Lidar For Basemaps
Lidar For Basemaps
Lidar For Basemaps
Lidar essentially bounces off of anything in its path, so the first step in making a basemap is
to apply a filter to separate bare-earth (hard ground) returns from everything else.
This is usually done by the vendor using special software, and the data is either stored in
separate files (bare-earth and non-bare-earth) or all the data is stored in a single file with a
classification value for each data point indicating what type it is.
Rule of Thumb:
I've found that its possible to make decent contours for 1:10000 orienteering maps
from lidar with an average point spacing of up to 2m, but having 1m or better is
ideal
for the additional small details that it provides. I've also made basemaps using 3-4m
sparsely sampled lidardata, but once you get into the 4m range you are at the same
resolution as USGS topo maps. Lidar at this resolution lacks the fine detail required
for orienteering maps, but contours derived from this data is more accurate than the
USGS topos, so its still more desirable in that respect.
All returns
(earth, and
non-earth)
Bare-earth
Non-earth
Vegetation Products
Vegetation Height: clearings, old farm field boundaries for wire
fences, holes in canopy for rootstocks
Understory Vegetation: set boundary conditions to select for
vegetation under canopy and make an understory density image.
clearings, powerlines
sometimes fences
streams,
although this
is better
from the
filtered bare
earth data
Water absorbs
the lidar pulses,
so wet features
appear black in
the intensity
image
Ive written my own contour program using a data analysis language called IDL. This
program writes contours directly to an OCAD readable file (unfortunately its OCAD5 - I
wrote it a very long time ago!)
Once nicety is the ability to write a single dataset with 0.5m contours that the end-user can
select to be any desired contour interval (0.5, 1, 2.5 and 5m) with proper index contours
every 5 lines, just by hiding or un-hiding the appropriate contour symbols in OCAD - all
from a single OCAD file.
Select appropriate
contour symbols to
hide/unhide. Here is a
2.5m interval example
from the same 0.5m
contour dataset in the
previous slide.
1)
Shaded relief
2)
Mathematical gradient
3)
Stream bed
Y-direction grad
Stream bed
3) Unsharp Mask. This is a high-pass filter. It removes the slowly varying terrain
relief, leaving behind a high-contrast image of the sharp features. It is nondirectional, and the contrast can be adjusted by changing the size of the smoothing
kernel and the stretch before writing the .bmp template for use in OCAD.
This technique is borrowed from film astrophotography, where a defocused negative
of an image (say a galaxy) is combined in an enlarger with a positive of the original,
thus performing a subtraction optically. Very useful for identifying small globular
Star clusters hiding under the bright glow of the galactic bulge and disk. For the
Lidar data we do it by digitally smoothing the original
Smoothed out-of-focus
bare earth image
Stream bed
3) Unsharp Mask. This is a high-pass filter. It removes the slowly varying terrain
relief, leaving behind a high-contrast image of the sharp features. It is nondirectional, and the contrast can be adjusted by changing the size of the smoothing
kernel and the stretch before writing the .bmp template for use in OCAD.
This technique is borrowed from film astrophotography, where a defocused negative
of an image (say a galaxy) is combined in an enlarger with a positive of the original,
thus performing a subtraction optically. Very useful for identifying small globular
Star clusters hiding under the bright glow of the galactic bulge and disk. For the
Lidar data we do it by digitally smoothing the original
Smoothed out-of-focus
bare earth image
Stream bed
3) Unsharp Mask. This is a high-pass filter. It removes the slowly varying terrain
relief, leaving behind a high-contrast image of the sharp features. It is nondirectional, and the contrast can be adjusted by changing the size of the smoothing
kernel and the stretch before writing the .bmp template for use in OCAD.
This technique is borrowed from film astrophotography, where a defocused negative
of an image (say a galaxy) is combined in an enlarger with a positive of the original,
thus performing a subtraction optically. Very useful for identifying small globular
Star clusters hiding under the bright glow of the galactic bulge and disk. For the
Lidar data we do it by digitally smoothing the original
Smoothed out-of-focus
bare earth image
Stream bed
3) Unsharp Mask. This is a high-pass filter. It removes the slowly varying terrain
relief, leaving behind a high-contrast image of the sharp features. It is nondirectional, and the contrast can be adjusted by changing the size of the smoothing
kernel and the stretch before writing the .bmp template for use in OCAD.
This technique is borrowed from film astrophotography, where a defocused negative
of an image (say a galaxy) is combined in an enlarger with a positive of the original,
thus performing a subtraction optically. Very useful for identifying small globular
Star clusters hiding under the bright glow of the galactic bulge and disk. For the
Lidar data we do it by digitally smoothing the original
Smoothed out-of-focus
bare earth image
Stream bed
3) Unsharp Mask. This is a high-pass filter. It removes the slowly varying terrain
relief, leaving behind a high-contrast image of the sharp features. It is nondirectional, and the contrast can be adjusted by changing the size of the smoothing
kernel and the stretch before writing the .bmp template for use in OCAD.
This technique is borrowed from film astrophotography, where a defocused negative
of an image (say a galaxy) is combined in an enlarger with a positive of the original,
thus performing a subtraction optically. Very useful for identifying small globular
Star clusters hiding under the bright glow of the galactic bulge and disk. For the
Lidar data we do it by digitally smoothing the original
Smoothed out-of-focus
bare earth image
Stream bed
3) Unsharp Mask. This is a high-pass filter. It removes the slowly varying terrain
relief, leaving behind a high-contrast image of the sharp features. It is nondirectional, and the contrast can be adjusted by changing the size of the smoothing
kernel and the stretch before writing the .bmp template for use in OCAD.
This technique is borrowed from film astrophotography, where a defocused negative
of an image (say a galaxy) is combined in an enlarger with a positive of the original,
thus performing a subtraction optically. Very useful for identifying small globular
Star clusters hiding under the bright glow of the galactic bulge and disk. For the
Lidar data we do it by digitally smoothing the original
Smoothed out-of-focus
bare earth image
Stream bed
Manmade
pond
Dry ditch or
gully
Small stream
Earth bank
or cliff - high side
Is bright, low side
Is dark
Small or indistinct
trails
Likely just a
reentrant
Large stream
Large trails
pit
In the Sprint
buildings
sometimes fences
small clearings
with rough open
uniform
canopy
singletrees
rough edging
on powerlines saplings
small clean
openings possible rootstocks
By over 2 minutes
Ive started making basemaps - mostly from Lidar - for clubs and AR groups, with
all proceeds donated directly to the US Orienteering Team. That is, the cost of
my labor is donated to the team, and you get a fieldchecker-ready basemap in
OCAD.
If you or your club might be interested in making a lidar basemap for your next
mapping project, please contact me and we can discuss the options.
Eddie Bergeron, US Orienteering Team
bergeron@stsci.edu