I have to align an ancient temple using LSPs in Greek Grid and photos in WGS84, which I choose to do pillar by pillar, all the pillars are aligning just fine in the Greek Grid CoordSys.
Although the photos are in WGS84, RC understands the difference just fine…
Almost… because two specific pillars are aligning as shown in the picture bellow…
I believe there is something wrong with the weighting strategies for the Geotags but not sure how to fix that, any ideas?
I supose that a factor of 0 to 1 is used as weight factor for the Geotags, is that correct? I also believe that mainly this is because the elevation captured with my Canon’s GPS is totally wrong…
I tried several weights and several tactics but nothing seems to work, only managed to split the scene in two correct components, that couldn’t merge in anyway!
I left open the Alignment settings window in 1D!
Hi ptokmaki@gmail.com
First make a backup save of the RC project, then delete the component and use alignment settings from this screenshot and try to align it again.
Alignment_settings_80k.jpg
Thank you so much, it worked!
Could you explain a few bits about the magic behind numbers…?
Hi ptokmaki@gmail.com
In short, these settings help to detect more points with higher quality. There are no magic numbers overall, I personally change them per project data IF I see some issues. But in most cases these values are good to start with. If it does not align, then it is in most cases not properly captured…
Hello Wishgranter,
Just a little question:
Which is the best distortion model to use for a normal lens under a normal situation? Division ? Brown 3?
Hi Manuel
In “normal situation” is the BROWN3 the most reliable distortion model…
Manuel wrote:
Hello Wishgranter,
Just a little question:
Which is the best distortion model to use for a normal lens under a normal situation? Division ? Brown 3?
Just some additional notes:
Division is recommended for fish-eyes optics (for example GoPro), reliably covers simple distortions. It works with radial distortion modelled by means of a one-parameter division model.
Brown3 For optics with less than 180°. This is a polynomial model of radial distortion with 3 modelling parameters. It is used as a default.
Brown4 is able to cover different distortion in the middle and borders of an image. This is also a polynomial model of radial distortion but with 4 coefficients.
Brown3 with tangential2 / Brown4 with tangential2 extend Brown3/Brown4 with tangential distortion, which can compensate offset of optics. Majority of current optics has it negligibly small.
K + Brown3 with tangential2 / K + Brown4 with tangential2 These models add to the Brown model a possibility to optimize the whole camera calibration, including Skew and Aspect ratio. If K + Brown3 with tangential2 / K + Brown4 with tangential2 is not used, by default RC assumes zero skew and aspect ratio as 1.
Brown’s distortion model can be used for radial distortion. The more parameters, meant as changes in polynomials, the more complex deformation.