A new methodology to estimate the discrete-return point density on airborne lidar surveys

Abstract

The distribution of the discrete-return point density in airborne lidar flights obtained from an oscillating mirror laser scanner is analysed and alternative formulations to determine its value are presented. The point density in a lidar swath varies and can best be fitted with a potential function. This study confirms that calculating the overall point density with traditional statistical parameters yields biased results owing to the abnormally high densities of the swath boundaries. New formulas for calculating the representative mean are proposed: a weighted arithmetic mean (WAM) based on a potential function; geometric mean (GM); and harmonic mean (HM). All three means give more weight to the central sectors across the strip and less to the boundary sectors where extreme data redundancy exists. The WAM based on a potential function yields equivalent estimates as the HM; the GM yields slightly higher estimates. The results obtained improve the mean estimation and, more importantly, allow users to estimate better the mean point density on airborne lidar surveys, which are usually overestimated approximately by 15%.