Automated BIM Model Generation Using Drawing Recognition and Line-Text Extraction

Recently, the number of BIM-applied construction projects has been increased with the expectation that it will turn labor-oriented industries into knowledge-intensive ones. Prior to application of BIM data, 3D modeling process should be preceded. The time and effort for modeling varies depending on the type of work, and therefore, most BIM applications has been focused on a certain work trade which shows high effectiveness. As a result, a doubt as to whether this new technology is effective in the construction industry has emerged. In this paper, we propose a quasi-automated BIM modeling methodology that can shorten the requisite time for modeling regardless of the type of work. This approach would resolve biased utilizations of BIM as well as suspicion issues regarding its efficacy. BIM is semantically-based and object-oriented, that is, individual element has its own geometrical shape with specific attribute as architectural element. This distinctive feature enables the proposed method. Based on drawing recognition and following line-text extraction, an object-oriented model is automatically generated. An effort demanded for modeling process can be reduced dramatically by automated modeling process. This procedure is demonstrated specifically with the practical experiment as well as detailed dataflow algorithm.