RIEGL VZ600i

Discuss Riegl hardware here.
Post Reply
pfirmsto
I have made 20-30 posts
I have made 20-30 posts
Posts: 28
Joined: Wed Sep 22, 2021 11:19 am
2
Full Name: Peter Firmstone
Company Details: Zeus Project Services Pty Ltd
Company Position Title: Director
Country: Australia
Linkedin Profile: No
Has thanked: 14 times
Been thanked: 13 times

Re: RIEGL VZ600i

Post by pfirmsto »

Sub cm accuracy for earthworks.... :lol:
1cm over 1 square km is 25 tonnes. Airport runway? I don't know the application, I'm guessing for argument's sake.
If i were to do a job with a gls or sx12 hat large i just wouldn't quote. What a pita this would be to scan.

Loads of people moan and have comments about a Riegl system but have never used one. With the use of targets you can get sub cm accuracy overall. It's just procedures followed.
Hmm, you were previously recommending RTK, not targets, but ok, let's consider targets...

I haven't seen anyone moaning about the Reigl, maybe I missed those comments. Would I want to use a Riegl system for the entire job? How would it handle vibrations from an operating plant when it's tripod mounted on gridmesh walkways? The NavVis VLX is more accurate in an environment subject to vibrations. Each tool has its place. The Reigl can no more replace the SX12, than it can replace the NavVis VLX, but it can be used with both those tools, sure.

You're doing pretty well if you're getting sub cm accuracy overall, did you perform least squares analysis on control for target positioning?

We had a lot of base station data, 18 weeks of static observations processed by AUSPOS, position accuracy was 1 mm after using GNSS vectors and covariance matrices from AUSPOS in least squares, with the rest of the control network sub 4mm, most around 3mm. Other GNSS locations we processed 2 weeks of static data for 4mm position accuracy, and 4 weeks for 3mm, we checked GNSS for level and position daily, they ran off solar. I'm quoting absolute position accuracy in GDA2020 @ 95% confidence, referring back to fiducial coordinates, for our control, then I need to add in accuracy for the scans, which is additional to control, which means I'm approaching the accuracy you claim to achieve using the Reigl? Most people do at most 24 hours static GNSS observations, and that's not going to be sub cm.

The Reigl isn't designed to position over a point, it's resected into position when using targets, when you position an SX12 or GLS over control, with precise centring prisms for targets at 100m, and the entire network is least squares adjusted, it gives you confidence. Scanner targets have 2mm std deviation target centring. Prisms can be placed at distance outside the scan. The Reigl is faster and more accurate than a Leica RTC360, but it can't match the accuracy of a least squares adjusted GLS and SX12 network, nor can the latter match the speed of a Reigl or RTC. If scan targets aren't measured by a total station and adjusted using least squares, then the registration will accumulate errors.

My point is, we have to go to extreme lengths to achieve the sort of accuracy you're quoting, and we're taking care to level and centre over every control point with precision.

If there's not much information in the environment to register to, RTK (30 second occupations) with IMU isn't going to provide high accuracy, but then maybe high accuracy isn't required, but that's what you recommended.

I can't just grab one tool and recommend it for all jobs, I know you like your scanner, and you want to tell everyone how great it is, there's nothing wrong with that, but I suspect you might have some familiarity bias. :) I think there's more to be understood about the application before making a recommendation in this case.
User avatar
landmeterbeuckx
V.I.P Member
V.I.P Member
Posts: 1622
Joined: Tue May 01, 2012 5:19 pm
11
Full Name: Lieven Beuckx
Company Details: Studiebureau Beuckx
Company Position Title: Owner
Country: Belgium
Linkedin Profile: Yes
Has thanked: 185 times
Been thanked: 557 times

Re: RIEGL VZ600i

Post by landmeterbeuckx »

pfirmsto wrote: Sun Feb 11, 2024 11:45 am
Sub cm accuracy for earthworks.... :lol:
1cm over 1 square km is 25 tonnes. Airport runway? I don't know the application, I'm guessing for argument's sake.
If i were to do a job with a gls or sx12 hat large i just wouldn't quote. What a pita this would be to scan.

Loads of people moan and have comments about a Riegl system but have never used one. With the use of targets you can get sub cm accuracy overall. It's just procedures followed.
Hmm, you were previously recommending RTK, not targets, but ok, let's consider targets...

I haven't seen anyone moaning about the Reigl, maybe I missed those comments. Would I want to use a Riegl system for the entire job? How would it handle vibrations from an operating plant when it's tripod mounted on gridmesh walkways? The NavVis VLX is more accurate in an environment subject to vibrations. Each tool has its place. The Reigl can no more replace the SX12, than it can replace the NavVis VLX, but it can be used with both those tools, sure.

You're doing pretty well if you're getting sub cm accuracy overall, did you perform least squares analysis on control for target positioning?

We had a lot of base station data, 18 weeks of static observations processed by AUSPOS, position accuracy was 1 mm after using GNSS vectors and covariance matrices from AUSPOS in least squares, with the rest of the control network sub 4mm, most around 3mm. Other GNSS locations we processed 2 weeks of static data for 4mm position accuracy, and 4 weeks for 3mm, we checked GNSS for level and position daily, they ran off solar. I'm quoting absolute position accuracy in GDA2020 @ 95% confidence, referring back to fiducial coordinates, for our control, then I need to add in accuracy for the scans, which is additional to control, which means I'm approaching the accuracy you claim to achieve using the Reigl? Most people do at most 24 hours static GNSS observations, and that's not going to be sub cm.

The Reigl isn't designed to position over a point, it's resected into position when using targets, when you position an SX12 or GLS over control, with precise centring prisms for targets at 100m, and the entire network is least squares adjusted, it gives you confidence. Scanner targets have 2mm std deviation target centring. Prisms can be placed at distance outside the scan. The Reigl is faster and more accurate than a Leica RTC360, but it can't match the accuracy of a least squares adjusted GLS and SX12 network, nor can the latter match the speed of a Reigl or RTC. If scan targets aren't measured by a total station and adjusted using least squares, then the registration will accumulate errors.

My point is, we have to go to extreme lengths to achieve the sort of accuracy you're quoting, and we're taking care to level and centre over every control point with precision.

If there's not much information in the environment to register to, RTK (30 second occupations) with IMU isn't going to provide high accuracy, but then maybe high accuracy isn't required, but that's what you recommended.

I can't just grab one tool and recommend it for all jobs, I know you like your scanner, and you want to tell everyone how great it is, there's nothing wrong with that, but I suspect you might have some familiarity bias. :) I think there's more to be understood about the application before making a recommendation in this case.
The rtk which is used wil be used for fast registration and sometimes when the data isn't critical it will be used for final delivery.

I always tie in with fixed points whereas i'm as opposed to many here will be onsite for other things concerning these prohects like stakeout. So i always place tiepoints wich are measured seperately with rtk, total staion or combined and usage of a digital level involved. It all depends on the usage of the data further in the proces.

The accuracy of sx12,... and other scanning robots is of course very good when using prosms but the scanning data itself will also have noise and there isn't absolute accuracy of 0mm. So your scanning data will also be off compared to points measured with TS and DL.

A riegl can be used in traditional surveying methods. It has got a laser plumb so above centered can be done but isn't used by 99.99% of Riegl users i think.

I'm maybe biased but this is in fact a Riegl thread about a new instrument with new functionality. Something that other brands like Topcon (old ancient scanners), Trimble (just a rebranded Z&F),.... lack. You may not believe me but Riegl is ahead of it's competitors with a few years : registration in the instrument was possible 5 years ago, pictures while scanning 4 years ago (still not possible with any other TLS scanner),....
LSBbvba
Surveying services - 3D Laserscanning
Tel : +32477753126
www.lsbbvba.be
[email protected]
pfirmsto
I have made 20-30 posts
I have made 20-30 posts
Posts: 28
Joined: Wed Sep 22, 2021 11:19 am
2
Full Name: Peter Firmstone
Company Details: Zeus Project Services Pty Ltd
Company Position Title: Director
Country: Australia
Linkedin Profile: No
Has thanked: 14 times
Been thanked: 13 times

Re: RIEGL VZ600i

Post by pfirmsto »

justin.scott wrote: Mon Jan 29, 2024 7:14 am Hi All, thanks for the great content. My company (AUS based) are the proud owners of the new VZ600i system. We're having great success scanning urban environments but have some more challenging terrain coming up on future jobs.

What's the best way to manage accurate scan registration in rural environments where there are not many vertical features (such as buildings, trees etc) to capture overlapping data to enable to registration?

Reflectors/cylinder targets are an option but we don't want to have multiple targets in every scan or the efficiency will be great diminished.

Cheers,
Justin
Hey Justin,

Summing up, there are different tools for different jobs, the VZ is ideal for urban environments, if you want accuracy in a rural environment, where there's not many features for registration, use a scanner with a survey workflow, so you're not dependant on environment features for registration, use static GNSS positions at the perimeter of the area you wish to scan, then use GNSS vectors and covariance matrices from AUSPOS Sinex files in a least squares adjustment, this will keep your registrations tight and well aligned.

You may find areas in the environment with enough features for cloud to cloud registration algorithms, there's no harm testing the VZ600i in difficult areas and checking results or registering against the scans from an Leica MS60, Trimble SX12 or Topcon GLS-2200 using a traditional traverse survey workflow.

The VZ is fast, another strategy might be to the cover entire area quickly with the VZ, if you've got an accurate framework to register to (Point clouds from MS60 / SX12 / GLS), use the VZ scans that register to those and discard the ones that won't register.

Be very careful when considering using cloud to cloud registration algorithms where there are insufficient registration features, better safe than sorry. Advice from someone who's been burnt before. People will tell you the latest c2c methodology is accurate and works, until it doesn't, it works in urban environments, not rural areas with limited features for registration.

Cheers,

Peter.
pfirmsto
I have made 20-30 posts
I have made 20-30 posts
Posts: 28
Joined: Wed Sep 22, 2021 11:19 am
2
Full Name: Peter Firmstone
Company Details: Zeus Project Services Pty Ltd
Company Position Title: Director
Country: Australia
Linkedin Profile: No
Has thanked: 14 times
Been thanked: 13 times

Re: RIEGL VZ600i

Post by pfirmsto »

There seems to be some confusion around accuracy, this is a great video showcasing different lidar technologies, but more importantly, their accuracy. The accuracy quoted is additional to control, total accuracy is control accuracy and registered point cloud accuracy, relative to control. Eg if control accuracy is 10mm and registered point cloud accuracy is 15mm, relative to control, then total accuracy is 25mm.

https://www.youtube.com/watch?v=0RXU_8stFOM
dhirota
V.I.P Member
V.I.P Member
Posts: 959
Joined: Sun Nov 01, 2009 11:18 pm
14
Full Name: Dennis Hirota
Company Details: Sam O Hirota Inc
Company Position Title: President
Country: USA
Linkedin Profile: Yes
Location: Hawaii, USA
Has thanked: 91 times
Been thanked: 381 times

Re: RIEGL VZ600i

Post by dhirota »

We have decided to use the SONY alpha7Riv external camera body since that is the camera that Riegl is currently calibrating rather than the SONY alpha 7Rv. The current value of using the external SONY is the Auto ISO which gives superior color image exposures, especially in moving exterior cloud cover and bright sun than the non-auto ISO of the internal Riegl camera. If you are not capturing color images than it does not matter. The SONY produces 5 or 6 images per scan position (SP) versus 36 images for the Riegl internal camera per SP.

Here are some SONY images from some rockface scans. Notice the Hawaii blue sky and white clouds are excellent.

1.SONY AUTO ISO IMAGE.jpg
2.ROCKFACE OCTREE COMPOSITE SONY AUTO ISO COLOR.jpg
3.ROCKFACE OCTREE COMPOSITE REFLECTANCE CHROME-ORANGE.jpg

Here are some details on clearing moving objects using the Riegl object removing using Riscan Pro that I posted on another LSF thread, but thought people interested in the VZ600i might be interested in the detail timings.

viewtopic.php?p=110524#p110524

Here are the Riegl RiSCAN PRO 2.19 reports from the wizard that does the post processing with a few check marks and a single click, from the above thread

URBAN ENVIRONMENT, FASTER SCAN SPEED, INTERNAL RIEGL CAMERA, MANY SCAN POSITIONS (SP), 5 TARGET RESPONSES per PULSE

viewtopic.php?p=107398#p107398

220 SCAN POSITIONS.jpg
NO MANUAL PROCESSING
POST PROCESS MOVING OBJECTS = 9.27 seconds per SP = 34 minutes
POST PROCESS OCTREE FOR REMOVAL = 29.45 seconds per SP = 108 minutes

UNCLEANED 220 SP = 3,538,850,848 points
CLEANED 220 SP = 1,655,381,938 points
VZ600i Laser pulse repetition rate =2,200kHz; natural targets range=220m; max targets per pulse=5; 9mm @10m; 31 seconds/SP including Riegl internal images.

NON-URBAN ENVIRONMENT, SLOWER SCAN SPEED, EXTERNAL SONY CAMERA, FEWER SCAN POSITIONS (SP), 15 TARGET RESPONSES per PULSE
4. OAHU-KANEOHE-PALI ROCKFACE-SONY AUTO-ISO.jpg
5. OAHU-KANEOHE-PALI VALLEY-ROCKFACE-REFLECTANCE.jpg
18 SCAN POSITIONS.jpg
NO MANUAL PROCESSING
POST PROCESS MOVING OBJECTS = 17.27 seconds per SP = 5 minutes 11 seconds
POST PROCESS OCTREE FOR REMOVAL = 33.33 seconds per SP = 10 minutes

UNCLEANED 18 SP = 512,133,470 points
CLEANED 18 SP = 296,987,186 points
VZ600i Laser pulse repetition rate =600kHz; natural targets range=420m; max targets per pulse=15; 6mm @10m; 112 seconds/SP including external SONY 7Riv images.


Although there is a tunnel and highway, the moving objects are probably the wind moving the many trees along the rock face above the tunnel/highway.

I have 600 SP in the urban environment using our Riegl VZ600i that average between 7 to 10 seconds per SP for flagging moving point cloud objects and about 30 seconds per SP for generating the OCTREE, 5mm grid point cloud. There seems to be a difference in the efficiency of removal with the number of SP, with more being better.

POST PROCESSING USING: AMD 7950X ZEN4 16 cores CPU 5.5GHz; 192 GB RAM DDR5; BOOT NVMe CRUCIAL T700 4TB 12GB/sec
You do not have the required permissions to view the files attached to this post.
Dennis Hirota, PhD, PE, LPLS
www.samhirota.com
[email protected]
Post Reply

Return to “Riegl”