Methods for estimating relative accuracy of ALS data.

Abstract

Listen

Translate article

The data acquired by airborne laser scanning is a set of measurement points with certain XYZ coordinates (the so-called cloud of points), located at places where laser pulses sent by the scanning system are refl ected. The coordinates of these points are determined based on the distance measured from the scanner system to the ground along the predetermined laser beam, and the scanning angle at which the beam is sent. The position and orientation of the scanner during the measurement is determined with the use of the GPS/INS system integrated with the scanner [7]. The coverage of the developed area with data from the ALS is done by executing a fl ight in multiple parallel strips that partly overlap, thus ensuring the continuity of the development. The accuracy of the designated coordinates of ALS points depends on the accuracy of observation (distance, scan angle, position and orientation) and on the mathematical model used to determine the coordinates of these points [2]. In connection with the occurrence of errors in the measurement devices and in the defi nition of the mathematical model, there exist some discrepancy in the data from the adjacent strips in overlapping area. Therefore, in order to improve the position of the measuring points, we use ALS data alignment methods by basing on the adjustment of data from adjacent strips and their best integration in the exterior coordinate system. The use of ALS data for the purposes of subsequent developments, such as the DTM (Digital Terrain Model) and DSM (Digital Surface Model) requires the determination of their accuracy. By measuring the diff erences between the adjacent strips, we can estimate the relative accuracy of the ALS, and by sett ing the position of the examined control objects in the exterior coordinate system independently, we can also estimate their absolute accuracy. However, the specifi c nature of data