Changes between Version 30 and Version 31 of Tutorial/CorrelationImageVelocimetryOptimisation


Ignore:
Timestamp:
Feb 2, 2015, 10:11:28 AM (6 years ago)
Author:
vaillant1p
Comment:

--

Legend:

Unmodified
Added
Removed
Modified
  • Tutorial/CorrelationImageVelocimetryOptimisation

    v30 v31  
    6464
    6565== Example of an intense vortex ==
    66 With the GUI '''uvmat''' open an image in '/UVMAT_DEMO04_PIVchallenge_2001A/images' taken from PIV challenge ([http://www.pivchallenge.org/ see website]). A short description of the experimental conditions is given in the file 'readme.txt' in the folder of the images. It is said that the field of view is 170 x140 mm, so that, since the image size is 1280x1024 px, the scale is 7.4 px/cm. but no time Dt **. The flow is the vortex behind a wing with poor particle seeding in the core.
     66With the GUI '''uvmat''' open an image in '/UVMAT_DEMO04_PIVchallenge_2001A/images' taken from PIV challenge ([http://www.pivchallenge.org/ see website]). A short description of the experimental conditions is given in the file 'readme.txt' in the folder of the images. It is said that the field of view is 170 x140 mm, so that, since the image size is 1280x1024 px, the scale is 7.4 px/cm. but there is no information about the time interval Dt.
    6767
    68 From '''uvmat''' select '''[Run/PIV]''' so the GUI '''civ_input''' pops up. With the option''' [!ImportParam]''', import the parameters from the file 'Images.ref.civ/0_XML/A_1_1.xml' and '''[RUN]''' the civ calculation. The result should look like the reference result 'Image.ref.civ/A_1_1-2.nc'. As expected form the dark spot at the vortex centre, the core is not properly resolved in civ1. The result is improved in civ2.
     68The flow is the vortex behind a wing with poor particle seeding in the core.
    6969
    70 Let us improve again the result by a third civ iteration. This is launched in the GUI '''series/civ_series''', opening  the netcdf file result of the previous Civ2 calculation instead of an image. The source image then automatically appears in the second line of the input table of '''series'''. Unselect the option '''[CIV1]''', '''[FIX1]''' and '''[PATCH1]''', and select the check box '''[CIV3]''' in the frame Civ2.''' '''Then run the civ calculation, with result in a new folder 'Images.civ_1'. The final result is shown in the figure below, with a cut across the vortex core at y=500 pixel.
     70From '''uvmat''' select '''[Run/PIV]''' so the GUI '''civ_input''' pops up. With the option''' [!ImportParam]''', import the parameters from the file 'Images.ref.civ/0_XML/frame_1.xml', select the mask 'mask_1.png' for '''[CIV1]''' and ''''[CIV2]''' in 'Images.ref.mask' and '''[RUN]''' the civ calculation on 'frame_1.png' only. The result should look like the reference result 'Images.ref.civ/frame_1-2.nc'. As expected from the dark spot at the vortex centre, the core is not properly resolved in civ1. The result is improved in civ2.
     71
     72Let us improve again the result by a third civ iteration. This is launched in the GUI '''series''', opening  the netcdf file result of the previous Civ2 calculation instead of an image. The source image then automatically appears in the second line of the input table of '''series'''. Unselect the option '''[CIV1]''', '''[FIX1]''' and '''[PATCH1]''', and select the check box '''[CIV3]''' in the frame '''[CIV2]'''. Choose the mask 'mask_1.png' and then '''[RUN]''' the civ calculation, saving the result in a new folder 'Images.civ_1'. The final result is shown in the figure below, with a cut across the vortex core at y=500 pixel.
    7173
    7274figure**