Semi-automatic true orthophoto production by using LIDAR data

Light Detection and Ranging (LIDAR) has been used for years with a variety of applications, including the efficient creation of digital terrain models (DTMs) for large-scale, high- accuracy mapping. LIDAR technology offers fast, real-time collection of 3-D points so that allows the production of higher-accuracy orthophotos than traditional stereo-compilation methods. The technology is moving quickly toward offering more efficient collection techniques for applications such as city modelling and true orthophoto production. Traditional orthophoto production suffers from the fact that buildings (and any other objects above ground) are not correctly placed in the orthophoto. True orthophoto production overcomes these deficiencies by taking the digital surface model (DSM) into account. However, the resulting true orthophotos still suffer from occluded areas and unsharp edges of buildings and roads. The experimental investigations with different data sets show that it is essential to use high quality DSMs to produce high quality true orthophotos. In this paper, aspects of the improvement of true orthophoto production will be discussed: (a) in the first approach which has been widely used in photogrammetry, the DSM is used for generating true orthophotos and (b) in comparison to the DSM based method, semi-automatically collected vector data of buildings are superimposed on DTM and are used as input data for generating true orthophotos in the second approach. The processes and problems of both true orthophoto production procedures will be investigated in detail and results include the comparison of true orthophoto productions based on DSM (from LIDAR data) and DTM plus vector data. For the experimental investigation, a LIDAR data set for downtown Stuttgart is used.