Topic: Camera informations

Report Abuse Report Abuse
jediknight (Over 1 year ago)

I'm trying to implement a filter for meshlab ( that imports the point cloud and the camera informations about each shot taken, to allow the user to align the view in the 3D viewport to a given shot.

In order to achieve such a result, i need some information about the camera. I spent some time analizing the json string returned by the photosynth server and the source code of synthexport ( and I was able to find out some of the information I need.
I found that each camera is described by some parameter, among which there are some of those I'm interested in.
Anyways I still miss these ones:
- position of the radial distortion center on the image plane in pixels
- CCD sensor width and height in mm
- dimension in mm of a single pixel
Moreover I can't exactly figure out the meaning of all the informations about the camera given in json.

Does anybody know their meaning?
Any help is appreciated
Nathanael (Over 1 year ago)

I'm afraid I won't be able to be of as much help on this topic, but here are a few leads anyway.

I know you already looked at Christoph's SynthExport app. I suppose if you don't get an answer here, you could try asking the community over there on this thread, if you haven't already:

Kean Walmsley, Josh Harle, and Henri Astre would be good people to ask in addition to Christoph Hausner. 

Also, is your effort to write this functionality what ALoopingIcon is referencing ( ) over at the photogrammetry forum?
Nathanael (Over 1 year ago)
Sadly the few remaining shreds of official technical information aren't documenting the .json formatting, but I'll post them here anyway, as off-topic as they may be. Perhaps you'll see something that my untechnical eyes didn't see.

The Photosynth viewer has a JavaScript API to interact with that is documented here:

Also related to the viewer, you may have noticed that 'Share' links have unique variables attached to the normal view page and Photosynth collection ID where, 'i' is the Photosynth projector, 'z' is the zoom level, 'g' is the clump of photos you are navigated to, 'p' is the pan position, and 'm' is whether the viewer is in 2D mode.

I don't know if there's any overlap in variables between the .json file and those used by the viewer, but that's about all I can offer.

Photosynth team, we could use your help here.
Nathanael (Over 1 year ago)
For cross referential purposes, JediKnight also started a corresponding topic on the old forum which can also be checked for answers:
jediknight (Over 1 year ago)
Hi Nathanael,

to answer your question (with a "little" delay, sorry):
"Also, is your effort to write this functionality what ALoopingIcon is referencing ( ) over at the photogrammetry forum?"

Yes, that's the functionality I'm working to. It's almost complete now, there are still some issues concerning the camera matching between the 3d view and the shots...