How Single Photon LiDAR (SPL) and Geiger Mode LiDAR (GML) compare to "traditional" linear LiDAR
516 views
Skip to first unread message
Martin Isenburg
Feb 13, 2018, 9:16:17 PM2/13/18
to LAStools - efficient command line tools for LIDAR processing
Hello,
After posting the summary and link
below I got a few questions such as:
(1) Are there comparative data sets available showing the results obtained using the different technologies for the same physical area?
and
(2) Are there links available which can be shared for results which are more technical, quantitative and rigorous?
The ultimate "Photon Shoot Out" would be to have several currently available scanning systems fly the same test area with settings both for their respective "advertised" maximal commercial production capability as well with settings for their respective "highest density and accuracy"
capability ... (-:
====================
So what's the deal with Single Photon LiDAR (SPL) and Geiger Mode LiDAR (GML) now that our industry and various government agencies have had a year or two to evaluate whether they measure up to the initial hype by the PR departments of Sigma Space Corporation who produced the SPL and Harris Geospatial Solutions who operate the GML. According to these two industry veterans, only the SPL may be able to deliver the accuracy and the richness of vegetation-penetrating linear LiDAR systems, whereas the GML is more like the point clouds derived from dense-matching photogrammetry ...
Regards,
Martin @rapidlasso
Jason Stoker
Feb 14, 2018, 8:57:27 AM2/14/18
to LAStools - efficient tools for LiDAR processing
It is a bit dated now, but we did compare the three technologies over the same area in CT a couple of years ago: http://www.mdpi.com/2072-4292/8/9/767
I would offer to share this data, but the vendors collected it gratis for us (so technically they own it) and the issues we reported have been fixed in later iterations of the sensors, so I’m guessing they do not want people revisiting old issues they have corrected.
We (USGS) have contracted for GML in South Dakota, Illinois and North Carolina. We have contracted for SPL in another part of South Dakota and Hawaii. All data are in various stages of collection and acceptance. Once data passes our Q/A it will be made available for anyone to use and test.
I would offer to share this data, but the vendors collected it gratis for us (so technically they own it) and the issues we reported have been fixed in later iterations of the sensors, so I’m guessing they do not want people revisiting old issues they have corrected.
We (USGS) have contracted for GML in South Dakota, Illinois and North Carolina. We have contracted for SPL in another part of South Dakota and Hawaii. All data are in various stages of collection and acceptance. Once data passes our Q/A it will be made available for anyone to use and test.
Olsen, Richard (RC) (CIV)
Feb 14, 2018, 6:18:21 PM2/14/18
to last...@googlegroups.com
At ILMF 2018, there was a fairly well done, quantitative comparison of the systems done in the context of forested areas in Florida.
by Alvan Karlin; Southwest Florida Water Management District; email ?
Not published, unfortunately, but I have in my notes that the briefing is ‘available’.
Riegl was more accurate than the Harris system by about a factor of 10 according to my notes (0.01 inch mean error, vs 0.11 inch mean error); Harris was faster (and apparently flew at night); final products he was interested in seemed to be equivalent in quality (?). the overall time advantage, or area collection rate, favored the Harris system by a factor of 2 or so.
( I was very curious as to how flying at night might have made the solar noise problem more manageable for the Harris system. )
Per Martin’s comment below about a need for a shootout – I’ll comment that it really is going to be very application specific.
RC Olsen
Naval Postgraduate School
--
Download LAStools at
http://lastools.org
http://rapidlasso.com
Be social with LAStools at
http://facebook.com/LAStools
http://twitter.com/LAStools
http://linkedin.com/groups/LAStools-4408378
Manage your settings at
http://groups.google.com/group/lastools/subscribe
Stoker, Jason
Jul 16, 2019, 2:05:27 AM7/16/19
to <lastools@googlegroups.com>
Hi everyone- thought some of you here might be interested in digging in to some Geiger mode lidar data collected for the state of North Carolina in collaboration with 3DEP we recently published.
We look forward to any feedback you want to provide on these data, especially as it relates/compares to traditional linear-mode data. I found Martin's dive into the SPL100 (or 99, ha) data we published in South Dakota very helpful! We are still working on improving our specifications to better accommodate these data, so any feedback/thoughts you can provide back to us after digging in would be great.
Points have been classified according to NC requirements, including an automated building, road and vegetation classification.
There are currently 12 of 21 counties that have made it through our 3DEP Q/A and are published- available for free, open, public use on our FTP site. About 3.3 TB of data spanning 21,840 files. Data are all in LAZ format.
Data (LPC and DEMs can also be downloaded from the National Map viewer. GML data will have the "NC_Phase4" prefix (https://viewer.nationalmap.gov/basic) These will also be available via our Amazon S3 requester pays bucket (I think 7 counties are up there now), and we hope to push point clouds to EPT in the Amazon public dataset bucket soon too.For those not as interested in a specific AOI, you can access county-level projects directly via FTP at these locations.
Derived DEMs are available as well here ftp://rockyftp.cr.usgs.gov/vdelivery/Datasets/Staged/Elevation/OPR/, using the same "NC_Phase4_[County]_[Year]" naming convention.
Screenshot of current data extent below.
Cheers,
jason
jason
Jason M. Stoker, Ph.D
US Geological Survey
National Geospatial Program
Office: 970-226-9227
Cell: 605-496-3513
My USGS Profile
US Geological Survey
National Geospatial Program
Office: 970-226-9227
Cell: 605-496-3513
My USGS Profile
Gottfried Mandlburger
Jul 23, 2019, 5:05:01 AM7/23/19
to last...@googlegroups.com
Dear Jason,
Thank you for the link to this very impressive data collection. Following the discussion on subject matters (i.e. previous posts with the same subject), I'm wondering if the unfiltered data containing the noise points typical for single photon sensitive LiDAR is (or can be made) available for any of the cited data. I'm sure the community in general and the LAStools group in particular would benefit, if more unfiltered GmLiDAR data was available as is the case for the SPL100 data (or 99 respectively) of Navarra.
Do you see any chance for this to happen?
Kind regards,
Gottfried
PS:
Thank you for the link to this very impressive data collection. Following the discussion on subject matters (i.e. previous posts with the same subject), I'm wondering if the unfiltered data containing the noise points typical for single photon sensitive LiDAR is (or can be made) available for any of the cited data. I'm sure the community in general and the LAStools group in particular would benefit, if more unfiltered GmLiDAR data was available as is the case for the SPL100 data (or 99 respectively) of Navarra.
Do you see any chance for this to happen?
Kind regards,
Gottfried
PS:
On 12.09.2018 11:23, Paski wrote:
Finally we have in our FTP site (ftp://ftp.cartografia.navarra.es/) for free download all LiDAR data from Single Photon Lidar (ftp://ftp.cartografia.navarra.es/5_LIDAR/5_4_2017_NAV_ca_EPSG25830/).
---
You received this message because you are subscribed to the Google Groups "LAStools - efficient tools for LiDAR processing" group.
To unsubscribe from this group and stop receiving emails from it, send an email to lastools+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/lastools/CANNEDyioAmRONgyLhVfGW6892dtC6kVKykM4Gk%3D9x0BFSU4u_Q%40mail.gmail.com.
-- Dr. Gottfried Mandlburger Tel.: +43 1 58801 12235 Fax.: +43 1 58801 12299 http://www.ipf.tuwien.ac.at _____ _____ _____ /____// ___// / TU Wien // __ / /__ / // / Department of Geodesy and Geoinformation //__/// /__ / // / Research Group Photogrammetry /____//____//____/ Gusshausstrasse 27-29, A-1040 Vienna
Stoker, Jason
Jul 23, 2019, 10:24:08 AM7/23/19
to <lastools@googlegroups.com>
Hi Gottfried, this is a great question. Unfortunately for this Geiger-mode collection neither 3DEP nor the state of North Carolina requested a copy of the 'raw' data in contracting. In GML terms this would be called "L1 or L2 data". Desciption of levels is on page 19 here: https://gwg.nga.mil/focus_groups/csmwg/LIDAR_Formulation_Paper_Version_1.1_110801.pdf We only received the "L3" product, which is the fully processed, classified lidar point cloud. All of the filtering is done on the back end, versus in the classification processes we are used to (every return is a point and we flag/classify noise points). We did not have the software nor the storage to process all the raw data. Processing is very different compared to linear mode or even single photon lidar, and happens much earlier in the process than traditional systems. You could think of it like processing SfM data, where your 'raw' data is all the frames of imagery and your product is the derived point cloud. In this case the raw data is all the 3D photons collected frame by frame by the Harris system. I'm not sure of the exact specifications, but it could be something like 100x the storage needed for the raw data versus the final product we provide to the public. At our fixed (and sometimes decreasing) budgets, these increases in storage and processing are hard to justify to my leadership.
I've attached a paper I found summarizing how Harris GML processing workflow is done.
The collection of the L1 data by Harris allows them to fly once and process the L1 data in different ways to filter data and extract different ANPDs depending on the application. Some of the lower costs we were supposed to realize are due to the fact that Harris could produce different L3 products for different customers from the same flights- tuning the data to produce QL1 data for us, versus 20-40pts/sq m for a utility customer for example. This whole concept is different than what we are used to where our mindset has been that we 'own all the points' from a flight from a traditional sensor. I have personally wrestled with this concept, as it is so different than what I have been doing for 20 years, and we (3DEP) are still trying to design how and what we request from a GML system to test to meet our specification.
I would love to hear anyone's thoughts as to juggling the value of keeping the raw data versus the cost of storing the 'noise'. As a scientist and researcher I have one answer, but as a manager dealing with budgets sometimes I have another.
Jason M. Stoker, Ph.D
US Geological Survey
National Geospatial Program
Office: 970-226-9227
Cell: 605-496-3513
My USGS Profile
US Geological Survey
National Geospatial Program
Office: 970-226-9227
Cell: 605-496-3513
My USGS Profile
To view this discussion on the web visit https://groups.google.com/d/msgid/lastools/a3b9ffa2-aada-8eaa-9186-15f3c7ffd818%40geo.tuwien.ac.at.
Gottfried Mandlburger
Jul 24, 2019, 4:58:22 AM7/24/19
to last...@googlegroups.com
Jason,
Thank's for your open words and for sharing the Harris paper (which I
came across earlier). I especially appreciate your twofold view on the
topic as a scientist and a manager. And, data load is certainly an issue.
I'm totally aware of the different data properties of conventional
(linear-mode) LiDAR, SPL, and GmLiDAR. I'm also inclined to follow your
argument of single photon sensitive LiDAR being somehow similar to data
from Dense Image Matching, although the reason for noise is completely
different in DIM (image orientation, texture, stereo restitution, etc.)
than in SPL/GmLiDAR (dark count, solar background, aerosols, etc.).
> The collection of the L1 data by Harris allows them to fly once and
> process the L1 data in different ways to filter data and extract
> different ANPDs depending on the application.
This is exactly what people from Leica told when I was discussing the
possibility of obtaining L1/L2 data of the SPL100 sensor. Thing is, that
we have to rely on what the manufacturers think is the best way of data
processing for a certain application. I don't mistrust the (software)
engineers, neither at Harris nor at Leica/SigmaSpace. But, of course,
this is not what we would call an open data approach.
Besides filtering of noise, there is also the issue of data smoothing,
which could be done in the one or the other way. This especially applies
to GmLiDAR data with the inherent (high) overlap of consecutive footprints.
We are discussing matters here in the LASTools forum, and from the
perspective of a lidar data processing software I see a clear benefit
for having access to earlier processing stages than L3. I'm sure the
community would benefit, not least for raising awareness. However, I
totally understand that this might not be feasibly for, e.g., all the
3DEP data but. Thus, it would be great to have L1/2/3 data for a
representative flight block.
This is certainly a scientist's-only view on the subject.
Kind regards,
Gottfried
On 2019-07-23 16:14, 'Stoker, Jason' via LAStools - efficient tools for
> My USGS Profile <https://www.usgs.gov/staff-profiles/jason-stoker>
Thank's for your open words and for sharing the Harris paper (which I
came across earlier). I especially appreciate your twofold view on the
topic as a scientist and a manager. And, data load is certainly an issue.
I'm totally aware of the different data properties of conventional
(linear-mode) LiDAR, SPL, and GmLiDAR. I'm also inclined to follow your
argument of single photon sensitive LiDAR being somehow similar to data
from Dense Image Matching, although the reason for noise is completely
different in DIM (image orientation, texture, stereo restitution, etc.)
than in SPL/GmLiDAR (dark count, solar background, aerosols, etc.).
> The collection of the L1 data by Harris allows them to fly once and
> process the L1 data in different ways to filter data and extract
> different ANPDs depending on the application.
possibility of obtaining L1/L2 data of the SPL100 sensor. Thing is, that
we have to rely on what the manufacturers think is the best way of data
processing for a certain application. I don't mistrust the (software)
engineers, neither at Harris nor at Leica/SigmaSpace. But, of course,
this is not what we would call an open data approach.
Besides filtering of noise, there is also the issue of data smoothing,
which could be done in the one or the other way. This especially applies
to GmLiDAR data with the inherent (high) overlap of consecutive footprints.
We are discussing matters here in the LASTools forum, and from the
perspective of a lidar data processing software I see a clear benefit
for having access to earlier processing stages than L3. I'm sure the
community would benefit, not least for raising awareness. However, I
totally understand that this might not be feasibly for, e.g., all the
3DEP data but. Thus, it would be great to have L1/2/3 data for a
representative flight block.
This is certainly a scientist's-only view on the subject.
Kind regards,
Gottfried
On 2019-07-23 16:14, 'Stoker, Jason' via LAStools - efficient tools for
> <mailto:gottfried....@geo.tuwien.ac.at>> wrote:
>
> Dear Jason,
>
> Thank you for the link to this very impressive data collection.
> Following the discussion on subject matters (i.e. previous posts
> with the same subject), I'm wondering if the unfiltered data
> containing the noise points typical for single photon sensitive
> LiDAR is (or can be made) available for any of the cited data. I'm
> sure the community in general and the LAStools group in particular
> would benefit, if more unfiltered GmLiDAR data was available as is
> the case for the SPL100 data (or 99 respectively) of Navarra.
>
> Do you see any chance for this to happen?
>
> Kind regards,
> Gottfried
>
> PS:
> On 12.09.2018 11:23, Paski wrote:
>> Finally we have in our FTP site
>> (ftp://ftp.cartografia.navarra.es/) for free download all LiDAR
>> data from Single Photon Lidar
>> (ftp://ftp.cartografia.navarra.es/5_LIDAR/5_4_2017_NAV_ca_EPSG25830/).
>
>
>
>
> Dear Jason,
>
> Thank you for the link to this very impressive data collection.
> Following the discussion on subject matters (i.e. previous posts
> with the same subject), I'm wondering if the unfiltered data
> containing the noise points typical for single photon sensitive
> LiDAR is (or can be made) available for any of the cited data. I'm
> sure the community in general and the LAStools group in particular
> would benefit, if more unfiltered GmLiDAR data was available as is
> the case for the SPL100 data (or 99 respectively) of Navarra.
>
> Do you see any chance for this to happen?
>
> Kind regards,
> Gottfried
>
> PS:
> On 12.09.2018 11:23, Paski wrote:
>> Finally we have in our FTP site
>> (ftp://ftp.cartografia.navarra.es/) for free download all LiDAR
>> data from Single Photon Lidar
>> (ftp://ftp.cartografia.navarra.es/5_LIDAR/5_4_2017_NAV_ca_EPSG25830/).
>
>
>
ftp://rockyftp.cr.usgs.gov/vdelivery/Datasets/Staged/Elevation/LPC/Projects/USGS_LPC_NC_Phase4_Alexander_2017_LAS_2019/
>
ftp://rockyftp.cr.usgs.gov/vdelivery/Datasets/Staged/Elevation/LPC/Projects/USGS_LPC_NC_Phase4_Alleghany_2016_LAS_2019/
>
ftp://rockyftp.cr.usgs.gov/vdelivery/Datasets/Staged/Elevation/LPC/Projects/USGS_LPC_NC_Phase4_Anson_2016_LAS_2019/
>
ftp://rockyftp.cr.usgs.gov/vdelivery/Datasets/Staged/Elevation/LPC/Projects/USGS_LPC_NC_Phase4_Cabarrus_2016_LAS_2019/
>
ftp://rockyftp.cr.usgs.gov/vdelivery/Datasets/Staged/Elevation/LPC/Projects/USGS_LPC_NC_Phase4_Catawba_2017_LAS_2019/
>
ftp://rockyftp.cr.usgs.gov/vdelivery/Datasets/Staged/Elevation/LPC/Projects/USGS_LPC_NC_Phase4_Cleveland_2016_LAS_2019/
>
ftp://rockyftp.cr.usgs.gov/vdelivery/Datasets/Staged/Elevation/LPC/Projects/USGS_LPC_NC_Phase4_Davidson_2017_LAS_2019/
>
ftp://rockyftp.cr.usgs.gov/vdelivery/Datasets/Staged/Elevation/LPC/Projects/USGS_LPC_NC_Phase4_Davie_2017_LAS_2019/
>
ftp://rockyftp.cr.usgs.gov/vdelivery/Datasets/Staged/Elevation/LPC/Projects/USGS_LPC_NC_Phase4_Forsyth_2017_LAS_2019/
>
ftp://rockyftp.cr.usgs.gov/vdelivery/Datasets/Staged/Elevation/LPC/Projects/USGS_LPC_NC_Phase4_Gaston_2016_LAS_2019/
>
ftp://rockyftp.cr.usgs.gov/vdelivery/Datasets/Staged/Elevation/LPC/Projects/USGS_LPC_NC_Phase4_Iredell_2017_LAS_2019/
>
ftp://rockyftp.cr.usgs.gov/vdelivery/Datasets/Staged/Elevation/LPC/Projects/USGS_LPC_NC_Phase4_Lincoln_2016_LAS_2019/
>
ftp://rockyftp.cr.usgs.gov/vdelivery/Datasets/Staged/Elevation/LPC/Projects/USGS_LPC_NC_Phase4_Alleghany_2016_LAS_2019/
>
ftp://rockyftp.cr.usgs.gov/vdelivery/Datasets/Staged/Elevation/LPC/Projects/USGS_LPC_NC_Phase4_Anson_2016_LAS_2019/
>
ftp://rockyftp.cr.usgs.gov/vdelivery/Datasets/Staged/Elevation/LPC/Projects/USGS_LPC_NC_Phase4_Cabarrus_2016_LAS_2019/
>
ftp://rockyftp.cr.usgs.gov/vdelivery/Datasets/Staged/Elevation/LPC/Projects/USGS_LPC_NC_Phase4_Catawba_2017_LAS_2019/
>
ftp://rockyftp.cr.usgs.gov/vdelivery/Datasets/Staged/Elevation/LPC/Projects/USGS_LPC_NC_Phase4_Cleveland_2016_LAS_2019/
>
ftp://rockyftp.cr.usgs.gov/vdelivery/Datasets/Staged/Elevation/LPC/Projects/USGS_LPC_NC_Phase4_Davidson_2017_LAS_2019/
>
ftp://rockyftp.cr.usgs.gov/vdelivery/Datasets/Staged/Elevation/LPC/Projects/USGS_LPC_NC_Phase4_Davie_2017_LAS_2019/
>
ftp://rockyftp.cr.usgs.gov/vdelivery/Datasets/Staged/Elevation/LPC/Projects/USGS_LPC_NC_Phase4_Forsyth_2017_LAS_2019/
>
ftp://rockyftp.cr.usgs.gov/vdelivery/Datasets/Staged/Elevation/LPC/Projects/USGS_LPC_NC_Phase4_Gaston_2016_LAS_2019/
>
ftp://rockyftp.cr.usgs.gov/vdelivery/Datasets/Staged/Elevation/LPC/Projects/USGS_LPC_NC_Phase4_Iredell_2017_LAS_2019/
>
ftp://rockyftp.cr.usgs.gov/vdelivery/Datasets/Staged/Elevation/LPC/Projects/USGS_LPC_NC_Phase4_Lincoln_2016_LAS_2019/
Reply all
Reply to author
Forward
0 new messages