Development of an algorithm to generate a Lidar pit-free canopy height model
Lidar-derived Canopy Height Models (CHMs) are commonly used for extractmg relevant forest information.Often irregular height variations - allso called data pits or simply pits - are present in the CHM.Tbese pits typically appear when the frrst Lidar retum is far below the canopy which tends to happen for two reasons.The first reason is that a laser beam deeply penetrates through the branches and the foliage before producing the frrst retum (Persson et al.,2002).The second reason is multiple laser beams - possibly from different flight lines - produce their frrst retum in close horizontal proximity but with a great height difference because they see the canopy or the groLmd from different angles (Leckie et al.,2003).These pits hamper the correct extraction of forestry metrics from the CHM.Previous studies recommend to apply smoothing such as a median filter or a Gaussian filter to reduce pits.However,smoothing modifies all height values of the CHM not just those corresponding to pits and a filter may smooth away small trees when its support is on the order of the crown diameter (Solberg et al.,2006).We now describe the implementation of a new algorithm for generating pit-free CHM rasters from different density Lidar point clouds and visually compare them to those obtained with standard Gaussian smoothing.
Anahita Khosravipour Andrew K.Skidmore Martin Isenburg Tiejun Wang Yousif A.Hussin
ITC, University of Twente, The Netherlands rapidlasso GmbH, Germany
国际会议
北京
英文
125-128
2013-10-09（万方平台首次上网日期，不代表论文的发表时间）