Hi,
I am currently using the following processing pipeline as building block to prepare point clouds for further processing and would love some feedback. My goal is to make the batch file as universal as possible to process various types of point clouds (varying density, size, classified, unclassified etc.) and only modify a few parameters (Ex. tile size, EPSG etc.).
What sort of tile size and buffer size would you recommend to ensure fast, efficient processing and avoiding edge artifacts?
SET curdir=%~dp0
SET cores=%NUMBER_OF_PROCESSORS%
SET epsg=2950
SET vertical_epsg=5713
:: Index, optimize
lasindex -i -i %curdir%*.laz -dont_reindex -cpu64 -cores %cores%
lasoptimize -i %curdir%*.laz -odix _o -epsg %epsg% -rescale 0.01 0.01 0.01 -vertical_epsg %vertical_epsg% -olaz -cpu64 -cores %cores%
:: Index, tile
lasindex -i -i %curdir%*_o.laz -dont_reindex -cpu64 -cores %cores%
lastile -i -i %curdir%*_o.laz -tile_size 100 -buffer 20 -odir %curdir%tuiles -odix tile -olaz -cpu64 -cores %cores%
:: Dups, noise, control
lasduplicate -i %curdir%/*_ot.laz -odix d -olaz -cpu64 -cores %cores%
lasnoise -i %curdir%/*_otd.laz -buffered 100 -step_xy 3 -step_z 0.5 -isolated 10 -odix n -olaz -cpu64 -cores %cores%