大规模复杂场景真实感快速绘制
文献类型：学位论文
| 作者 | 鲍冠伯 |
| 学位类别 | 工学博士 |
| 答辩日期 | 2012-05-28 |
| 授予单位 | 中国科学院研究生院 |
| 授予地点 | 中国科学院自动化研究所 |
| 导师 | 张晓鹏 ; 胡包钢 |
| 关键词 | 多分辨率模型 图形处理器 实时绘制 实例化绘制 森林植物 Level of detail (LOD) GPU real-time rendering instancing forest plants |
| 其他题名 | Real-time Realistic Rendering of Large-scale Complex Scenes |
| 学位专业 | 计算机应用技术 |
| 中文摘要 | 大规模复杂场景的真实感快速绘制是计算机图形学领域中重要的研究课题，在数字娱乐、虚拟仿真、科普教学、园林规划和数字农林业研究等方面有着广泛的应用。本文以三维几何植物模型为对象，研究大规模复杂场景的真实感实时绘制。本文的主要内容包括植物模型的广告牌简化和绘制，森林场景真实感实时绘制和大规模复杂场景的硬件实例化绘制。本文主要的工作和贡献有： 1) 基于广告牌技术的思想设计并实现了一个自动对植物模型进行简化的工具，该工具无需人工交互可以将复杂的植物三维网格模型自动地简化成一个由广告牌和网格组成的混合简化模型以及一个由三个相互正交的广告牌组成的纯广告牌模型。其中混合简化模型在森林绘制时用于表示前景的树木，纯广告牌模型用于表示森林背景处的树木。树木模型的混合简化算法将树木模型中的树冠部分生成一组任意朝向的广告牌云来代替，而保持原来树干部分的网格模型，这样既极大减少了模型的多边形数量又保持了原有模型的外形轮廓，在实时绘制时混合简化模型中广告牌随视线方向动态地改变，具有良好的视觉效果。 2) 提出了一种森林场景真实感快速绘制方法。该方法综合集成了三维树木模型的多分辨率模型（LOD）的生成，复杂场景的阴影实时计算，以及适宜于网络应用的森林场景数据的组织与传输。我们提出了一个叶序模型的提取方法，在保持视觉效果的情况下，把网格模型和纹理更好地匹配起来；我们在多阴影图方法上结合了一个阴影图的LOD更新策略，不仅减少了绘制的几何复杂度，而且减少了GPU的显存消耗；我们采用了一种场景数据的编码方法，利于对森林场景数据进行网络传输。该方法实现了适宜于网络应用的森林场景真实感实时绘制。 3) 提出了一种基于硬件的实例化绘制算法，该算法在支持OpenGL 4的显卡上能够实时绘制包含一百多万个复杂三维几何模型的大规模场景，同时实时绘制场景的阴影，无需场景管理算法的帮助。该算法克服了目前显卡只能将几何图元输出到有限个顶点流（Vertex stream）的限制，能够同时绘制多个场景对象的多级LOD模型。模型的视锥裁剪和LOD选择操作完全在GPU中完成，无需CPU交互，极大减少了每帧调用的图形绘制命令数量，提高了绘制效率。与Kinect的结合应用显示了该算法的有效性。 4）我们进一步通过合并多个模型对象，在几何着色器中控制单级LOD的输出以及模型位置点排序，改进了硬件实例化绘制算法，实现了对带有不同透明度的植物线状细枝模型的半透明的硬件实例化绘制。 |
| 英文摘要 | Real-time rendering of large-scale complex scenes is important in many applications, such as video games, Internet graphics, and landscape and cityscape scene design and visualization. We take three-dimensional geometric tree models to study the real-time rendering of large-scale complex scenes. The main content of this thesis includes simplifying plant models using billboards，realistic real-time scenes rendering through the web and hardware instancing of large complex scenes. Main contributions of this thesis are as follows: 1) We design a tree model simplification tool, which can automatically generate both hybrid polygon/image-based tree models for near distance viewing and three-cross-billboard representations of trees for medium and far distance viewing during a preprocessing phase. In the hybrid polygon/image-based tree model, a set of arbitrarily oriented billboards is used to represent the geometry of the tree foliage. The three-cross-billboards can be used in the background for the fast distance representation. This approach allows real-time rendering and achieves a good parallax effect, which is important when the camera is moving: the tree seems to have a real 3D shape and not flat as only with traditional billboards. We also develop a fast forest rendering system using the simplified plant models, which includes dynamic changing of foliage billboards, texture atlases, and quick distance sorts. 2) We present a new framework for rendering large-scale forest scenes realistically and quickly that integrates extracting level of detail (LOD) tree models, rendering real-time shadows for large-scale forests, and transmitting forest data for network applications. We construct a series of LOD tree models to compress the overall complexity of the forest in view-dependent forest navigation. A new leaf phyllotaxy LOD modeling method is presented to match leaf models with textures, balancing the visual effect and model complexity. To progressively render the scene from coarse to fine, sequences of LOD models are transferred from simple to complex. The forest can be rendered after obtaining a simple model of each tree, allowing users to quickly see a sketch of the scene. We also adopt a LOD strategy for shadow maps. Experiments show that the large-scale forest scenes can be rendered with smooth shadows in real-time. 3) We present a novel method for rendering over one million highly detailed trees in real-time with real-time shadows. By taking advantage o... |
| 语种 | 中文 |
| 其他标识符 | 200818014629084 |
| 源URL | [http://ir.ia.ac.cn/handle/173211/6428]  |
| 专题 | 毕业生_博士学位论文 |
| 推荐引用方式 GB/T 7714 | 鲍冠伯. 大规模复杂场景真实感快速绘制[D]. 中国科学院自动化研究所. 中国科学院研究生院. 2012. |
入库方式： OAI收割
来源：自动化研究所
浏览0
下载0
收藏0
其他版本
除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。