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 - aphextwin

Pages: 1 [2] 3
General / Re: TOP cam for low-altitude UAV
« on: July 29, 2013, 02:09:02 PM »
yes... I see no other way getting the images for such a model.

General / Re: TOP cam for low-altitude UAV
« on: July 29, 2013, 12:19:06 PM »
Hi David,

3. this model is not altered by zbrush or any other software, it?s an OBJ-Export from PS.

2. No, besides the SfM, the photos are only used for texture-generation (which is done in PS)

1. Blender for animation and rendering as it is simple to use, quick and effective.

I hope these informations give you a little overview.

General / TOP cam for low-altitude UAV
« on: July 27, 2013, 06:54:04 PM »
Just bought this one:

- Built-in Intervalometer with EVERY option you wish
- CCD sensor (global shutter)
- fixed zoom lens
- light weight (about 220g)
- 10MP quite enough (imho)

Just in case smbdy is interested, I did this with this camera:

General / Re: Capturing a Coin and Merging Chunks
« on: July 22, 2013, 10:35:55 PM »
Mine's not working yet, but I'm sure it will... just need more time =/

Feature Requests / Re: Cut and trim by masks
« on: July 16, 2013, 11:46:55 PM »

General / Re: Capturing a Coin and Merging Chunks
« on: July 16, 2013, 11:36:52 PM »
I just got a crazy idea concerning our coin-issue... I'll try something tomorrow and tell you if it works... if not, I'll tell you anyways, but not this evening... If it works, alle problems that involve coin-reconstruction are solved ;-)

General / Re: not align phots
« on: July 16, 2013, 02:10:32 PM »
There are some issues that you need to refine:
- The object needs to be focused on every image (which is not the case on most images)
- try to hold the brightness on the object on every image
- try to get "most of the image" on the sensor (get closer to the object, not get parts of the object, better the whole object)


General / Re: not align phots
« on: July 16, 2013, 01:37:17 PM »
Can you provide your images?

General / Re: Capturing a Coin and Merging Chunks
« on: July 16, 2013, 10:52:49 AM »

when you select all the information that comes with this image you'll notice that in the image you provided only about 80000 px are valid "information" but only few are used as features. The image contains 786432 px total, so only about 10% of the sensor data are actually used, the rest is noise. when the coin is in a perpendicular position, the information is even much less. Imho you need to maximize the data that comes with each foto so I'd recommend you to get the coin closer to the lens / use a bigger lens. Also I'd try to use a polarizer. You'll probably notice some light shining on the coin in certain rotations. You're shooting with ISO 200. Is this the lowest ISO of the camera? Besides there is no noise visible, taking a higher ISO will get you very little noise. When shooting still, there is no need to increase the ISO, so I'd bring it to the minimum of the camera. For PS this is a definite no go ;-)
just my 2 cents...

ps: I'd not use any macro stacks. I tried this once with the equipment in our university. Total waste of time, because the lens distortion is different with each stacked image.

General / Re: Capturing a Coin and Merging Chunks
« on: July 15, 2013, 03:18:07 PM »

can you provide a sample-image of your series taken for the coin?

I was not able to accomplish this task with PS some month ago, but I was able to make a 3d-scan of a coin with a SL-scanner and a macro-lens for the projector.
I guess my problem was that I didn?t geht the coin large enough on the camerasensor. And I don?t own a macro lens with a focal length that was great enough...

General / Re: masking problems please help
« on: July 15, 2013, 03:09:27 PM »
Sometimes it?s a good idea to use two chunks when processing an object that needs to be turned upside down by hand. I printed out some very small markers that I use for alignment reasons. Works very well for me, probably you give it a try :) If you need a textures model you can take extra-texture-photos without these markers and use them for texturing.

General / Re: How to overcome RAM Limitation
« on: July 10, 2013, 02:56:41 PM »
It won't swap the ram.

In my experience when the limit is reached it performs by a simple crash with the option to send the crash report to agisoft :-)

I desperately hope they are making good progress with distributed computing.

General / Re: About Markers
« on: July 01, 2013, 10:57:40 AM »
That was exactly what helped me understanding markers. Thanks a lot!


Markers can be used for different purposes:
- georeferencing. When marker has at least two projections on aligned images and have got coordinates specified in Ground Control pane it can be used for referencing. All checked markers are used for georeferencing when Update button is pressed.
- optimization of camera alignment. The same requirement for markers as in previous point. Marker projections and coordinates are included as parameters for camera position refinement. Optimization is recommended for non-linear model distortion removal.
- chunk alignment.
- creating scale-bars.
- helping PhotoScan to re-align the sub-set of wrongly positioned cameras (more detailed procedure is described in User Manual on page 9).

Feature Requests / Move To Chunk
« on: June 29, 2013, 06:17:37 PM »
When a certain amount of images in workspace is selected a nice feature would be if you could move these images to a certain chunk by selecting the chunk in a sub menu. Drag and drop is possible but if you have a lot of images somewhere down in the list a move-to-chunk-feature would be nicer.

General / Re: About Markers
« on: June 26, 2013, 01:04:29 PM »
nobody can help?

Pages: 1 [2] 3