Progress monitoring, a critical tool for successful construction projects, provides information on the present state of construction, which can then be compared with the original plan. The comparison can be used in decision-making to control variations in project performance. However, current methods for acquiring and updating information on the progress of a project using digital cameras and laser-based systems are inadequate for monitoring project progress effectively due to their high computational processing, cost, and time requirements. A recently developed stereo vision system has the potential to provide information about construction project progress more efficiently, rapidly, and cheaply. In this paper, the framework for a rapid 3D object recognition technique and some preliminary experimental results are described. The results of the experiment show possibilities for applying the technique to automatic project progress monitoring and control systems.