CASIA OpenIR  > 毕业生  > 硕士学位论文
Alternative TitleAnalysis of Scene Reflection and Illumination
Thesis Advisor卢汉清 ; 沈向洋
Degree Grantor中国科学院研究生院
Place of Conferral中国科学院自动化研究所
Degree Discipline模式识别与智能系统
Keyword反射分量分离 反射属性估计 光照估计 Diffuse/specular Separation Reflectance Estimation Illumination Estimation
Abstract本文论述了如何由图像恢复场景反射和光照方面信息的问题。世界的外观 是由光和物质之间的相互作用决定的。光照条件或物体表面反射属性的变化会 引起场景外观的显著变化,镜面反射也使得世界的外观更加多样化。对反射和 光照的理解,长期以来都是物理学和人类感知方面研究的重要问题。而在计算 机视觉和图形学方面,由图像分析反射和光照,也日渐成为非常重要的课题。 本文的研究工作主要关注下面三个问题,即:从图像中进行反射分量分 离,反射属性估计,和光照估计。 首先,本文提出了一种方法,基于颜色分析和多基线立体视觉,分离图像 中的漫反射分量和镜面反射分量,同时估计高光区域的准确景深。这种方法利 用极线几何约束颜色分析进行高光检测,并利用高光检测的结果来约束立体景 深的求解。避免了以往方法的局限性,如要求准确分割、区域不变的反射属 性、或经过控制的光照等等。 第二,本文提出了一种在一幅图像中估计表面反射属性的方法。这种方法 采用软组合的策略,根据像素之间反射属性相似度聚集BRDF数据,并渐进估 计反射模型参数。算法中还引入了基于形状连续性的、与漫反射率无关软组 合,有助于对镜面反射特性做出更准确的估计。 第三,本文提出了一种综合明暗、阴影、高光三方面信息,对有纹理的场 景估计方向性光源的方法。与假设已知或统一反射属性的以往方法不同,本文 利用了三条线索问的物理一致性和线索本身的完整性,处理纹理为光照估计带 来的不利影响。文中还设计了与纹理无关的光强估计技术。 对合成场景和真实场景的实验证明了这些方法的有效性。
Other AbstractIn this thesis, we address the problem of reflection and illumination analysis from images. The appearance of the world is determined by the interaction be- tween light and matter. Changes in illumination conditions or surface reflectance properties can lead to significant variations in the appearance of a scene. Specular reflections also contribute to the diversity of world appearance. Understanding reflection and light has been one of the longest standing research areas in physics and human perception. The analysis of reflection and illumination from images has also become an important and interesting problem in computer vision and graphics research. In this thesis, we focus on the following three subproblems: diffuse/specular separation, reflectance estimation, and illumination estimation from images. First, we propose a method based on color analysis and multibaseline stereo that simultaneously estimates the diffuse/specular separation and the true depth of specular reflections. This method overcomes the limiting assumptions of the previous approaches, such as accurate segmentation, regionally uniform reflectance or structured lighting. Secondly, we propose an iterative method that aggregates BRDF data and estimates reflectance values in a single image by soft grouping, where pixels con- tribute to one another's estimate according to their degree of reflectance similarity. Estimation of specular reflectance is further improved by albedo-independent soft grouping of pixels based on shape continuity. Thirdly, we propose a method that integrates cues from shading, shadow and specular reflections for estimating directional illumination in a textured scene. Unlike previous works which all assume known or uniform reflectance, our method can deal with the effects of textures by capitalizing on physical consistencies that exist among the lighting cues. A technique for texture-independent estimation of light source intensities is also presented. Experiments on synthetic and real scenes were carried on to validate our approaches.
Other Identifier693
Document Type学位论文
Recommended Citation
GB/T 7714
李元臻. 对场景反射和光照的估计与分析[D]. 中国科学院自动化研究所. 中国科学院研究生院,2003.
Files in This Item:
File Name/Size DocType Version Access License
硕士生学位论文-693.pdf(12646KB) 暂不开放CC BY-NC-SAApplication Full Text
Related Services
Recommend this item
Usage statistics
Export to Endnote
Google Scholar
Similar articles in Google Scholar
[李元臻]'s Articles
Baidu academic
Similar articles in Baidu academic
[李元臻]'s Articles
Bing Scholar
Similar articles in Bing Scholar
[李元臻]'s Articles
Terms of Use
No data!
Social Bookmark/Share
All comments (0)
No comment.

Items in the repository are protected by copyright, with all rights reserved, unless otherwise indicated.