Messages

31

General / Re: Refined camera calibration parameters

« on: January 11, 2013, 03:29:22 PM »

Does anyone know what is the reason for the extreme deformation of the model with photos (or frames)  extracted from a video? Is it the lack of EXIF-data?
I intend to use this method, taking photos from videos, more intensively in the future and would like to avoid any negative surprises.

32

General / Re: Accuracy (round N)

« on: December 20, 2012, 01:55:23 PM »

Don't know where is the problem. The results look fine, partly even incredibly good. The 3 spikes might have to to with bad overlapping or any other reasons, but not really a problem.

33

General / Re: Virtual interior of the scenery

« on: December 12, 2012, 11:58:45 PM »

 
Very impressive! Could you please leave a few words about the "Making of..."? What software used, how's the process step by step, what's important to take care of and so on.

34

General / Re: GEoTIFF to AutoCAD .dxf - Help needed

« on: December 07, 2012, 04:32:11 PM »

Thanks. This is of course the most appreciated solution to me.
But...as I am not experienced with AutoCAD...can the DEM be created from out of AutoCAD?

35

General / GEoTIFF to AutoCAD .dxf - Help needed

« on: December 07, 2012, 03:45:52 PM »

My customer needs a DEM in AutoCAD.dxf file format. Does anyone know how to do that?
I've tried a vectoring tool named Vextractor which is expected to do the conversion from .tiff files to .dxf, but the result is simply nonsense. It just creates a few erratic vertical and horizontal lines.
I know that ArcGIS, QGIS and so on are the usual tools, but too complex to learn for such a primitive requirement.
Besides the special request for .dxf: I don't understand why Agisoft still doesn't provide a simple tool to create .jpg files to show the contents of a DEM (isohypses) for avoiding this .tiff-nightmare. Having to spend hundreds of dollars for Global Mapper (which is usually suggested by Agisoft to use) cannot be a solution. This is absurd!

Oliver

36

Feature Requests / Re: Mosaic Edition

« on: December 05, 2012, 01:48:47 AM »

I agree about the attractiveness of this feature in Pix4D, though the editing of the seamlines is a nightmare to get used to. I played around with it for an hour or so.
On the other hand, if you use it intensively on a model and say have a lot of buildings in it, you'd get a picture where every building raises into another direction. Not what I'd call a useful orthophoto.

What I'd consider much more desirable is finding a method that applies a clean texture to vertical objects, like walls. I am not an expert in it but I am surprised that there is no tool which can use existing vertical textures (like through the use of very wide angle photo lenses and intensive overlapping) and put it onto the mesh without those terrible and extreme distortions.

37

General / Re: The size of the resulting PSZ and JPG

« on: October 02, 2012, 07:01:07 PM »

What PDF-Reader is necessary to view these big datasets in 3D? With normal Acrobat Reader X it doesn't work on my machine? Is it any special PDF Reader?

br
Oliver

38

General / Re: align/merge chunks with markers

« on: June 05, 2012, 04:29:03 PM »

I had this twice too.

39

General / Re: How to handle strong altitude differences on the ground?

« on: June 05, 2012, 04:27:53 PM »

Thank you for these arguments.
Yes, I can control the altitude of the drone precisely an assure a constant distance to the ground.
Think I will try both ways but assume, b) will work fine.

40

General / How to handle strong altitude differences on the ground?

« on: June 04, 2012, 04:28:13 PM »

Does anyone have experience with measuring areas with altitude differences of up to 80 meters?
I will have to determine volumes of different heaps in a stone pit (stone quarry), which has lows of 300 m above sea level and highs of 380 meters. Normally we fly in altitudes of 80 m above ground. To get a homogenous and correct mesh, is it better to
a) use a greater overall flying height and take all shots from the same altitude (which resulted in 60-140 m distance to the ground) or
b) could I keep a constant distance to the ground by altering the altitude of the microdrone?

I'd prefer b) as the ground resolution would be better and equal on all photos, but I'm not sure if PS would calculate a wrong 3D model.

Thank you for help.
Oliver

41

General / Re: Input resolution vs. output resolution?

« on: June 04, 2012, 04:08:26 PM »

I didn't test this yet but I'm pretty sure that it is nonsense to believe, the resolution would get worse, just because it is RAW. The main purpose of using the RAW format in photography is being able to change white balance, color dynamic and other characteristics of the picture after making it. The resolution doesn't change, compared to JPEG.
If you really wanted to alter the white balance, make sure that you do it equally with all pictures. I see no reason to use RAW, as the aesthetic requirements are very limited in aerial photography.

42

General / Re: What Vertical Accuracy Should I get?

« on: May 27, 2012, 06:05:48 PM »

My last project with 160 photos produced the following results with 10 GCPs and 5 control points.
The GCPs are the 0s:

Error(m)         X error         Y error      Z error
0.000000         0.000000         -0.000000   0.000000         
0.019324         -0.003320      0.003339      -0.018741         
0.046620         -0.030979      0.010814      0.033117         
0.052064         -0.007972      0.021518      -0.046734         
0.000000         0.000000         -0.000000   0.000000         
0.000000         -0.000000      -0.000000   -0.000000         
0.081491         0.034416         -0.034463   0.065336         
0.000000         -0.000000      -0.000000   -0.000000         
0.000000         0.000000         -0.000000   0.000000         
0.000000         0.000000         -0.000000   -0.000000         
0.000000         0.000000         -0.000000   -0.000000         
0.000000         0.000000         -0.000000   -0.000000         
0.000000         0.000000         -0.000000   -0.000000         
0.000000         0.000000         -0.000000   0.000000         
0.084857         0.008853         0.015547      -0.082950

I  used a 16 mm lense and a 24 MP camera. Certainly important is the overlapping of photos. I used 65% in both dimensions.

What's interesting to me is: how do you find out the CE95 and LE95 values? It would be great to have a tool in PS for them. Customers are interested in an objective criteria for the accuracy.
Greetings
Oliver

43

General / Re: export points and marker's image coordinates

« on: May 15, 2012, 05:34:18 PM »

Thank you, good to know this.

What's interesting to me is, what precision do you guys achieve with your models? My own errors range between 1 and 10 cm horizontally and around 20 cm vertically. Currently I am thinking about ways how to improve these values, especially vertically.

Oliver

44

General / Re: KMZ won't open in Google Earth

« on: May 11, 2012, 10:16:17 AM »

1,2 GB...Jesus! I even had troubles with 70 MB in GE. How did you do that?
Yes, I always used the standard settings.
Try a higher compression rate of the texture .jpg file.

45

General / Re: KMZ won't open in Google Earth

« on: May 10, 2012, 12:32:25 AM »

How big is the file? Beyond 30 MB it didn't work well in my tests. Expect at least 20 seconds up to 1 minute until it opens. Could be that you need to change the "height" parameter, which behaves somehow strange. It worked when I used "Absolute height" and played around with different values.
How did you create the .kmz? The most stable way for me was to import the .dae as a "Model". The texture  is imported automatically. After that, you save it and the .kmz is constructed correctly by GE. In other methods I had troubles with the relative paths.
Oliver