Field monitoring is usually employed to follow the construction of underground space in densely populated areas. Such monitoring is intended to verify that settlements and lateral deformations around the excavation lie within preestablished parameters. It also allows the identification of impact on nearby structures and utilities, so that engineers can minimize such impact by adjusting the construction processes or excavation rate. In order to facilitate the interpretation of field monitoring it is imperative that a detailed timeline record of different construction activities be developed. Application of image-based techniques to perform a semi-automated tracking of excavation activities is explored in this paper. Special attention is paid to the development of the new Enhanced Pattern Detection and Comparison (EPDC) algorithm. The proposed algorithm facilitates the use of digital cameras and webcam images to highlight locations where different construction activities take place, thus allowing project engineers to create an accurate timeline record of the excavation activities. Copyright ASCE 2008.