Small movie of a construction site with excavation
- landmeterbeuckx
- V.I.P Member
- Posts: 1615
- 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: 183 times
- Been thanked: 548 times
-
- I have made 30-40 posts
- Posts: 34
- Joined: Sat Feb 15, 2020 10:13 pm
- 4
- Full Name: Veit Streicher
- Company Details: Navvis
- Company Position Title: Consultant
- Country: Germany
- Linkedin Profile: Yes
- Has thanked: 60 times
- Been thanked: 14 times
Re: Small movie of a construction site with excavation
Nice Movie. However I guess you need a mobile system to cover the blind spots...
- landmeterbeuckx
- V.I.P Member
- Posts: 1615
- 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: 183 times
- Been thanked: 548 times
Re: Small movie of a construction site with excavation
No blind spots, the X-ray views from Riscan make it seem that where the scanner was positioned there was no data. In reflectance view all data is there.
At the bottom there was some water after some showers.
At the bottom there was some water after some showers.
- jcoco3
- Global Moderator
- Posts: 1724
- Joined: Sun Mar 04, 2012 5:43 pm
- 12
- Full Name: Jonathan Coco
- Company Details: Consultant
- Company Position Title: Owner
- Country: USA
- Linkedin Profile: No
- Has thanked: 70 times
- Been thanked: 157 times
Re: Small movie of a construction site with excavation
Nice! Got to love the refreshing look of range independent reflectivity. Void circles under Riegl scanners are larger than I am used to, but at least they can fill them in easily from a distance. I was recently reading about TVG (time varied gain) as it relates to sonar processing, and the results seem very similar to the signal processing that Riegl performs on their data. Wonder why more manufactures don't apply the same type of processing? Doesn't Leica also do this now, but perhaps in a different way? Who else? Guess its time to do some digging on the forum
- smacl
- Global Moderator
- Posts: 1409
- Joined: Tue Jan 25, 2011 5:12 pm
- 13
- Full Name: Shane MacLaughlin
- Company Details: Atlas Computers Ltd
- Company Position Title: Managing Director
- Country: Ireland
- Linkedin Profile: Yes
- Location: Ireland
- Has thanked: 627 times
- Been thanked: 657 times
- Contact:
Re: Small movie of a construction site with excavation
I've been doing some work recently on histogram based intensity correction in SCC with some interesting results but it would certainly be interesting to experiment with inverse distance based correction. Problem is that it demands a format that includes scanner position which is doable for some instruments and data formats but not for all. For those in development such as myself, it would be really beneficial to be able produce a range and observer position / setup number for each point.jcoco3 wrote: ↑Mon Jun 08, 2020 1:58 pm Nice! Got to love the refreshing look of range independent reflectivity. Void circles under Riegl scanners are larger than I am used to, but at least they can fill them in easily from a distance. I was recently reading about TVG (time varied gain) as it relates to sonar processing, and the results seem very similar to the signal processing that Riegl performs on their data. Wonder why more manufactures don't apply the same type of processing? Doesn't Leica also do this now, but perhaps in a different way? Who else? Guess its time to do some digging on the forum
- landmeterbeuckx
- V.I.P Member
- Posts: 1615
- 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: 183 times
- Been thanked: 548 times
Re: Small movie of a construction site with excavation
The holes are visible because i didn't decimate this pointcloud. These views are excellent for visualizations. I use them a lot. Because Riegl has so many viewtypes styles these x-rays can be different in dozens of ways.jcoco3 wrote: ↑Mon Jun 08, 2020 1:58 pm Nice! Got to love the refreshing look of range independent reflectivity. Void circles under Riegl scanners are larger than I am used to, but at least they can fill them in easily from a distance. I was recently reading about TVG (time varied gain) as it relates to sonar processing, and the results seem very similar to the signal processing that Riegl performs on their data. Wonder why more manufactures don't apply the same type of processing? Doesn't Leica also do this now, but perhaps in a different way? Who else? Guess its time to do some digging on the forum
Yan Koch is doing something similar with his new project ScanSap. He's also using these views. Did some tests with some of my Riegl data and it came out nice.
- landmeterbeuckx
- V.I.P Member
- Posts: 1615
- 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: 183 times
- Been thanked: 548 times
- jcoco3
- Global Moderator
- Posts: 1724
- Joined: Sun Mar 04, 2012 5:43 pm
- 12
- Full Name: Jonathan Coco
- Company Details: Consultant
- Company Position Title: Owner
- Country: USA
- Linkedin Profile: No
- Has thanked: 70 times
- Been thanked: 157 times
Re: Small movie of a construction site with excavation
Yep, understood, and I wasn't being a critic. They are just large in diameter as compared to most other terrestrial scanners and your video simply reminded me of that. I forget sometime how much larger the Riegl void is, because of the difference in field of view. As you said decimation or spatially filtering would reduce their appearance, but that's not what you were trying to achieve.The holes are visible because i didn't decimate this pointcloud.
Well that's a new set of terms for me Is it similar to Time Varied Gain in this case? I realize that for sonar and time of flight scanners, that "time" varied gain may be more appropriate terminology, but for phase shift I guess it would likely be something else.inverse distance based correction
Why is position needed, I would have thought that only the range would be needed? From what I understand reading on the concept of time varied gain, it is known that the signal(sound) amplitude will degrade/attenuate over distance and a correction (appears to be linear) can be applied by knowing the time or distance. It might not be that simple with a laser, I was just thinking that if fewer photons make it back(less intensity/amplitude) and we could determine what that loss would be over a distance (in a lab setting) for any laser then apply the same type of correction...That was me asking the question, why is it not that simple?Problem is that it demands a format that includes scanner position which is doable for some instruments and data formats but not for all. For those in development such as myself, it would be really beneficial to be able produce a range and observer position / setup number for each point.
Is it absolute position or relative that is needed? Seems like you get both of those from either GPS or the registration effort, but I can understand thats its not applied to every point in the file, and that makes software processing more difficult. Last year I was poking around with a hokuyo and got a rudimentary taste of how efficient the file formats must be when drinking data from such a fire-hose. I was a bit amazed at how streamlined the format is. Anything that needs to be state once is done in the header or head of each block group, and much of everything else that can be assumed is simply left out of the file. No angles are stored in the file, you are just supposed to know the step size and frequency of the device(from the spec sheet), and if you need that information to do your processing then you have to put it back in yourself or figure something else out
- landmeterbeuckx
- V.I.P Member
- Posts: 1615
- 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: 183 times
- Been thanked: 548 times
Re: Small movie of a construction site with excavation
Sometimes that is a problem but only a small one. We can scan so fast with this instrument if a position is critical we just move away 3m and scan again.jcoco3 wrote: ↑Mon Jun 08, 2020 4:59 pmYep, understood, and I wasn't being a critic. They are just large in diameter as compared to most other terrestrial scanners and your video simply reminded me of that. I forget sometime how much larger the Riegl void is, because of the difference in field of view. As you said decimation or spatially filtering would reduce their appearance, but that's not what you were trying to achieve.The holes are visible because i didn't decimate this pointcloud.
- smacl
- Global Moderator
- Posts: 1409
- Joined: Tue Jan 25, 2011 5:12 pm
- 13
- Full Name: Shane MacLaughlin
- Company Details: Atlas Computers Ltd
- Company Position Title: Managing Director
- Country: Ireland
- Linkedin Profile: Yes
- Location: Ireland
- Has thanked: 627 times
- Been thanked: 657 times
- Contact:
Re: Small movie of a construction site with excavation
We've a number of interference factors (error sources) such as refraction, PPM error, dust, rain, etc... that scale with distance. Not sure how much it effects laser scanners but with reflectorless total stations, the beam diverges slightly with distance which also effects return signal strength. Basically, for the same reflectivity, the returned position and signal strength will be weaker the further the laser is from the object being measured.jcoco3 wrote: ↑Mon Jun 08, 2020 4:59 pminverse distance based correction. Well that's a new set of terms for me Is it similar to Time Varied Gain in this case? I realize that for sonar and time of flight scanners, that "time" varied gain may be more appropriate terminology, but for phase shift I guess it would likely be something else.
Relative position for the most part, but relative to the job rather than just the current setup in the context of static scanning. The strength of the returned signal is highly dependent on the angle of incidence at which the laser beam hits the target object. e.g. if it is close to parallel it is very weak, if it is close to normal, it is far stronger. If we're looking at getting rid of noise from a dense point cloud with plenty of redundancy, this is very useful information. Range is more important, but observer position is valuable too, particularly in the context of SLAM where you've a bunch of extra interference factors at the equipment end. From a purely functional standpoint, position also gives you the possibility to re-colour the scan with external photography and carry out the type of point cloud rationalization techniques you see in the likes of NavVis. Coming from a survey background myself as a software dude, I'm always happier seeing all the raw observations rather than reduced coordinates as these provide the best opportunity for correction, manipulation and QA.Why is position needed, I would have thought that only the range would be needed? From what I understand reading on the concept of time varied gain, it is known that the signal(sound) amplitude will degrade/attenuate over distance and a correction (appears to be linear) can be applied by knowing the time or distance. It might not be that simple with a laser, I was just thinking that if fewer photons make it back(less intensity/amplitude) and we could determine what that loss would be over a distance (in a lab setting) for any laser then apply the same type of correction...That was me asking the question, why is it not that simple?
Is it absolute position or relative that is needed? Seems like you get both of those from either GPS or the registration effort, but I can understand thats its not applied to every point in the file, and that makes software processing more difficult. Last year I was poking around with a hokuyo and got a rudimentary taste of how efficient the file formats must be when drinking data from such a fire-hose. I was a bit amazed at how streamlined the format is. Anything that needs to be state once is done in the header or head of each block group, and much of everything else that can be assumed is simply left out of the file. No angles are stored in the file, you are just supposed to know the step size and frequency of the device(from the spec sheet), and if you need that information to do your processing then you have to put it back in yourself or figure something else out