Digital image reasoning for tracking excavation activities

Efficient and safe construction of underground space is a key aspect in the development of infrastructure within densely populated urban environments. Construction processes are usually adjusted, based on information collected from field monitoring, to control induced ground movements such that the impact on nearby structures and utilities is minimized. An important component of field monitoring includes the development of a detailed timeline record of various construction activities especially soil removal. This paper explores the use of two image based techniques - Close-Range Photogrammetry and Image Reasoning - to perform semi-automated tracking of excavation activities. A new image reasoning algorithm, enhanced pattern detection and comparison (EPDC), is introduced to quickly identify changes in poor contrast excavation surfaces. EPDC is illustrated using laboratory and field trials. The proposed image reasoning algorithm paves the way for new uses of large numbers of digital camera and webcam images now available at many construction sites to acquire detailed construction staging information.