Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.

Messages - Bruno Andrieu

Pages: 1 [2] 3 4
General / decrasing the number of tie points : how is it done ?
« on: November 04, 2014, 12:21:57 AM »
In the 1.1 version, there is now a tool to reduce the size of the sparce point cloud
("demarrier les points unitives" in the french version).

There was already the  "gradual slection"  to suppress outliers.  But I do not see how the new tool proceed.
can you explan the rule used to chose the tie points that are eliminated by this tool and in which case is it useful ?


General / What means the colour of the points in the photo panel ?
« on: November 02, 2014, 01:24:44 AM »
Hi when looking on a photo in the photo panel we can see the point from used to build the sparse point cloud.

In grey I think it is all the candidate points in the initial image (?)
in blue it looks as the tie points (?)

They are also points in red and I do not know what they are ?

can you help ?

General / Re: what are "points unitives" in 1.1 release
« on: November 02, 2014, 12:53:59 AM »
Thanks,  I understand that tie points are the points of the sparse point cloud.

But note that the use of "point unitive" in the french version is not correct.
You could have used "point unitif", or better :  "point de liaison".

And : what would be really helpfull is to find a glossary in the help.

General / Re: Dense Cloud quality setting, what does it control?
« on: October 29, 2014, 11:21:08 AM »

Obviously the major effect is the higher density of the point cloud.
Beside this I could notice :
- On poor images, lower quality setting may allow photoscan to identify points in regions were no points are found at higher quality setting.  Because of the this, some time I merge dense points clouds calculated at two different quality settings (eg H and vH)
- 3D point positionning is less accurate in low quality setting ( probably more errors occur at the boudary between different objects)


General / Re: what are "points unitives" in 1.1 release
« on: October 27, 2014, 02:24:52 AM »
Thanks ,
But  I do not figure what they are and how to use them (?)

General / what are "points unitives" in 1.1 release
« on: October 26, 2014, 04:40:51 PM »

I photoscan 1.1 I find the new term "point unitive" (in the french version) this is for instance in the option for spare clous point generation and in the tool for the dense cloud.

Can you explain what it means ?



Feature Requests / Re: Marker point order broken.
« on: October 26, 2014, 04:19:24 PM »

What I would find useful is to be able to rank the marker list, either by alphabetic order or as it is now  by order of creating. At the moment it is sometime difficult to find a marker in the list ( as I use plenty of them).

Also , to name them, It would be nice that user can give a prefix (other than "point" and the new marker would be created with that prefix, until a new prefix is given.

I would find this useful so as to give consistent marker names on several chunks that we want later to align based on marker.

General / Re: image correction useful ?
« on: October 26, 2014, 01:55:59 PM »
Thanks for the answers.
My question was specifically about the use of software such as Dxomark, that do "professionnal  quality" image correction with output looking better  than when using  the software within the cameras.

When  you disable all camera software (eg on nikon 7000 or 5300) and do no correction at all, an enlarged image (eg 400%) show some defects (stringes, color noise, low FTM, etc).

I checked that for  "high iso" (> 400) and "medium quality" lenses, the 3D point clouds I obtained were much more dense when using with Dxomark corrected images than with  no correction at all.
However they are a lot of parameters in the software and my question is to clarify which corrections do really help and which should better be avoided (not speaking about image cropping, etc, which obviously is forbidden).


General / image correction useful ?
« on: October 25, 2014, 05:26:50 PM »
Hello Agisoft team
I use Nikon camera with standard nikon lenses (24mm or 35mm, fixed focal lenses) to take images.
Softwares as Nikon Capture or Dxomark allow correcting for lense distorsion. Would you recommend to use these corrections before using photos in Agisoft or rather work with the uncorrected images ?

Could you comment also for other options such as image sharpness improvment and noise reduction ?




Feature Requests / helping to identfy mispositionned markers
« on: October 25, 2014, 04:23:29 PM »
I frequently  mark points  on the images  to improve camera alignment. Doing that some mispositionning happen and it is  a long process, when a marker has a high rmse,  to go through all photos looking for positionning errors.

It would be great  that in the photo panel and/or on each photo we can see by the colour of the marker the distance between the actual position of the marker and the expected position (given the position of the marker on other aligned photos).

That is the green colour that is used low would change to orange or red if the marker is likely mispositioned on a photo.


Feature Requests / Re: Point Cloud classification by color
« on: May 25, 2014, 01:59:08 PM »

this would be a very usefull feature indeed.
With 2D images classifying pixels by colour has plenty of scientific applications. So this would be a nice extension to 3D images.

So ++


Feature Requests / Re: exort ortho_image using point cloud
« on: May 17, 2014, 10:58:42 AM »
Thanks Alexey

If I build mesh from sparse cloud then I have a mesh with a low resolution, and I cannot really use it.

It would be usefull for me  to understand which data are used to build orthophotos: can you explain ?

It would be useful for me to generate high resolutions 2D views of my models. At the moment I use copies of my screen which displays the point cloud, but this gives not full resolution on the whole model. Can you explain how I should proceed ?


Feature Requests / exort ortho_image using point cloud
« on: April 13, 2014, 02:02:21 PM »

Is it possible to export ortho images computed from the point cloud instead of the mesh ?

building high resolution  textures  usually fails  with my computer and  images from vH points cloud would be sufficient.


Pages: 1 [2] 3 4