Identifying the local properties of a structure, either to perform structural health monitoring or to fine tune a numerical model, requires the updating of a large number of parameters. With a high spatial density, but a low dynamic range response information, high-speed-camera measurements have the potential to identify a large number of localized parameters. In contrast, accelerometer measurements provide low-spatial-density modal shapes, but a high dynamic range, and introduce the problem of mass loading. In this research, modal shapes from a high-speed camera are used, providing full-field response information about the observed structure and an over-determined optimization problem. Since the high-speed camera has a lower dynamic range than the accelerometer and the signal-to-noise ratio is low where the displacement amplitude is small, location-specific weighting methods were introduced. The numerical and real experiments showed that the accelerometer’s positioning is important for successful updating, while with a high-speed-camera measurement this is not relevant. This research showed that due to the spatial over-determination, the model updating based on high-speed-camera data, was significantly better than the low-spatial-resolution, accelerometer-based approach.