Forum

Author Topic: Volumes  (Read 12664 times)

Peter

  • Jr. Member
  • **
  • Posts: 71
    • View Profile
Volumes
« on: January 21, 2013, 10:09:03 AM »
In the the tutorials there is an option when doing volume calculations that does not involve gps gcp, just set scale in model byt setting an known distance between two points.

How much accuracy will i loose with this method, maybe none?
What is the accuracy when doing volume measurement with good photos and gcp (under idela conditions that is), i have heard number of 1%

George

  • Full Member
  • ***
  • Posts: 193
    • View Profile
Re: Volumes
« Reply #1 on: February 02, 2013, 12:27:51 PM »
I would like to support Peter's question about volumes.
It's a bit confusing tool the "Volumes calculator" at Pscan.
Fact is that Pscan doesn't know what is the gound/zero or reference level under say pile of gravel or soil etc.
The tools has serious limitations tho.
Talking about accuracy of the tool is practically difficult issue as well.
Processing volumes calculations at any GIS package is much better and the accuracy will be almost the accuracy of Pscan geometry reconstruction - almost 100%.
Otherwise, Alexey could you clarify please the logic behind the tool?
Thanks.

Peter

  • Jr. Member
  • **
  • Posts: 71
    • View Profile
Re: Volumes
« Reply #2 on: February 04, 2013, 03:13:45 PM »
Yes, som light on this would be intresting!

Alexey Pasumansky

  • Agisoft Technical Support
  • Hero Member
  • *****
  • Posts: 15282
    • View Profile
Re: Volumes
« Reply #3 on: February 04, 2013, 05:25:07 PM »
Hello Peter,

The accuracy will be higher if you use several precise GCPs and not just single distance. The use of GCPs also allows to generate georeferenced DEM and to measure volume in the external GIS applications based on such DEM. It is recommended way since such GIS applications have tools to specify the area of measurement, the ground surface and are able to calculate difference volumes.
While PhotoScan uses more simplified method based on the closed model volume and is not able to control the shape of the bottom surface and requires additional manual job like model cropping and performing close holes operation.
Best regards,
Alexey Pasumansky,
Agisoft LLC

Peter

  • Jr. Member
  • **
  • Posts: 71
    • View Profile
Re: Volumes
« Reply #4 on: February 11, 2013, 11:29:32 AM »
Yes but if you isolate the area to be measured and do the close hole procedure, then it would be an reliable method for volume measurement?

As in the tutorial.

Alexey Pasumansky

  • Agisoft Technical Support
  • Hero Member
  • *****
  • Posts: 15282
    • View Profile
Re: Volumes
« Reply #5 on: February 13, 2013, 03:58:51 PM »
Hello Peter,

The problem with the measurement accuracy could appear in case the bottom of the model created by PhotoScan using Close Holes tool differs greatly from the ground surface shape.
Another problematic situation will appear if it is require to measure the volume of slightly overlapping piles - in this case it will be almost impossible to specify the models bottom along the ground, since at the side of overlap model is not close to the ground level.

So we assume that the proposed method will not work accurately for the described cases.

Best regards,
Alexey Pasumansky,
Agisoft LLC

Peter

  • Jr. Member
  • **
  • Posts: 71
    • View Profile
Re: Volumes
« Reply #6 on: February 13, 2013, 06:39:10 PM »
Yes i am with you on that, but what if the ground plane differs from when the reference scanning was done then it is not reliable method. The ground level tends to change if they move around material

So as long as the piles are not overlapping pscan would to a great job at calculating the piles?