Structural health monitoring techniques (SHM) have become a useful means to document in-service load tests or collect long-term data from ambient traffic on bridges or other civil structures [1-3]. Most of the used SHM data acquisition systems consist of physical sensor networks that are attached to the structure’s surface, and transmit collected data either wired [4] or wirelessly [5] to a hub. From there the data is downloaded to a laptop or transmitted via Internet connection to the bridge engineer. The sensors record data from external stimuli such as temperature, humidity, or load, or internal structural responses such as strain or displacement. In order to save deployment time and costs, remote sensing approaches have more recently been studied for SHM applications such as laser vibrometers [6, 7], LIDAR [8], GPS [9], or image-based methods [7, 10]. These techniques are promising for global monitoring, i.e. modal analysis, but often lack the desired resolution for accurate dynamic response characterization or effective local damage detection [11, 12]. One reason is that typically only a small finite number of points can be monitored simultaneously which leads to sparse data. In addition, these techniques are still expensive and require specialized equipment that needs to be operated by trained technicians. Some researchers have used videos to detect vehicle location and correlate that with structural response measured by traditional sensor networks [13]. This project, however, proposes a different approach. Motivated by the recent wide availability of inexpensive high-quality high-speed digital video cameras combined with innovative video signal processing algorithms, it is time to consider the next generation of monitoring techniques that uses the captured digital video to extract information of structural dynamic performance directly.