Bringing Information to the Field: Automated Photo Registration and 4D BIM

Abstract

Effective data and information management is one of the key factors to successful construction project management. However, in current practice, the data and information are hard to search, retrieve, and reuse by the project participants. As a representative example, a large number of as-constructed photos are taken during the construction phase for documentary purposes. They have great value in progress tracking, scheduling, and decision making when their contents and context are grasped and brought to field crews. Traditionally, the contents and context of the photos are manually grasped—a practice that is inefficient and is prone to errors—and results are not shared among project participants in time. A promising alternative is to register the photos to four-dimensional (4D) building information modeling (BIM) (three-dimensional BIM plus time) to identify BIM objects in the photo and retrieve their contents in digital form. This paper presents a method for automated registration of daily photos to 4D BIM and identification of BIM objects. Specifically, a content-based image retrieval approach was devised to find the closest photo-taken location by comparing photos with 4D BIM, and then a two-dimensional grid concept was used to extract BIM objects from the model that is associated with the photos. The method is demonstrated through the implementation of photos of a real construction project, and results illustrate that the method can find the closest photo-taken location and retrieve the contents and context of the given photos.