Support Website Contact Support Blog

Criss Cross Obliques With Less Sidelap Versus One Direction?

I’m doing a topographic survey of a rural area with about 500ft of elevation change (I’m breaking it up into sections where each section has only 250ft of elevation change).  Will output a DXF with contour lines and possibly a DEM or DTM.  I only need accuracy within 12 inches.  Will have ground control points.

Obviously when you are generating output for 3D features, it’s more accurate to also have obliques.  One can see the crispness of the structures and lack of blobbing/noise that the use of obliques contributes. 

Given the accuracy requirements and that this is mainly for land features, not structures, I am wondering the following:

(a) In general you want 1/3 GSD so 4in/pixel for 12-inch accuracy, that’s low, I will be twice better GSD.  And I’ll have 80/70 overlap which is also pretty decent.  In that case, are obliques in reality a waste of time?

(b) If I did do obliques, what would produce better results, run a mission in one direction with 80/70 frontlap/sidelap, or run two missions in two different directions (criss cross) but with 50/40 overlap/sidelap.  In other words is having more overlap in one perspective more valuable than having two different perspective obliques with less overlap?  Again this is for land topography measurement, not 3D modeling of structures.

 

 

Normally for a topographic survey you only need nadir pictures and single grid flight.

The oblique and/or double grid it’s for having good building or side reconstructions (normally for 3d bases).

No need and it’s a waste of time if your output it’s only a topographic survey.

But be careful, if you have trees and want to pick the level in the base of the tree it’s good to have some obliques too… or higher overlap

Hi Jeff,

As mentioned by  @Jorge , nadir grid flights should be enough for topographic surveys.

I would suggest you check out our article about Selecting the Image Acquisition Plan Type.
In a general manner, if your terrain has no vegetation or buildings, a single grid in nadir is enough. If there are some buildings, you could consider doing a “double-grid” as we call it in our app, which would be the equivalent of the “criss-cross” you are referring to.
Note that, as you will map a rural area, the overlap (or the flight height) could be increased if you are going to map fields for example (see here).

Additionally, regarding the change in elevation in your area, the following links could help you plan your missions:
Can Pix4Dmapper process Images taken at different Flight Heights?
Image Acquisition Plan for Terrain with Height Variations

Regards,