Messages

13186

General / Re: Presenting hundereds of models

« on: March 26, 2014, 04:52:27 PM »

Hello Mark,

Thanks for the screenshot. Maybe you can improve the export resolution to get more detailed output?

As some sort of work around you can use screenshot save using Python script:

Code: [Select]

from PyQt4.QtGui import QPixmap, QApplication
QPixmap.grabWindow(QApplication.desktop().winId(), x1, y1, x2, y2).save('d:/test.png', 'png')

13187

General / Re: Presenting hundereds of models

« on: March 26, 2014, 04:35:51 PM »

Hello Mark,

Can you please provide the DEM export results for any of the examples (performed through GUI export DEM dialog) so that we could see why this way is not acceptable?

13188

Face and Body Scanning / Re: Background masking process

« on: March 25, 2014, 02:25:15 PM »

Hello Mark,

The easiest way is to load images taken from different cameras to different chunk and then to apply background masking for each chunk separately. Then just drag and drop images to the single chunk in the Workspace pane.

More complicated way is to write short script to perform the operation.

13189

General / Re: Report

« on: March 25, 2014, 02:05:32 PM »

Hello ristag,

Projections is a number of used tie point projections on the original photos.
Error (in pix) - mean reprojection error.

And what additional information are you interested in?

13190

Feature Requests / Re: Kinect Data Stream Integration

« on: March 24, 2014, 03:22:12 PM »

Hello Jeff,

Currently we are not planning such feature. Even it seems to be possible in principle we are not sure that this feature is demanded and will be easy to operate.

To work both with Kinect and regular cameras you'll need to align the camera positions from both sets in the same space.

13191

Face and Body Scanning / Re: Background masking process

« on: March 24, 2014, 03:06:59 PM »

Hello Mark,

If you have the folder with the background images you can use template for import mask filename like {filename}_background.{fileext} - such option will help PhotoScan to find, lets say, 01_01_background.cr2 background image for 01_01.cr2 file in the user defined folder (if Apply to chunk option is used).

Another unrelated comment is to convert CR2 files to TIFFs to prevent any color issues, since RAW images are not officially supported by PhotoScan.

13192

Other Languages / Re: Экспорт ортофото больших размеров

« on: March 22, 2014, 11:59:33 PM »

Я имел в виду опцию нарезки на блоки в окне экспорта ортофото. Но это, конечно, работает только для единой сцены.

13193

Other Languages / Re: Экспорт ортофото больших размеров

« on: March 21, 2014, 02:02:55 PM »

Здравствуйте, Александр,

А функция нарезки ортофото на блоки Вам не подходит?

13194

General / Re: Export Undistort Photos

« on: March 21, 2014, 12:36:53 PM »

Hello bmc130,

You can use the following template for export: {filename}.{fileext}

13195

Feature Requests / Re: Instance chunks to seperate files but load into one master Photoscan file?

« on: March 20, 2014, 02:49:56 PM »

Hello,

We are planning to implement similar functionality, but at the moment we are working on different features for the next major update. So I can say that discussed feature should not be expected in the version 1.1, but the idea is not lost for sure.

13196

General / Re: Photoscan Standard - object orientation?

« on: March 18, 2014, 07:41:53 PM »

Hello bcoyle,

The rotation is applied at the same moment it is performed in the Model view. Could you please check if this problem is reproducible only for FBX or for other formats too (like OBJ, for example)?

13197

General / Re: GCP error low - high errors on DEM

« on: March 18, 2014, 01:49:35 PM »

Hello stihl,

Apart from Marker accuracy parameter, Projection accuracy is also used. So by increasing projection accuracy close to zero you'll get errors on marker location also closer to zero. But you should also note that in some cases camera distortion model used could not estimate camera orientations so that all markers fit their positions precisely.

If you wish you can send us the project file with the aligned cameras and markers placed, so that we could try different settings of optimization.

13198

Other Languages / Re: Фотосъемка

« on: March 18, 2014, 01:47:39 PM »

Если Вы производите съёмку со штатива, то можете воспользоваться функцией автоматического создания масок по пустому кадру без объекта, сделанному с той же точки съёмки и теми же параметрами (Меню Инструменты -> Импортировать маски -> Из фона).

Можете отправить нам набор, с которым Вы работаете, чтобы мы смогли более детально просмотреть кадры и дать дополнительные рекомендации на основе конкретного набора.

Возможно, тьюториал по обработке Вам поможет разобраться с основными параметрами реконструкции: http://www.agisoft.ru/tutorials/photoscan/07/

13199

Other Languages / Re: Фотосъемка

« on: March 18, 2014, 12:23:46 PM »

Добрый день,

Такой сценарий съёмки возможен, но при этом следует использовать маски для закрытия областей фона, поскольку объект движется относительно него. В остальном процесс обработки не отличается от стандартного.

Однако, поясните, пожалуйста, о каких параметрах X и Y идёт речь?

13200

General / Re: Importing georeferenced .obj model to Meshlab

« on: March 18, 2014, 01:03:30 AM »

Hello Edo,

If using projected coordinate system, please use Shift fields for X and Y coordinates in Export model dialog to subtract large numbers from the vertices' coordinates.
Note that models with degrees as units will not be displayed properly in MeshLab due to different scale along X and Y axis.