Study on Visual Appearance Measurement and Realistic Rendering for Non-metallic Materials

碩士 === 國立臺灣科技大學 === 色彩與照明科技研究所 === 107 === How to accurately describe the visual experience of materials has always been an issue that researchers are trying to persue. In order to solve this problem, CIE (Commission Internationale de’Eclairage) gave a clear definition for visual appearance in 2008....

Full description

Bibliographic Details
Main Authors: Yu-Cheng Hsieh, 謝侑城
Other Authors: Tzunghan Lin
Format: Others
Language:zh-TW
Published: 2019
Online Access:http://ndltd.ncl.edu.tw/handle/t64px7
id ndltd-TW-107NTUS5123003
record_format oai_dc
spelling ndltd-TW-107NTUS51230032019-10-24T05:20:28Z http://ndltd.ncl.edu.tw/handle/t64px7 Study on Visual Appearance Measurement and Realistic Rendering for Non-metallic Materials 非金屬材質視覺外貌量測及擬真渲染研究 Yu-Cheng Hsieh 謝侑城 碩士 國立臺灣科技大學 色彩與照明科技研究所 107 How to accurately describe the visual experience of materials has always been an issue that researchers are trying to persue. In order to solve this problem, CIE (Commission Internationale de’Eclairage) gave a clear definition for visual appearance in 2008. Visual appearance consists of the properties of color, texture, gloss, and transparency that are evaluated by quantitative analysis. According to these analysis, the rendering process of PBR (Physically Based Rendering) is introduced in this research. Those basic quantity maps, called base color, normal, roughness and metallic maps, are used in Metallic/Roughness mainstream workflows of PBR. In this research, the data with different volumetric SVBRDF (Spatially Varying Bidirectional Reflectance Distribution Function) was captured by a self-developed SVBRDF measurement system. The purpose of Experiment 1 is verifying the reality of PBR texture of our device by psychophysical experiments. There are up to 20 samples, including textiles and wall-papers, that are verified. The result of Experiment 1 is considered as the base of Experiment 2. In Experiment 2, we chose the four qualified samples to evaluate and attend to decrease the number of LED light bulbs in our system, which is important cues for roughness map estimation in SVBRDF measurement. Totally, six types of arrangement, including these cases of 33, 30, 18, 15, 9, and 3 LED light bulbs, are evaluated. The roughness maps under different number of lights were tested based on Experiment 1 by psychophysical experiment, as well. The results of Experiment 2 show that in case of 3 LED light bulbs is acceptable for the material with higher roughness. And, in case of 9 LED light bulbs is not enough for the material of lower roughness. By contrast, in case of 15 LED light bulb is just acceptable in this situation. Tzunghan Lin 林宗翰 2019 學位論文 ; thesis 106 zh-TW
collection NDLTD
language zh-TW
format Others
sources NDLTD
description 碩士 === 國立臺灣科技大學 === 色彩與照明科技研究所 === 107 === How to accurately describe the visual experience of materials has always been an issue that researchers are trying to persue. In order to solve this problem, CIE (Commission Internationale de’Eclairage) gave a clear definition for visual appearance in 2008. Visual appearance consists of the properties of color, texture, gloss, and transparency that are evaluated by quantitative analysis. According to these analysis, the rendering process of PBR (Physically Based Rendering) is introduced in this research. Those basic quantity maps, called base color, normal, roughness and metallic maps, are used in Metallic/Roughness mainstream workflows of PBR. In this research, the data with different volumetric SVBRDF (Spatially Varying Bidirectional Reflectance Distribution Function) was captured by a self-developed SVBRDF measurement system. The purpose of Experiment 1 is verifying the reality of PBR texture of our device by psychophysical experiments. There are up to 20 samples, including textiles and wall-papers, that are verified. The result of Experiment 1 is considered as the base of Experiment 2. In Experiment 2, we chose the four qualified samples to evaluate and attend to decrease the number of LED light bulbs in our system, which is important cues for roughness map estimation in SVBRDF measurement. Totally, six types of arrangement, including these cases of 33, 30, 18, 15, 9, and 3 LED light bulbs, are evaluated. The roughness maps under different number of lights were tested based on Experiment 1 by psychophysical experiment, as well. The results of Experiment 2 show that in case of 3 LED light bulbs is acceptable for the material with higher roughness. And, in case of 9 LED light bulbs is not enough for the material of lower roughness. By contrast, in case of 15 LED light bulb is just acceptable in this situation.
author2 Tzunghan Lin
author_facet Tzunghan Lin
Yu-Cheng Hsieh
謝侑城
author Yu-Cheng Hsieh
謝侑城
spellingShingle Yu-Cheng Hsieh
謝侑城
Study on Visual Appearance Measurement and Realistic Rendering for Non-metallic Materials
author_sort Yu-Cheng Hsieh
title Study on Visual Appearance Measurement and Realistic Rendering for Non-metallic Materials
title_short Study on Visual Appearance Measurement and Realistic Rendering for Non-metallic Materials
title_full Study on Visual Appearance Measurement and Realistic Rendering for Non-metallic Materials
title_fullStr Study on Visual Appearance Measurement and Realistic Rendering for Non-metallic Materials
title_full_unstemmed Study on Visual Appearance Measurement and Realistic Rendering for Non-metallic Materials
title_sort study on visual appearance measurement and realistic rendering for non-metallic materials
publishDate 2019
url http://ndltd.ncl.edu.tw/handle/t64px7
work_keys_str_mv AT yuchenghsieh studyonvisualappearancemeasurementandrealisticrenderingfornonmetallicmaterials
AT xièyòuchéng studyonvisualappearancemeasurementandrealisticrenderingfornonmetallicmaterials
AT yuchenghsieh fēijīnshǔcáizhìshìjuéwàimàoliàngcèjínǐzhēnxuànrǎnyánjiū
AT xièyòuchéng fēijīnshǔcáizhìshìjuéwàimàoliàngcèjínǐzhēnxuànrǎnyánjiū
_version_ 1719277011907641344