Rule-based Learning Techniques to Derive Automated Digital Terrain Model Using Airborne LiDAR Data

Abstract

Constructing an accurate Digital Terrain Model is costly and time-consuming, leading to more challenges in urban environments due to the presence of different objects. This research performs the step by step analysis of LiDAR data using a rule-based algorithm to create an automatic DTM. This method needs no extra data and has a precision equal to that of a DTM, which is constructed manually. The DTM constructed in this research was compared to the DTM constructed manually to investigate the accuracy of the results. It was found that the mean difference between the elevations in both DTMs in the rural and urban areas was equal to zero and 0.10 m, respectively, while the mean difference between the slopes was 1.2 and 1.6%, respectively. However, in the areas which lacked buildings, the elevation and slope characteristics were equal, revealing identical DTMs, which was also confirmed by sig=.441 from t-test. Although sig=0.0 in the t-test shows a difference between the two DTMs in the urban and rural areas, it does not reveal the value of this difference. Thus, the RMSE method was used to examine this difference, leading to the values of ±0.20m, ±0.05m, and ±0.04m for the urban, rural, and areas without buildings, respectively. Considering that the precision required for urban and rural planning is 0.4m, it is totally acceptable to use the proposed algorithm instead of the manual method.