Vehicle inspection at ports of entry is a critical component of border security. One part of the vehicle screening process involves customs and border protection (CBP) personnel performing a preliminary inspection of the underside of random vehicles by looking under the vehicle through a mirror mounted on a stick, searching for anomalies or foreign objects present on the undercarriage structure and components of a vehicle. If any deviances are detected, the vehicle is directed to a secondary and more thorough inspection. This paper presents a project that aims at automating this preliminary undercarriage inspection by using an automatic under-vehicle inspection system (AUVIS) and image processing algorithms to assist personnel in identifying vehicles for secondary inspection. The inspection system will image the undercarriage of every car in a lane as they approach the CBP agent and use a novel change detection algorithm to compare the captured image to a corresponding reference image of the vehicle's make and model. The software will detect and highlight any differences between the two images to provide a rapid and objective recommendation for secondary inspection.