IFCCompressor: A content-based compression algorithm for optimizing Industry Foundation Classes files

Abstract

As the commonly used open and neutral file format for Building Information Modeling (BIM) data, IFC (Industry Foundation Classes) aims to facilitate interoperability between various software platforms in the AEC industry. However, the IFC files generated from different systems often contain enormous redundant information, which will greatly limit IFC-based data storage and exchange, management, transmission and other applications. To address this issue, this paper presents a content-based compression algorithm, named IFCCompressor, for optimizing IFC data files. Its goal is to make a large IFC file as small as possible by reducing its redundant information. The algorithm is achieved through an iterative compression procedure based on an IFC model's tree structure. The optimization procedure can be lossless or be constrained by an error bound. Compared with pure compression (ZIP) regardless of information content, the presented algorithm starts with a comprehensive analysis of structure and content of IFC file, and then eliminates its redundant information without changing the original file format. Unlike partial model extraction methods, our algorithm results in a complete IFC model but with a more compact IFC physical file. In contrast with some commercial IFC optimization tools such as Solibri IFC Optimizer, the algorithm can make the size of IFC files smaller. The experimental results show that the algorithm is particularly effective for some office/residential building models with a large number of duplicated components. The compression rate with our algorithm is generally very high (the average is 40.32%) for tested cases. The online IFCCompressor tool and its demonstration can be accessed at: http://cgcad.thss.tsinghua.edu.cn/liuyushen/IFCCompressor/.