Messages

16

General / Re: Agisoft PhotoScan 1.2.0 pre-release

« on: August 09, 2015, 04:53:48 PM »

I might... can't really remember since I was testing. If I removed something I am sure it is output files. Is it possible to change some kind of .xml so it doesnt read the orthomosaic data? I was wondering if the project must be started from scratch or rescued.


Any notes on the pixel accuracy values (reference tab) or why the blending process in texture generation is taking so long now?
Thank you for the changes! Agisoft really feels like a vivid creature evolving

17

General / Re: Agisoft PhotoScan 1.2.0 pre-release

« on: August 09, 2015, 04:01:26 PM »

I might have deleted that file.
I was playing with different outputs like orthomosaic, a feature I've never used in head scans.

18

General / Re: Agisoft PhotoScan 1.2.0 pre-release

« on: August 09, 2015, 02:08:46 PM »

Laplacian looks promising! As I've observed its strength is dependant on the number of polygons?
Best,

Manuel.

19

General / Re: Zbrush - Workflow

« on: August 09, 2015, 01:24:25 PM »

Could someone suggest a proper workflow to close holes caused by messy and dense hair ? It usually happens on the top of my head scans..

I was wondering if zBrush has a brush similar to photoshop's clone brush, so I can copy the surrounding topology with the same polygon density... I suspect that, before anything, I should perform a "close holes" but it wont close with the same density.

Anyway I am sure there are many zBrush's professionals here  

Thank you!

20

General / Re: Agisoft PhotoScan 1.2.0 pre-release

« on: August 09, 2015, 11:46:59 AM »

Great! Yes it is a .psx file with the filename.files

I've sent them via WeTransfer (it'll take about 3 hours though, my uploading speed is crap...) 
I've also noticed that after optimizing the sparse point cloud the pixel accuracy seems to have way lower number than before on the new version. Has something changed ? Texture blending time also seems to take more... Maybe it is just this project, I could try with older ones.

21

General / Re: Agisoft PhotoScan 1.2.0 pre-release

« on: August 08, 2015, 06:47:02 PM »

Maybe Bilateral would've been better! Although whats important is to take the first step! Well done! It was a feature hugely requested. As a suggestion I find the laplacian filter to be "too strong" even just applying one pass.

I've been playing with the new version at it seems like the project file gets corrupted very easily... I've sent a few times the error report. I keep trying to open the project file but it keeps crashing before opening the model. It wasnt corrupted until the last stage of exporting the model.

Best.

22

General / Re: Agisoft PhotoScan 1.2.0 pre-release

« on: August 07, 2015, 03:12:26 PM »

• Added Smooth Model command.

Wow! How does your implementation work? is it just Gaussian based? Bilateral?

Great news!

23

General / Re: Zbrush - Workflow

« on: July 03, 2015, 12:19:34 AM »

Thank you Lee.

You've just saved me a lot of time!!
I'll definitely send an email to Pixologic 

24

General / Re: Zbrush - Workflow

« on: July 02, 2015, 10:08:12 PM »

Could someone explain this to me ? Its driving me crazy!
It keeps happening 50% of the time. Somehow the model is being moved and scaled by zBrush. The funny thing is that I never touch the move tool or the scale tool. Could it be related to zRemeshing or projecting details from one subtool to the other? ...  So many hours of work lost!

Thank you all,
Manuel Franquelo.

PD: Thank you Alexey!

25

General / Re: Correcting model's axis

« on: June 11, 2015, 03:29:56 PM »

Yeah sorry, i've just got into the Wiki for the PY scripts and found the appropiate one for 1.1
Its this one: -i meant match the bounding box axis, not copy it-

#rotates model coordinate system in accordance of bounding box for active chunk
#scale is kept
#compatibility: Agisoft PhotoScan Professional 1.1.0

import PhotoScan
import math

doc = PhotoScan.app.document
chunk = doc.chunk

R = chunk.region.rot      #Bounding box rotation matrix
C = chunk.region.center      #Bounding box center vector

if chunk.transform.matrix:
   T = chunk.transform.matrix
   s = math.sqrt(T[0,0] ** 2 + T[0,1] ** 2 + T[0,2] ** 2)       #scaling
   S = PhotoScan.Matrix( [[s, 0, 0, 0], [0, s, 0, 0], [0, 0, s, 0], [0, 0, 0, 1]] ) #scale matrix
else:
   S = PhotoScan.Matrix( [[1, 0, 0, 0], [0, 1, 0, 0], [0, 0, 1, 0], [0, 0, 0, 1]] )

T = PhotoScan.Matrix( [[R[0,0], R[0,1], R[0,2], C[0]], [R[1,0], R[1,1], R[1,2], C[1]], [R[2,0], R[2,1], R[2,2], C[2]], [0, 0, 0, 1]])

chunk.transform.matrix = S * T.inv()      #resulting chunk transformation matrix

Thanks!

26

General / Re: Correcting model's axis

« on: June 11, 2015, 03:12:22 PM »

Thank you Alex,

This is the scrypt:

#scale is kept
#compatibility: Agisoft PhotoScan Professional 0.9.0


import PhotoScan
import math

doc = PhotoScan.app.document

chunk = doc.activeChunk

R = chunk.region.rot      #Bounding box rotation matrix
C = chunk.region.center      #Bounding box center vector

if chunk.transform:
   T = chunk.transform
   s = math.sqrt(T[0,0]*T[0,0] + T[0,1]*T[0,1] + T[0,2]*T[0,2])       #scaling
   S = PhotoScan.Matrix( [[s, 0, 0, 0], [0, s, 0, 0], [0, 0, s, 0], [0, 0, 0, 1]] ) #scale matrix
else:
   S = PhotoScan.Matrix( [[1, 0, 0, 0], [0, 1, 0, 0], [0, 0, 1, 0], [0, 0, 0, 1]] )

T = PhotoScan.Matrix( [[R[0,0],R[0,1],R[0,2], C[0]], [R[1,0],R[1,1],R[1,2],C[1]], [R[2,0],R[2,1],R[2,2],C[2]], [0,0,0,1]])

chunk.transform = S * T.inv()      #resulting chunk transformation matrix

27

General / Correcting model's axis

« on: June 11, 2015, 03:06:59 PM »

Dear all,

I can´t seem to be able to correct the model's axis. I am trying to do a turntable movie on zBrush -as of now i didn´t need to correct the axis- but it seems that zBrush uses the model's axis and not the global zBrush axis to rotate the model across time.

I dont have references to create scale bars for this particular model... so i can´t scale it nor orientate it.
I've tried using a .py script that i've found to match the bounding box axis. But it doesnt seem to work...

Any ideas? :-) 

28

General / Re: Canon EOS 5DS R 50MP

« on: March 02, 2015, 11:43:51 PM »

Sigma 50mm ART lens will do the job

29

Feature Requests / Re: high frequency noise reduction

« on: February 01, 2015, 03:44:35 PM »

+1 I can´t seem to find any other denoising software worth the results...
I've tried meshlab but i can´t seem to get the results i want using Taubin...
There are many surface-preserving denoise algorithms out there, im sure agisoft super developing team can implement one of them! 

30

General / Re: Can´t build dense cloud

« on: January 29, 2015, 03:22:05 PM »

Great!
Thank you Alexey.