一种BIM家居场景模型纹理真实性还原方法  

BIM Home Scene Model Texture Authenticity Reduction Method

作　　者：杨亚龙[1,2,3] 胡奇志 苏亮亮 胡文瀚 Yang Yalong;Hu Qizhi;Su Liangliang;Hu Wenhan(Anhui Province Key Laboratory of Intelligent Building and Building Energy Saving,Anhui Jianzhu University,Hefei 230022,China;Anhui Institute of Strategic Study on Carbon Dioxide Emissions Peak and Carbon Neutrality in Urban-Rural Development,Hefei 230022,China;School of Electronic and Information Engineering,Anhui Jianzhu University,Hefei 230601,China)

机构地区：[1]安徽建筑大学智能建筑与建筑节能安徽省重点实验室,合肥230022 [2]安徽省建设领域碳达峰碳中和战略研究院,合肥230022 [3]安徽建筑大学电子与信息工程学院,合肥230601

出　　处：《智能建筑电气技术》2024年第1期1-5,共5页Electrical Technology of Intelligent Buildings

基　　金：安徽省重点研究与开发计划项目(202104a07020017);安徽省新时代育人质量工程项目(2022cxcysj144);国家自然科学基金(62001004);安徽建筑大学科研项目(2020XMK04)。

摘　　要：针对现有的BIM建模软件与建模方法制作出的模型纹理材质过于单一且真实感不足,不能很好地满足不同用户的实际需求等问题,本文提出了一种BIM家居场景模型纹理真实性还原方法。首先,根据家居场景中常见家具种类建立图像数据集;其次,采用神经网络算法获取到图像的纹理特征信息,得到相似的真实家具图像,并通过最小框图裁剪并合成对应的BIM家居场景中不同家具模型的材质贴图;最后反馈到建模软件实现BIM家居场景模型纹理真实性的还原。实验结果表明,经本文方法处理后的BIM家居场景模型能够较为真实的还原现实场景。For the existing BIM modeling software and methods often produce models with limited and insufficiently realistic textures,which fail to meet the actual needs of different users.This study proposes a method for restoring the authenticity of texture in BIM home scenario models.Firstly,an image dataset is established based on common types of furniture in home scenarios.Then,a neural network algorithm is employed to extract texture features from images,generating similar real furniture images.The corresponding texture maps of different furniture models in the BIM home scenario are obtained by cropping and synthesizing the minimum bounding box.Finally,the restored authenticity of texture in the BIM home scenario model is achieved by providing feedback to the modeling software.Experimental results demonstrate that the BIM home scenario models processed using this method can more realistically recreate realistic scenes.

关 键 词：BIM模型 纹理还原 神经网络 家居场景 

分 类 号：TU205[建筑科学—建筑设计及理论]