Deep Belief Network Based 3D Models Classification in Building Information Modeling

Li Wang, Zhi-kai Zhao, Na Xu

Abstract


3D models classification is a critical process of Building Information Modeling (BIM). A Deep Learning Approach is proposed to classify 3D models in BIM environment. The ray based feature extraction algorithm is used to extract features of 3D models and form features matrix. The Deep Belief Network constructed by Restricted Boltzmann Machines applies the features matrix and classifies the models adopting the effective training process. The process of training DBN is layer by layer. Experiments were taken on the public 3D model library of PSB model database. The results show that compared with several commonly used classification method, the proposed method of this paper has achieved good results in the 3D model classification for efficiently BIM.

Keywords


Deep learning, BIM, 3D models classification, Deep Belief Network, Feature extraction

Full Text:

PDF



International Journal of Online Engineering (iJOE).ISSN: 1861-2121
Creative Commons License
Indexing:
Web of Science ESCI logo Engineering Information logo INSPEC logo DBLP logo ELSEVIER Scopus logo EBSCO logo Ulrich's logoGoogle Scholar logo Microsoft® Academic Search