Rogue point elimination in airborne laser hydrography

Abstract

Elimination of rogue points from data gathered by an airborne laser depth sounding system is a vital part of the processing of such data for hydrographic purposes. Application of the probabilistic data association filter is made to achieve this function with extensions to 2-D operation and the inclusion of smoothing. The algorithms are shown to perform well on data gathered by the airborne depth sounder (WRELADS II). They are successful both in eliminating false detections of the sea bottom and in retaining a single true shallower depth surrounded by deeper water.