抄録
Digital canopy models (DCMs) were extracted by multi-temporal aerial photographs (1964∼2004) and digital photogrammetric technique for forest canopy monitoring. In this study, ground contorol points (GCPs) were extracted by IKONOS image and digital elevation model (DEM) from LiDAR data. Bundle block adjustment with self-calibration was used for aerial triangulation, because camera callibration reports were not available. The generated DCM which was obtained in 2004 was compared with DCM from LiDAR data (obtained in 2004). Most of all pixel by aerial photographs were corresponded with DCM from LiDAR data. However, small gap areas were difficult to detect. Excluding gap areas, root mean square error of DCM from aerial photo was 3.78 meter. Using extracted multi-temporal DCMs, forest canopy changes, such as tree growth, clear-cutting and fallen damage, were figured out. Multi-temporal DCMs were found to be effective in monitoring forest canopy.
