Tiff file format for pre-processed quad-stereo sets
← Older revision
Revision as of 19:19, 13 July 2018
(One intermediate revision by the same user not shown)Line 45:
Line 45:
==<font color='ForestGreen'>TIFF stacks for ML</font>== ==<font color='ForestGreen'>TIFF stacks for ML</font>==
−[[File:Ml tiff stack imagej.jpeg|thumb|400px|Fig.1 Opened in ImageJ]]+[[File:Ml tiff stack imagej.jpeg|thumb|400px|Fig.1 Opened in ImageJ, one of the layers]]
[[File:Ml tiff stack python values.jpeg|thumb|800px|Fig.2 Decoded 'value'-layers]] [[File:Ml tiff stack python values.jpeg|thumb|800px|Fig.2 Decoded 'value'-layers]]
[[File:Ml tiff stack python data layers.jpeg|thumb|500px|Fig.3 Data layers]] [[File:Ml tiff stack python data layers.jpeg|thumb|500px|Fig.3 Data layers]]
Oleg
Tiff file format for pre-processed quad-stereo sets
Python
← Older revision Revision as of 00:49, 13 July 2018 Line 110: Line 110: values data: (242, 324, 3)</font> values data: (242, 324, 3)</font> −[[File:Ml tiff stack python tile packing.jpeg|thumb|400px|Fig.4 Single layer packing]]+[[File:Ml tiff stack python tile packing.jpeg|thumb|400px|Fig.4 Single layer packing (actually, see the CVPR2018 presentation, p.23)]] '''2.''' Use '''pack_tile.py''' from [https://git.elphel.com/Elphel/python3-imagej-tiff python3-imagej-tiff] to: '''2.''' Use '''pack_tile.py''' from [https://git.elphel.com/Elphel/python3-imagej-tiff python3-imagej-tiff] to: * <font color='DarkGreen'>'''Pack image tiles for better structuring for ML'''</font> * <font color='DarkGreen'>'''Pack image tiles for better structuring for ML'''</font> OlegTiff file format for pre-processed quad-stereo sets
Python
← Older revision Revision as of 00:49, 13 July 2018 (One intermediate revision by the same user not shown)Line 110: Line 110: values data: (242, 324, 3)</font> values data: (242, 324, 3)</font> − +[[File:Ml tiff stack python tile packing.jpeg|thumb|400px|Fig.4 Single layer packing (actually, see the CVPR2018 presentation, p.23)]] '''2.''' Use '''pack_tile.py''' from [https://git.elphel.com/Elphel/python3-imagej-tiff python3-imagej-tiff] to: '''2.''' Use '''pack_tile.py''' from [https://git.elphel.com/Elphel/python3-imagej-tiff python3-imagej-tiff] to: * <font color='DarkGreen'>'''Pack image tiles for better structuring for ML'''</font> * <font color='DarkGreen'>'''Pack image tiles for better structuring for ML'''</font> OlegTiff file format for pre-processed quad-stereo sets
Python
← Older revision Revision as of 20:20, 12 July 2018 Line 110: Line 110: values data: (242, 324, 3)</font> values data: (242, 324, 3)</font> − +[[File:Ml tiff stack python tile packing.jpeg|thumb|400px|Fig.4 Single layer packing]] '''2.''' Use '''pack_tile.py''' from [https://git.elphel.com/Elphel/python3-imagej-tiff python3-imagej-tiff] to: '''2.''' Use '''pack_tile.py''' from [https://git.elphel.com/Elphel/python3-imagej-tiff python3-imagej-tiff] to: * <font color='DarkGreen'>'''Pack image tiles for better structuring for ML'''</font> * <font color='DarkGreen'>'''Pack image tiles for better structuring for ML'''</font> OlegFile:Ml tiff stack python values.jpeg
Oleg uploaded a new version of File:Ml tiff stack python values.jpeg
New page
== Licensing =={{CC }} Oleg
Tiff file format for pre-processed quad-stereo sets
← Older revision
Revision as of 20:08, 12 July 2018
(2 intermediate revisions by the same user not shown)Line 60:
Line 60:
ImageJ - it has a native support for stacks, each stack has a name label stored (along with related xml info) in the ImageJ tiff tags. To read tiff tags in ImageJ, go '''Image > Show Info...''' ImageJ - it has a native support for stacks, each stack has a name label stored (along with related xml info) in the ImageJ tiff tags. To read tiff tags in ImageJ, go '''Image > Show Info...'''
====Python==== ====Python====
−Use '''imagej_tiff.py''' from [https://git.elphel.com/Elphel/python3-imagej-tiff python3-imagej-tiff] to:+'''1.''' Use '''imagej_tiff.py''' from [https://git.elphel.com/Elphel/python3-imagej-tiff python3-imagej-tiff] to:
* <font color='DarkGreen'>'''Get tiff tags values (Pillow)'''</font> * <font color='DarkGreen'>'''Get tiff tags values (Pillow)'''</font>
* <font color='DarkGreen'>'''Parse Properties xml data stored by ImageJ in the tiff tags'''</font> * <font color='DarkGreen'>'''Parse Properties xml data stored by ImageJ in the tiff tags'''</font>
Line 109:
Line 109:
image data: (242, 324, 9, 9, 4) image data: (242, 324, 9, 9, 4)
values data: (242, 324, 3)</font> values data: (242, 324, 3)</font>
+
+
+'''2.''' Use '''pack_tile.py''' from [https://git.elphel.com/Elphel/python3-imagej-tiff python3-imagej-tiff] to:
+* <font color='DarkGreen'>'''Pack image tiles for better structuring for ML'''</font>
+ <font size=2>'''# The current packing is hardcoded in the script: (TY,TX,9,9,4) -> (TY,TX,100)'''
+ # Get tiles from a tiff stack
+ import pack_tile as pile
+ packed_tiles = pile.pack(tile)
+ # packed_tiles get stacked with '''estimated disparity''' (see '''test_nn_feed.py''')
+ packed_tiles = np.dstack((packed_tiles,values[:,:,0]))
Oleg
Tiff file format for pre-processed quad-stereo sets
← Older revision
Revision as of 20:08, 12 July 2018
(4 intermediate revisions by the same user not shown)Line 47:
Line 47:
[[File:Ml tiff stack imagej.jpeg|thumb|400px|Fig.1 Opened in ImageJ]] [[File:Ml tiff stack imagej.jpeg|thumb|400px|Fig.1 Opened in ImageJ]]
[[File:Ml tiff stack python values.jpeg|thumb|800px|Fig.2 Decoded 'value'-layers]] [[File:Ml tiff stack python values.jpeg|thumb|800px|Fig.2 Decoded 'value'-layers]]
+[[File:Ml tiff stack python data layers.jpeg|thumb|500px|Fig.3 Data layers]]
* What's in each stack is described in the [https://community.elphel.com/files/presentations/Elphel_TP-CNN_slides.pdf presentation for CVPR2018, pp.19-21]: * What's in each stack is described in the [https://community.elphel.com/files/presentations/Elphel_TP-CNN_slides.pdf presentation for CVPR2018, pp.19-21]:
** 5 layers in the stack: ['''diagm-pair''', '''diago-pair''', '''hor-pairs''', '''vert-pairs''', '''other'''] ** 5 layers in the stack: ['''diagm-pair''', '''diago-pair''', '''hor-pairs''', '''vert-pairs''', '''other''']
Line 59:
Line 60:
ImageJ - it has a native support for stacks, each stack has a name label stored (along with related xml info) in the ImageJ tiff tags. To read tiff tags in ImageJ, go '''Image > Show Info...''' ImageJ - it has a native support for stacks, each stack has a name label stored (along with related xml info) in the ImageJ tiff tags. To read tiff tags in ImageJ, go '''Image > Show Info...'''
====Python==== ====Python====
−Use '''imagej_tiff.py''' from [https://git.elphel.com/Elphel/python3-imagej-tiff python3-imagej-tiff] to:+'''1.''' Use '''imagej_tiff.py''' from [https://git.elphel.com/Elphel/python3-imagej-tiff python3-imagej-tiff] to:
* <font color='DarkGreen'>'''Get tiff tags values (Pillow)'''</font> * <font color='DarkGreen'>'''Get tiff tags values (Pillow)'''</font>
* <font color='DarkGreen'>'''Parse Properties xml data stored by ImageJ in the tiff tags'''</font> * <font color='DarkGreen'>'''Parse Properties xml data stored by ImageJ in the tiff tags'''</font>
Line 104:
Line 105:
−Upon opening tiff the image shape will be '''(height,width,layer)''' - for further processing it needs to be reshaped to '''(height_in_tiles, width_in_tiles, tile_height, tile_width, layer)''', example:+Upon opening tiff the image will be a numpy array of shape '''(height,width,layer)''' - for further processing it is reshaped to '''(height_in_tiles, width_in_tiles, tile_height, tile_width, layer)''', example:
<font size='2'>tiff stack shape: (2178, 2916, 5) <font size='2'>tiff stack shape: (2178, 2916, 5)
image data: (242, 324, 9, 9, 4) image data: (242, 324, 9, 9, 4)
values data: (242, 324, 3)</font> values data: (242, 324, 3)</font>
+
+
+'''2.''' Use '''pack_tile.py''' from [https://git.elphel.com/Elphel/python3-imagej-tiff python3-imagej-tiff] to:
+* <font color='DarkGreen'>'''Pack image tiles for better structuring for ML'''</font>
+ <font size=2>'''# The current packing is hardcoded in the script: (TY,TX,9,9,4) -> (TY,TX,100)'''
+ # Get tiles from a tiff stack
+ import pack_tile as pile
+ packed_tiles = pile.pack(tile)
+ # packed_tiles get stacked with '''estimated disparity''' (see '''test_nn_feed.py''')
+ packed_tiles = np.dstack((packed_tiles,values[:,:,0]))
Oleg
Tiff file format for pre-processed quad-stereo sets
← Older revision
Revision as of 23:19, 11 July 2018
(One intermediate revision by the same user not shown)Line 47:
Line 47:
[[File:Ml tiff stack imagej.jpeg|thumb|400px|Fig.1 Opened in ImageJ]] [[File:Ml tiff stack imagej.jpeg|thumb|400px|Fig.1 Opened in ImageJ]]
[[File:Ml tiff stack python values.jpeg|thumb|800px|Fig.2 Decoded 'value'-layers]] [[File:Ml tiff stack python values.jpeg|thumb|800px|Fig.2 Decoded 'value'-layers]]
+[[File:Ml tiff stack python data layers.jpeg|thumb|500px|Fig.3 Data layers]]
* What's in each stack is described in the [https://community.elphel.com/files/presentations/Elphel_TP-CNN_slides.pdf presentation for CVPR2018, pp.19-21]: * What's in each stack is described in the [https://community.elphel.com/files/presentations/Elphel_TP-CNN_slides.pdf presentation for CVPR2018, pp.19-21]:
** 5 layers in the stack: ['''diagm-pair''', '''diago-pair''', '''hor-pairs''', '''vert-pairs''', '''other'''] ** 5 layers in the stack: ['''diagm-pair''', '''diago-pair''', '''hor-pairs''', '''vert-pairs''', '''other''']
Line 104:
Line 105:
−Upon opening tiff the image shape will be '''(height,width,layer)''' - for further processing it needs to be reshaped to '''(height_in_tiles, width_in_tiles, tile_height, tile_width, layer)''', example:+Upon opening tiff the image will be a numpy array of shape '''(height,width,layer)''' - for further processing it is reshaped to '''(height_in_tiles, width_in_tiles, tile_height, tile_width, layer)''', example:
<font size='2'>tiff stack shape: (2178, 2916, 5) <font size='2'>tiff stack shape: (2178, 2916, 5)
image data: (242, 324, 9, 9, 4) image data: (242, 324, 9, 9, 4)
values data: (242, 324, 3)</font> values data: (242, 324, 3)</font>
Oleg
Tiff file format for pre-processed quad-stereo sets
TIFF stacks for ML
← Older revision Revision as of 22:50, 11 July 2018 Line 47: Line 47: [[File:Ml tiff stack imagej.jpeg|thumb|400px|Fig.1 Opened in ImageJ]] [[File:Ml tiff stack imagej.jpeg|thumb|400px|Fig.1 Opened in ImageJ]] [[File:Ml tiff stack python values.jpeg|thumb|800px|Fig.2 Decoded 'value'-layers]] [[File:Ml tiff stack python values.jpeg|thumb|800px|Fig.2 Decoded 'value'-layers]] +[[File:Ml tiff stack python data layers.jpeg|thumb|500px|Fig.3 Data layers]] * What's in each stack is described in the [https://community.elphel.com/files/presentations/Elphel_TP-CNN_slides.pdf presentation for CVPR2018, pp.19-21]: * What's in each stack is described in the [https://community.elphel.com/files/presentations/Elphel_TP-CNN_slides.pdf presentation for CVPR2018, pp.19-21]: ** 5 layers in the stack: ['''diagm-pair''', '''diago-pair''', '''hor-pairs''', '''vert-pairs''', '''other'''] ** 5 layers in the stack: ['''diagm-pair''', '''diago-pair''', '''hor-pairs''', '''vert-pairs''', '''other'''] OlegTiff file format for pre-processed quad-stereo sets
Python
← Older revision Revision as of 22:44, 11 July 2018 (4 intermediate revisions by the same user not shown)Line 45: Line 45: ==<font color='ForestGreen'>TIFF stacks for ML</font>== ==<font color='ForestGreen'>TIFF stacks for ML</font>== +[[File:Ml tiff stack imagej.jpeg|thumb|400px|Fig.1 Opened in ImageJ]] +[[File:Ml tiff stack python values.jpeg|thumb|800px|Fig.2 Decoded 'value'-layers]] * What's in each stack is described in the [https://community.elphel.com/files/presentations/Elphel_TP-CNN_slides.pdf presentation for CVPR2018, pp.19-21]: * What's in each stack is described in the [https://community.elphel.com/files/presentations/Elphel_TP-CNN_slides.pdf presentation for CVPR2018, pp.19-21]: ** 5 layers in the stack: ['''diagm-pair''', '''diago-pair''', '''hor-pairs''', '''vert-pairs''', '''other'''] ** 5 layers in the stack: ['''diagm-pair''', '''diago-pair''', '''hor-pairs''', '''vert-pairs''', '''other'''] Line 58: Line 60: ====Python==== ====Python==== Use '''imagej_tiff.py''' from [https://git.elphel.com/Elphel/python3-imagej-tiff python3-imagej-tiff] to: Use '''imagej_tiff.py''' from [https://git.elphel.com/Elphel/python3-imagej-tiff python3-imagej-tiff] to: −* get tiff tags values (Pillow)+* <font color='DarkGreen'>'''Get tiff tags values (Pillow)'''</font> −* parse Properties xml data stored in the tiff tags used by ImageJ+* <font color='DarkGreen'>'''Parse Properties xml data stored by ImageJ in the tiff tags'''</font> −* get tile dimensions from Properties+* <font color='DarkGreen'>'''Get tiles dimensions from Properties'''</font> −* read layers as numpy arrays for further computations or plotting+* <font color='DarkGreen'>'''Convert layers to numpy arrays (split into tiles or optionally not split) for further computations and plotting'''</font> −Example:+Example (more examples are commented out in the __main__ section of the script): <font size='1' style='line-height:0.5;'>'''~$ python3 imagej_tiff.py 1527256903_350165-ML_DATA-08B-O-FZ0.05-OFFS0.00000.tiff''' <font size='1' style='line-height:0.5;'>'''~$ python3 imagej_tiff.py 1527256903_350165-ML_DATA-08B-O-FZ0.05-OFFS0.00000.tiff''' time: 1531344391.7055812 time: 1531344391.7055812 OlegTiff file format for pre-processed quad-stereo sets
TIFF stacks for ML
← Older revision Revision as of 21:48, 11 July 2018 (15 intermediate revisions by the same user not shown)Line 3: Line 3: <font size='2'>'''1527256903_350165/''' <font size='2'>'''1527256903_350165/''' ├── 1527256903_350165.kml ├── 1527256903_350165.kml − ├── '''jp4'''+ ├── '''jp4''' (source files directory) │ ├── <font color='RoyalBlue'>1527256903_350165_0.jp4</font> │ ├── <font color='RoyalBlue'>1527256903_350165_0.jp4</font> │ ├── <font color='RoyalBlue'>...</font> │ ├── <font color='RoyalBlue'>...</font> Line 12: Line 12: ├── rating.txt ├── rating.txt ├── thumb.jpeg ├── thumb.jpeg − └── '''v03'''+ └── '''v03''' (model version) ├── <font color='Indigo'>1527256903_350165-00-D0.0.jpeg</font> ├── <font color='Indigo'>1527256903_350165-00-D0.0.jpeg</font> ├── <font color='Indigo'>...</font> ├── <font color='Indigo'>...</font> Line 26: Line 26: ├── <font color='DarkGoldenrod'>1527256903_350165.obj</font> ├── <font color='DarkGoldenrod'>1527256903_350165.obj</font> ├── <font color='DarkGoldenrod'>1527256903_350165.x3d</font> ├── <font color='DarkGoldenrod'>1527256903_350165.x3d</font> − └── '''ml'''+ └── '''ml''' (directory with processed data files for ML) ├── <font color='ForestGreen'>1527256903_350165-ML_DATA-08B-O-FZ0.05-OFFS-2.00000.tiff</font> ├── <font color='ForestGreen'>1527256903_350165-ML_DATA-08B-O-FZ0.05-OFFS-2.00000.tiff</font> ├── <font color='ForestGreen'>1527256903_350165-ML_DATA-08B-O-FZ0.05-OFFS-1.00000.tiff</font> ├── <font color='ForestGreen'>1527256903_350165-ML_DATA-08B-O-FZ0.05-OFFS-1.00000.tiff</font> Line 34: Line 34: where: where: * <font color='RoyalBlue'>'''*.jp4'''</font> - source files, 0..3 - quad stereo camera #1, 4..7 - quad stereo camera #2, and so on if there are more cameras in the system. * <font color='RoyalBlue'>'''*.jp4'''</font> - source files, 0..3 - quad stereo camera #1, 4..7 - quad stereo camera #2, and so on if there are more cameras in the system. −* <font color='Indigo'>'''*-D0.0.jpeg'''</font> - disparity = 0, images are undistorted to a common polynom+* <font color='Indigo'>'''*-D0.0.jpeg'''</font> - disparity = 0, images are undistorted to a distortion polynom common for each image * <font color='OrangeRed'>'''*.corr-xml'''</font> - ImageJ plugin's settings file? * <font color='OrangeRed'>'''*.corr-xml'''</font> - ImageJ plugin's settings file? * <font color='Maroon'>'''*-DSI_COMBO.tiff'''</font> - Disparity Space Image - tiff stack * <font color='Maroon'>'''*-DSI_COMBO.tiff'''</font> - Disparity Space Image - tiff stack Line 40: Line 40: * <font color='DarkGoldenrod'>'''*.x3d, *.png'''</font> - X3D format model with textures. The textures are shared with the OBJ format model * <font color='DarkGoldenrod'>'''*.x3d, *.png'''</font> - X3D format model with textures. The textures are shared with the OBJ format model * <font color='DarkGoldenrod'>'''*.obj, *.mtl, *.png'''</font> - OBJ format model with textures * <font color='DarkGoldenrod'>'''*.obj, *.mtl, *.png'''</font> - OBJ format model with textures −* <font color='ForestGreen'>'''ml/*.tiff'''</font> - TIFF stack of pre-processed images for ML+* <font color='ForestGreen'>'''*.tiff'''</font> - TIFF stack of pre-processed images for ML * *.kml, rating.txt, thumb.jpeg - files, related to the online viewer only * *.kml, rating.txt, thumb.jpeg - files, related to the online viewer only −==Stacked TIFF==+==<font color='ForestGreen'>TIFF stacks for ML</font>== − +* What's in each stack is described in the [https://community.elphel.com/files/presentations/Elphel_TP-CNN_slides.pdf presentation for CVPR2018, pp.19-21]: +** 5 layers in the stack: ['''diagm-pair''', '''diago-pair''', '''hor-pairs''', '''vert-pairs''', '''other'''] +*** '''diagm-pair''' +*** '''diago-pair''' +*** '''hor-pairs''' +*** '''vert-pairs''' +*** '''other''' - encoded values: estimated disparity, residual disparity and confidence for the residual disparity * The source files are processed using a plugin for ImageJ, the output file for each set is a tiff stack * The source files are processed using a plugin for ImageJ, the output file for each set is a tiff stack * There are a few ways to view the stack: * There are a few ways to view the stack: −** just open with ImageJ+====ImageJ==== +ImageJ - it has a native support for stacks, each stack has a name label stored (along with related xml info) in the ImageJ tiff tags. To read tiff tags in ImageJ, go '''Image > Show Info...''' +====Python==== +Use '''imagej_tiff.py''' from [https://git.elphel.com/Elphel/python3-imagej-tiff python3-imagej-tiff] to: +* get tiff tags values (Pillow) +* parse Properties xml data stored in the tiff tags used by ImageJ +* get tile dimensions from Properties +* read layers as numpy arrays for further computations or plotting +Example: + <font size='1' style='line-height:0.5;'>'''~$ python3 imagej_tiff.py 1527256903_350165-ML_DATA-08B-O-FZ0.05-OFFS0.00000.tiff''' + time: 1531344391.7055812 + time: 1531344392.5336654 + TIFF stack labels: ['diagm-pair', 'diago-pair', 'hor-pairs', 'vert-pairs', 'other'] + <?xml version="1.0" ?> + <properties> + <ML_OTHER_TARGET>0</ML_OTHER_TARGET> + <tileWidth>9</tileWidth> + <disparityRadiusMain>257.22231560274076</disparityRadiusMain> + <comment_ML_OTHER_GTRUTH_STRENGTH>Offset of the ground truth strength in the "other" layer tile</comment_ML_OTHER_GTRUTH_STRENGTH> + <data_min>-0.16894744988183344</data_min> + <comment_intercameraBaseline>Horizontal distance between the main and the auxiliary camera centers (mm). Disparity is specified for the main camera</comment_intercameraBaseline> + <ML_OTHER_GTRUTH>2</ML_OTHER_GTRUTH> + <data_max>0.6260986600450271</data_max> + <disparityRadiusAux>151.5308819757923</disparityRadiusAux> + <comment_disparityRadiusAux>Side of the square where 4 main camera subcameras are located (mm). Disparity is specified for the main camera</comment_disparityRadiusAux> + <comment_disparityRadiusMain>Side of the square where 4 main camera subcameras are located (mm)</comment_disparityRadiusMain> + <comment_dispOffset>Tile target disparity minum ground truth disparity</comment_dispOffset> + <comment_tileWidth>Square tile size for each 2d correlation, always odd</comment_tileWidth> + <comment_data_min>Defined only for 8bpp mode - value, corresponding to -127 (-128 is NaN)</comment_data_min> + <comment_data_max>Defined only for 8bpp mode - value, corresponding to +127 (-128 is NaN)</comment_data_max> + <comment_ML_OTHER_TARGET>Offset of the target disparity in the "other" layer tile</comment_ML_OTHER_TARGET> + <VERSION>1.0</VERSION> + <dispOffset>0.0</dispOffset> + <comment_ML_OTHER_GTRUTH>Offset of the ground truth disparity in the "other" layer tile</comment_ML_OTHER_GTRUTH> + <ML_OTHER_GTRUTH_STRENGTH>4</ML_OTHER_GTRUTH_STRENGTH> + <intercameraBaseline>1256.0</intercameraBaseline> + </properties> + Tiles shape: 9x9 + Data min: -0.16894744988183344 + Data max: 0.6260986600450271 + (2178, 2916, 5) + Stack of images shape: (242, 324, 9, 9, 4) + time: 1531344392.7290232 + Stack of values shape: (242, 324, 3) + time: 1531344393.5556033</font> + + + +Upon opening tiff the image shape will be '''(height,width,layer)''' - for further processing it needs to be reshaped to '''(height_in_tiles, width_in_tiles, tile_height, tile_width, layer)''', example: + <font size='2'>tiff stack shape: (2178, 2916, 5) + image data: (242, 324, 9, 9, 4) + values data: (242, 324, 3)</font> OlegTiff file format for pre-processed quad-stereo sets
TIFF stacks for ML
← Older revision Revision as of 20:42, 11 July 2018 (12 intermediate revisions by the same user not shown)Line 3: Line 3: <font size='2'>'''1527256903_350165/''' <font size='2'>'''1527256903_350165/''' ├── 1527256903_350165.kml ├── 1527256903_350165.kml − ├── '''jp4'''+ ├── '''jp4''' (source files directory) │ ├── <font color='RoyalBlue'>1527256903_350165_0.jp4</font> │ ├── <font color='RoyalBlue'>1527256903_350165_0.jp4</font> │ ├── <font color='RoyalBlue'>...</font> │ ├── <font color='RoyalBlue'>...</font> Line 12: Line 12: ├── rating.txt ├── rating.txt ├── thumb.jpeg ├── thumb.jpeg − └── '''v03'''+ └── '''v03''' (model version) ├── <font color='Indigo'>1527256903_350165-00-D0.0.jpeg</font> ├── <font color='Indigo'>1527256903_350165-00-D0.0.jpeg</font> ├── <font color='Indigo'>...</font> ├── <font color='Indigo'>...</font> Line 26: Line 26: ├── <font color='DarkGoldenrod'>1527256903_350165.obj</font> ├── <font color='DarkGoldenrod'>1527256903_350165.obj</font> ├── <font color='DarkGoldenrod'>1527256903_350165.x3d</font> ├── <font color='DarkGoldenrod'>1527256903_350165.x3d</font> − └── '''ml'''+ └── '''ml''' (directory with processed data files for ML) ├── <font color='ForestGreen'>1527256903_350165-ML_DATA-08B-O-FZ0.05-OFFS-2.00000.tiff</font> ├── <font color='ForestGreen'>1527256903_350165-ML_DATA-08B-O-FZ0.05-OFFS-2.00000.tiff</font> ├── <font color='ForestGreen'>1527256903_350165-ML_DATA-08B-O-FZ0.05-OFFS-1.00000.tiff</font> ├── <font color='ForestGreen'>1527256903_350165-ML_DATA-08B-O-FZ0.05-OFFS-1.00000.tiff</font> Line 34: Line 34: where: where: * <font color='RoyalBlue'>'''*.jp4'''</font> - source files, 0..3 - quad stereo camera #1, 4..7 - quad stereo camera #2, and so on if there are more cameras in the system. * <font color='RoyalBlue'>'''*.jp4'''</font> - source files, 0..3 - quad stereo camera #1, 4..7 - quad stereo camera #2, and so on if there are more cameras in the system. −* <font color='Indigo'>'''*-D0.0.jpeg'''</font> - disparity = 0, images are undistorted to a common polynom+* <font color='Indigo'>'''*-D0.0.jpeg'''</font> - disparity = 0, images are undistorted to a distortion polynom common for each image * <font color='OrangeRed'>'''*.corr-xml'''</font> - ImageJ plugin's settings file? * <font color='OrangeRed'>'''*.corr-xml'''</font> - ImageJ plugin's settings file? * <font color='Maroon'>'''*-DSI_COMBO.tiff'''</font> - Disparity Space Image - tiff stack * <font color='Maroon'>'''*-DSI_COMBO.tiff'''</font> - Disparity Space Image - tiff stack Line 40: Line 40: * <font color='DarkGoldenrod'>'''*.x3d, *.png'''</font> - X3D format model with textures. The textures are shared with the OBJ format model * <font color='DarkGoldenrod'>'''*.x3d, *.png'''</font> - X3D format model with textures. The textures are shared with the OBJ format model * <font color='DarkGoldenrod'>'''*.obj, *.mtl, *.png'''</font> - OBJ format model with textures * <font color='DarkGoldenrod'>'''*.obj, *.mtl, *.png'''</font> - OBJ format model with textures −* <font color='ForestGreen'>'''ml/*.tiff'''</font> - TIFF stack of pre-processed images for ML+* <font color='ForestGreen'>'''*.tiff'''</font> - TIFF stack of pre-processed images for ML * *.kml, rating.txt, thumb.jpeg - files, related to the online viewer only * *.kml, rating.txt, thumb.jpeg - files, related to the online viewer only −==Stacked TIFF==+==<font color='ForestGreen'>TIFF stacks for ML</font>== − +* What's in each stack is described in the [https://community.elphel.com/files/presentations/Elphel_TP-CNN_slides.pdf presentation for CVPR2018, pp.19-21]: +** 5 layers in the stack: ['''diagm-pair''', '''diago-pair''', '''hor-pairs''', '''vert-pairs''', '''other'''] +*** '''diagm-pair''' +*** '''diago-pair''' +*** '''hor-pairs''' +*** '''vert-pairs''' +*** '''other''' - encoded values: image disparity, estimated disparity and confidence for the estimated disparity * The source files are processed using a plugin for ImageJ, the output file for each set is a tiff stack * The source files are processed using a plugin for ImageJ, the output file for each set is a tiff stack * There are a few ways to view the stack: * There are a few ways to view the stack: −** just open with ImageJ+** Open with ImageJ - it has a native support for stacks, each stack has a name label stored (along with related xml info) in the ImageJ tiff tags. To read tiff tags in ImageJ, go '''Image > Show Info...''' +** Use '''imagej_tiff.py''' from [https://git.elphel.com/Elphel/python3-imagej-tiff python3-imagej-tiff] +*** Upon opening tiff the image shape will be '''(height,width,layer)''' - for further processing it needs to be reshaped to '''(height_in_tiles, width_in_tiles, tile_height, tile_width, layer)''': + <font size='2'>tiff stack shape: (2178, 2916, 5) + image data: (242, 324, 9, 9, 4) + values data: (242, 324, 3)</font> OlegTiff file format for pre-processed quad-stereo sets
← Older revision
Revision as of 19:39, 11 July 2018
(4 intermediate revisions by the same user not shown)Line 3:
Line 3:
<font size='2'>'''1527256903_350165/''' <font size='2'>'''1527256903_350165/'''
├── 1527256903_350165.kml ├── 1527256903_350165.kml
− ├── '''jp4'''+ ├── '''jp4''' (source files directory)
│ ├── <font color='RoyalBlue'>1527256903_350165_0.jp4</font> │ ├── <font color='RoyalBlue'>1527256903_350165_0.jp4</font>
│ ├── <font color='RoyalBlue'>...</font> │ ├── <font color='RoyalBlue'>...</font>
Line 12:
Line 12:
├── rating.txt ├── rating.txt
├── thumb.jpeg ├── thumb.jpeg
− └── '''v03'''+ └── '''v03''' (model version)
├── <font color='Indigo'>1527256903_350165-00-D0.0.jpeg</font> ├── <font color='Indigo'>1527256903_350165-00-D0.0.jpeg</font>
├── <font color='Indigo'>...</font> ├── <font color='Indigo'>...</font>
Line 26:
Line 26:
├── <font color='DarkGoldenrod'>1527256903_350165.obj</font> ├── <font color='DarkGoldenrod'>1527256903_350165.obj</font>
├── <font color='DarkGoldenrod'>1527256903_350165.x3d</font> ├── <font color='DarkGoldenrod'>1527256903_350165.x3d</font>
− └── '''ml'''+ └── '''ml''' (directory with processed data files for ML)
├── <font color='ForestGreen'>1527256903_350165-ML_DATA-08B-O-FZ0.05-OFFS-2.00000.tiff</font> ├── <font color='ForestGreen'>1527256903_350165-ML_DATA-08B-O-FZ0.05-OFFS-2.00000.tiff</font>
├── <font color='ForestGreen'>1527256903_350165-ML_DATA-08B-O-FZ0.05-OFFS-1.00000.tiff</font> ├── <font color='ForestGreen'>1527256903_350165-ML_DATA-08B-O-FZ0.05-OFFS-1.00000.tiff</font>
Line 34:
Line 34:
where: where:
* <font color='RoyalBlue'>'''*.jp4'''</font> - source files, 0..3 - quad stereo camera #1, 4..7 - quad stereo camera #2, and so on if there are more cameras in the system. * <font color='RoyalBlue'>'''*.jp4'''</font> - source files, 0..3 - quad stereo camera #1, 4..7 - quad stereo camera #2, and so on if there are more cameras in the system.
−* <font color='Indigo'>'''*-D0.0.jpeg'''</font> - disparity = 0, images are undistorted to a common polynom+* <font color='Indigo'>'''*-D0.0.jpeg'''</font> - disparity = 0, images are undistorted to a distortion polynom common for each image
* <font color='OrangeRed'>'''*.corr-xml'''</font> - ImageJ plugin's settings file? * <font color='OrangeRed'>'''*.corr-xml'''</font> - ImageJ plugin's settings file?
* <font color='Maroon'>'''*-DSI_COMBO.tiff'''</font> - Disparity Space Image - tiff stack * <font color='Maroon'>'''*-DSI_COMBO.tiff'''</font> - Disparity Space Image - tiff stack
Line 40:
Line 40:
* <font color='DarkGoldenrod'>'''*.x3d, *.png'''</font> - X3D format model with textures. The textures are shared with the OBJ format model * <font color='DarkGoldenrod'>'''*.x3d, *.png'''</font> - X3D format model with textures. The textures are shared with the OBJ format model
* <font color='DarkGoldenrod'>'''*.obj, *.mtl, *.png'''</font> - OBJ format model with textures * <font color='DarkGoldenrod'>'''*.obj, *.mtl, *.png'''</font> - OBJ format model with textures
−* <font color='ForestGreen'>'''ml/*.tiff'''</font> - TIFF stack of pre-processed images for ML+* <font color='ForestGreen'>'''*.tiff'''</font> - TIFF stack of pre-processed images for ML
* *.kml, rating.txt, thumb.jpeg - files, related to the online viewer only * *.kml, rating.txt, thumb.jpeg - files, related to the online viewer only
−==Stacked TIFF==+==<font color='ForestGreen'>TIFF stacks for ML</font>==
* The source files are processed using a plugin for ImageJ, the output file for each set is a tiff stack * The source files are processed using a plugin for ImageJ, the output file for each set is a tiff stack
* There are a few ways to view the stack: * There are a few ways to view the stack:
** just open with ImageJ ** just open with ImageJ
Oleg
Tiff file format for pre-processed quad-stereo sets
Created page with "==Image sets== Example tree of a single set: <font size='2'>'''1527256903_350165/''' ├── 1527256903_350165.kml ├── '''jp4''' │ ├── <font color='Royal..."
New page
==Image sets==Example tree of a single set:
<font size='2'>'''1527256903_350165/'''
├── 1527256903_350165.kml
├── '''jp4'''
│ ├── <font color='RoyalBlue'>1527256903_350165_0.jp4</font>
│ ├── <font color='RoyalBlue'>...</font>
│ ├── <font color='RoyalBlue'>1527256903_350165_3.jp4</font>
│ ├── <font color='RoyalBlue'>1527256903_350165_4.jp4</font>
│ ├── <font color='RoyalBlue'>...</font>
│ └── <font color='RoyalBlue'>1527256903_350165_7.jp4</font>
├── rating.txt
├── thumb.jpeg
└── '''v03'''
├── <font color='Indigo'>1527256903_350165-00-D0.0.jpeg</font>
├── <font color='Indigo'>...</font>
├── <font color='Indigo'>1527256903_350165-07-D0.0.jpeg</font>
├── <font color='OrangeRed'>1527256903_350165.corr-xml</font>
├── <font color='Maroon'>1527256903_350165-DSI_COMBO.tiff</font>
├── <font color='DarkSlateGray'>1527256903_350165-EXTRINSICS.corr-xml</font>
├── <font color='DarkGoldenrod'>1527256903_350165-img1-texture.png</font>
├── <font color='DarkGoldenrod'>...</font>
├── <font color='DarkGoldenrod'>1527256903_350165-img2001-texture.png</font>
├── <font color='DarkGoldenrod'>1527256903_350165-img_infinity-texture.png</font>
├── <font color='DarkGoldenrod'>1527256903_350165.mtl</font>
├── <font color='DarkGoldenrod'>1527256903_350165.obj</font>
├── <font color='DarkGoldenrod'>1527256903_350165.x3d</font>
└── '''ml'''
├── <font color='ForestGreen'>1527256903_350165-ML_DATA-08B-O-FZ0.05-OFFS-2.00000.tiff</font>
├── <font color='ForestGreen'>1527256903_350165-ML_DATA-08B-O-FZ0.05-OFFS-1.00000.tiff</font>
├── <font color='ForestGreen'>1527256903_350165-ML_DATA-08B-O-FZ0.05-OFFS0.00000.tiff</font>
├── <font color='ForestGreen'>1527256903_350165-ML_DATA-08B-O-FZ0.05-OFFS1.00000.tiff</font>
└── <font color='ForestGreen'>1527256903_350165-ML_DATA-08B-O-FZ0.05-OFFS2.00000.tiff</font></font>
where:
* <font color='RoyalBlue'>'''*.jp4'''</font> - source files, 0..3 - quad stereo camera #1, 4..7 - quad stereo camera #2, and so on if there are more cameras in the system.
* <font color='Indigo'>'''*-D0.0.jpeg'''</font> - disparity = 0, images are undistorted to a common polynom
* <font color='OrangeRed'>'''*.corr-xml'''</font> - ImageJ plugin's settings file?
* <font color='Maroon'>'''*-DSI_COMBO.tiff'''</font> - Disparity Space Image - tiff stack
* <font color='DarkSlateGray'>'''*-EXTRINSICS.corr-xml'''</font> - extrinsic parameters of the multicamera system
* <font color='DarkGoldenrod'>'''*.x3d, *.png'''</font> - X3D format model with textures. The textures are shared with the OBJ format model
* <font color='DarkGoldenrod'>'''*.obj, *.mtl, *.png'''</font> - OBJ format model with textures
* <font color='ForestGreen'>'''ml/*.tiff'''</font> - TIFF stack of pre-processed images for ML
* *.kml, rating.txt, thumb.jpeg - files, related to the online viewer only
==Stacked TIFF==
* The source files are processed using a plugin for ImageJ, the output file for each set is a tiff stack
* There are a few ways to view the stack:
** just open with ImageJ Oleg
Tensorflow with gpu
Setup (some details)
← Older revision Revision as of 17:07, 6 July 2018 Line 49: Line 49: # Export paths # Export paths <b>~$ export PATH=/usr/local/cuda-9.2/bin${PATH:+:${PATH}} <b>~$ export PATH=/usr/local/cuda-9.2/bin${PATH:+:${PATH}} − ~$ export LD_LIBRARY_PATH=/usr/local/cuda-9.2/lib64${LD_LIBRARY_PATH:+:${LD_LIBRARY_PATH}}</b></font>+ ~$ export LD_LIBRARY_PATH=/usr/local/cuda-9.2/lib64${LD_LIBRARY_PATH:+:${LD_LIBRARY_PATH}} + ~$ export LD_LIBRARY_PATH=/usr/local/cuda-9.2/extras/CUPTI/lib64/${LD_LIBRARY_PATH:+:${LD_LIBRARY_PATH}}</b></font> * Install TensorFlow (build from sources for cuda 9.2): * Install TensorFlow (build from sources for cuda 9.2): Oleg