scanning a garden results in 2 planes of points?

[JJ]

i've taken photos of a garden and have masked out sky and backgrounds.  the garden is about 50m long and 10m wide.

pictures taken with a micro four thirds camera - is this an issue?

as you can see i've ended up with 2 different planes of points.  is there a likely reason for this?

[ruyi7952]

Did you use GPS to record your location?
Do you notice the overlap rate?
Are the photos taken at the same focal length?

[JJ]

thanks for helping.  i am in the middle of having another go and it seems to be working better this time.  i used pair preselection and then aligned any remaining cameras separately.

to answer your questions though:

Did you use GPS to record your location?

no.  these were taken from the ground by the way.

Do you notice the overlap rate?

i probably didn't overlap as well as i should.  this garden is long and thin - about 50m long x 12m wide with some obstacles in the middle.

what would be the best way to approach photo taking in this situation?

Are the photos taken at the same focal length?

yes, all using the same 17mm lens (micro 4/3, so 2x crop factor and 34mm full frame equivalent FOV).

[JJ]

with Generic Preselection turned on the results were much better.

[ruyi7952]

yes.
Generic Preselection is the best option if the photo has no location information.

[JJ]

thanks for confirming.

so in this case i only needed a basic survey for dimensioning purposes.

if i wanted a more accurate scan how should i go about taking the photos?  on the short side (blue and red positions) it's pretty straight forward but if i want to photograph the yellow side, how far back should i stand? Pos1 or Pos2? or both? or some other positions?

[ruyi7952]

What you need is a precise relative position. Am I right? If so, you can print out some of the control points that can be automatically identified and take pictures of them. As for the distribution of control points, I recommend that you place them in a prominent position in the geometry element.