Some highly populated coastal areas suffer from the growing rate of extreme weather conditions. To monitor beach evolution, “shoreline position” has been adopted as an indicator. Unmanned aerial vehicle (UAV) with an integrated optical camera offers high-resolution remote-sensing data at low cost with high flexibility. In this study, video imagery from a UAV hovering above shoreline was analyzed. An image registration process was developed to deal with camera motion caused by wind and poor positioning. As the result, our image registration accuracy reached 0.56 pixel. We proposed two approaches for shoreline detection using: 1. cross-shore section image sequence, based on gray-scale intensity difference between wet sand and water; and 2. mean image, based on “color channel divergence” with automated threshold procedure. Shoreline positioning from cross-shore section image sequence agreed well with visually identified instantaneous shoreline. Shoreline positioning from mean image were highly correlated with shoreline positioning from cross-shore section image sequence, within an error of 1.53 m. In conclusion, using UAV to detect shoreline position is proved to be practical. With the mobility of UAV, it is suitable for short-term coastal change monitoring.