Accuracy Control of ALOS DSM

The possibility to create DSM from stereo pairs is based on the Pythagoras theorem and on the principles of photogrammetry that are applied to aerial photographs stereo pairs for the last seventy years. The application of these principles to digital satellite stereo data was inherent in the first satellite missions. During the last decades the satellite stereo-pairs were acquired across the track in different days (SPOT, ERS etc.). More recently the same-date along the track stereo-data acquisition seems to prevail (Terra ASTER, SPOT5 HRS) as it reduces the radiometric image variations (refractive effects, sun illumination, temporal changes) and thus increases the correlation success rate in any image matching.

One of the newest satellite sensors with stereo collection capability is ALOS PRISM. It can acquire stereopairs along the track with a 2,5m spatial resolution covering areas of 35X35km.

In this study we control the accuracy of a DSM created from ALOS data. The study area is Antiparos Island, Greece. The DSM was compared to a DSM created from airphotos stereopairs and to ground control points collected with a Differential GPS.

After a first control for random or systematic errors a statistical analysis was done. Points of certified elevation have been used to estimate the accuracy of the DSM. The elevation difference between the two DSMs was calculated. 2D RMSE, correlation and the percentile value were also computed and the results are presented.