and more common use in the field of entertainment,and this technology is also beginning to cut a striking figure in casting industry and scientific research...3D stereoscopic visualization technology is coming into more and more common use in the field of entertainment,and this technology is also beginning to cut a striking figure in casting industry and scientific research.The history,fundamental principle,and devices of 3D stereoscopic visualization technology are reviewed in this paper.The authors’research achievements on the 3D stereoscopic visualization technology in the modeling and simulation of the casting process are presented.This technology can be used for the observation of complex 3D solid models of castings and the simulated results of solidification processes such as temperature,fluid flow,displacement,stress strain and microstructure,as well as the predicted defects such as shrinkage/porosity,cracks,and deformation.It can also be used for other areas relating to 3D models,such as assembling of dies,cores,etc.Several cases are given to compare the illustration of simulated results by traditional images and red-blue 3D stereoscopic images.The spatial shape is observed better by the new method.The prospect of3D stereoscopic visualization in the casting aspect is discussed as well.The need for aided-viewing devices is still the most prominent problem of 3D stereoscopic visualization technology.However,3D stereoscopic visualization represents the tendency of visualization technology in the future;and as the problem is solved in the years ahead,great breakthroughs will certainly be made for its application in casting design and modeling and simulation of the casting processes.展开更多
New development in urban planning, cityscape, real estate management and the like, calls for new demands for 3D city model. There are so many objects in 3D city model such as building, river, road, and so on. The buil...New development in urban planning, cityscape, real estate management and the like, calls for new demands for 3D city model. There are so many objects in 3D city model such as building, river, road, and so on. The building model is very important in 3D city model. In recent years,a lot of research work about visualization has been done. In our opinions,visualization is only a part of 313 city model, while interactive operation about buildings is rather important as well. In order to implement interactive operation (create , edit, query,etc. ), good data structure and model must be developed.展开更多
This study proposes a detailed concept of how landscape changes can be transferred and communicated in a 3D environment using the storytelling approach. In 2018, Mocnik and Fairbairn argued that maps are good for repr...This study proposes a detailed concept of how landscape changes can be transferred and communicated in a 3D environment using the storytelling approach. In 2018, Mocnik and Fairbairn argued that maps are good for representing a geographic space but texts have a better benefit than maps for telling a story [1]. A framework is presented on how a landscape change can be retrieved out of textual descriptions. A prototype of a 3D model with a projection on top of it was developed. The case study tells the story of the landscape change in the city of Pristina. The key element of the visualization is a timeline. Several media like cadastral maps, orthophotos, texts, graphics, and background sound are used and combined in an animated light show and the visualization was evaluated within a user study.展开更多
In order to simplify the three-dimensional building group model, this paper proposes a clustering generalization method based on visual cognitive theory. The method uses road elements to roughly divide scenes, and the...In order to simplify the three-dimensional building group model, this paper proposes a clustering generalization method based on visual cognitive theory. The method uses road elements to roughly divide scenes, and then uses spatial cognitive elements such as direction, area, height and their topological constraints to classify them precisely, so as to make them conform to the urban morphological characteristics. Delaunay triangulation network and boundary tracking synthesis algorithm are used to merge and summarize the models, and the models are stored hierarchically. The proposed algorithm should be verified experimentally with a typical urban complex model. The experimental results show that the efficiency of the method used in this paper is at least 20% higher than that of previous one, and with the growth of test data, the higher efficiency is improved. The classification results conform to human cognitive habits, and the generalization levels of different models can be relatively unified by adaptive control of each threshold in the clustering generalization process.展开更多
文摘According to the characteristics of bore data,a model of 3D geologic body with generalized tri-prism as the primitive modeling element is constructed while the modeling process and key algorithms of modeling are presented here in detail.Using this method,the original bore data go through Delaunay triangulation to generate irregular triangular network on the surface,and then links stratum segments on the adjoining bores in session to form tri-prisms which would be pinched out.Finally stratified 3D geologic body model is built by an iterated search which searches for consecutive layer of the same property.The result shows that this method can effectively simulate stratified stratum modeling.
文摘Data acquisition and modeling are the two important, difficult and costful aspects in a Cybercity project. 2D-GIS is mature and can manage a lot of spatial data. Thus 3D-GIS should make the best of data and technology of 2D-GIS. Construction of a useful synthetic environment requires integration of multiple types of information like DEM, texture images and 3D representation of objects such as buildings. In this paper, the method for 3D city landscape data model and visualization based on integrated databases is presented. Since the data volume of raster are very huge, special strategies(for example, pyramid gridded method) must be adopted in order to manage raster data efficiently. Three different methods of data acquisition, the proper data structure and a simple modeling method are presented as well. At last, a pilot project of Shanghai Cybercity is illustrated.
文摘The 3D visualization model of slop with structural plane can displayed the characters of structural plane in slop directly, and illustrated the spatial combination. It is a modem and critical question in the field of geotechnical engineering. Based on the peculiarity of the reconnaissance and the research of the visualization by formers, systemized the method fit for building 3D visualization model of slop with structural plane. Write the special program with Visual C^-＋ computer language and illustrated it by OpenGL, the program can displayed and captured the random section plane. The program has a satisfied result by proving with the real projects.
文摘具身3D感知任务需通过自然语言指令,从第一人称视角理解动态环境.然而,当前基于大语言模型(large language models,LLMs)的具身3D感知方法仍存在局限:一些方法受限于语义输出范围与定位精度,另一些LLM方案缺乏能够有效聚合多视角语义与几何特征的统一编码器,而这些特征对实现精准语言对齐与精确定位至关重要.为了解决这个问题,我们提出M3Fusion,首个面向具身3D感知的端到端多视角多模态统一融合框架.该框架紧密整合来自多个第一人称视角的2D视觉语义与3D几何特征,将其投影至共享3D空间,形成统一的M3-tokens.这些token不仅能够无缝对齐语言指令以理解复杂任务,还能同步解码生成精确的3D边界框.我们提出一种专门的三阶段训练策略以实现跨模态对齐.在3D视觉定位(3D visual grounding,3DVG)与3D视觉问答(3DQA)数据集上的评估表明:我们的框架在定位精度上显著提升,同时保持问答能力,充分证明了统一表征与框架设计的优越性.
文摘3D stereoscopic visualization technology is coming into more and more common use in the field of entertainment,and this technology is also beginning to cut a striking figure in casting industry and scientific research.The history,fundamental principle,and devices of 3D stereoscopic visualization technology are reviewed in this paper.The authors’research achievements on the 3D stereoscopic visualization technology in the modeling and simulation of the casting process are presented.This technology can be used for the observation of complex 3D solid models of castings and the simulated results of solidification processes such as temperature,fluid flow,displacement,stress strain and microstructure,as well as the predicted defects such as shrinkage/porosity,cracks,and deformation.It can also be used for other areas relating to 3D models,such as assembling of dies,cores,etc.Several cases are given to compare the illustration of simulated results by traditional images and red-blue 3D stereoscopic images.The spatial shape is observed better by the new method.The prospect of3D stereoscopic visualization in the casting aspect is discussed as well.The need for aided-viewing devices is still the most prominent problem of 3D stereoscopic visualization technology.However,3D stereoscopic visualization represents the tendency of visualization technology in the future;and as the problem is solved in the years ahead,great breakthroughs will certainly be made for its application in casting design and modeling and simulation of the casting processes.
基金Project supported by the National Natural Science Foundation of China(No.69833010)
文摘New development in urban planning, cityscape, real estate management and the like, calls for new demands for 3D city model. There are so many objects in 3D city model such as building, river, road, and so on. The building model is very important in 3D city model. In recent years,a lot of research work about visualization has been done. In our opinions,visualization is only a part of 313 city model, while interactive operation about buildings is rather important as well. In order to implement interactive operation (create , edit, query,etc. ), good data structure and model must be developed.
文摘This study proposes a detailed concept of how landscape changes can be transferred and communicated in a 3D environment using the storytelling approach. In 2018, Mocnik and Fairbairn argued that maps are good for representing a geographic space but texts have a better benefit than maps for telling a story [1]. A framework is presented on how a landscape change can be retrieved out of textual descriptions. A prototype of a 3D model with a projection on top of it was developed. The case study tells the story of the landscape change in the city of Pristina. The key element of the visualization is a timeline. Several media like cadastral maps, orthophotos, texts, graphics, and background sound are used and combined in an animated light show and the visualization was evaluated within a user study.
文摘In order to simplify the three-dimensional building group model, this paper proposes a clustering generalization method based on visual cognitive theory. The method uses road elements to roughly divide scenes, and then uses spatial cognitive elements such as direction, area, height and their topological constraints to classify them precisely, so as to make them conform to the urban morphological characteristics. Delaunay triangulation network and boundary tracking synthesis algorithm are used to merge and summarize the models, and the models are stored hierarchically. The proposed algorithm should be verified experimentally with a typical urban complex model. The experimental results show that the efficiency of the method used in this paper is at least 20% higher than that of previous one, and with the growth of test data, the higher efficiency is improved. The classification results conform to human cognitive habits, and the generalization levels of different models can be relatively unified by adaptive control of each threshold in the clustering generalization process.
