Damage detection and localization allow for automated real-time monitoring of engineered systems. The benefits of such a system include improved safety, lower maintenance costs, and higher reliability. Many of the early works focus almost exclusively on numerical simulations of real systems, with very little experimentally acquired data used in detection. Introducing real world data complicates the analysis significantly by requiring noise reducing techniques to acquire legitimate results. Further, the cost of obtaining enough data to fully define a damaged system can quickly become prohibitive. This paper focuses on damage detection schemes carried out through empirical means. First a concept proving scheme is used by which data about the system is collected through accelerometer data. The damage detection scheme requires the reduction of a large set of data to one or two descriptive eigen parameters. Second, the scheme is repeated using optically gathered data through use of a high speed camera and software image manipulation tools. Damage detection is shown to be possible under the same conditions and initial parameters. Localization of the damage, however, is shown to require sensor information from multiple locations. Further still the optically based method is shown to supplement a failed detection by other means.