-- Lion.data ----
Lion Status BS Canon EOS 350D 50mm lens
INFO
Imager (Chip) dimensions of the used camera in mm
22.2 14.8
Image resolution in pixel
3456 2304
Interior orientations:
Focal length xo yo (xo and yo are the principle point coordinates in mm)
50.1727 -0.1163 -0.0477
Image distortions: (If any type of this distortions is not considered then his parameters will be set as 0)
Three radial (A1 A2 A3)
Two decentring (P1 P2)
Two for affinity (B1 B2)
1.95244e-005 3.52660e-008 -2.86435e-010
-1.37926e-005 -7.95748e-006
-8.21068e-005 -7.91334e-005
Type of exterior orientations
1 or OM-PH-K for Xo Yo Zo omega phi kappa in Degrees
2 or A-E-R for Xo Yo Zo Alpha elevation roll in Degrees
3 or OR-MX for Orientation Matrix Xo Yo Zo
OM11 OM12 OM13
OM21 OM22 OM23
OM31 OM32 OM33
A-E-R
Lion 001.jpg
-11.94132 -0.04854 -0.24880 -88.291238 -49.385139 -176.354413
Lion 003.jpg
1.21563 0.78480 7.73763 100.114852 -85.995881 -10.552748
Lion 005.jpg
11.39189 -0.25594 -0.81352 87.470769 -47.974523 -3.575856
Lion 006.jpg
20.57288 -1.87895 -15.25517 84.202754 4.364268 -7.170819
Lion 007.jpg
14.31320 -3.30490 -26.90039 76.644667 41.593681 -13.606379
Lion 008.jpg
-0.09715 -4.80307 -38.05739 2.198462 77.541761 -88.170208
Lion 009.jpg
-15.40439 -3.77216 -28.40165 -73.187465 43.254039 -162.040277
Lion 010.jpg
-23.39352 -1.69409 -11.44748 -84.135265 -4.585547 -173.485897
Lion 023.jpg
33.90188 2.33820 -33.74153 91.771641 30.017836 -93.756728
Lion 038.jpg
31.49309 2.17684 -35.19394 93.118045 34.275667 -92.242779
Lion 016.jpg
34.51276 2.43346 -32.02294 92.431718 27.438128 -92.928185
Lion 036.jpg
-12.73367 5.86597 4.03858 -112.290680 -52.729972 -158.802688
Lion 039.jpg
-22.17290 5.15946 -1.09332 -101.113997 -28.861841 -166.888133
Lion 026.jpg
3.24805 6.37802 6.96328 150.509730 -70.737415 -59.990579
-----------------------------------------------------------------------------------------------------
2) Create a batch file as shown in the following figure 6 by answering questions in an interactive way. The required information is the lion mesh, the 14 photos together with their interior and exterior orientations.
---------- Lion.bat ------------------------------------------------
Options
Photo Occlusions Finder Algorithm could not be activated in this process
0
Consider the Images Priority according to their Order (0/1)
0
Give priority for the following number of images as group_1 (ranging from 1 to NPhotos)
0
Give priority for the following number of images as group_2
0
Give priority for the following number of images as group_3
0
Activate the Local Patch Reassign LPR function for homogeneous texture (0 for OFF)
0
Assign Partially Occluded Parts as Occluded (0/1)
0
Remove the un-textured parts (0/1)
0
Color Criteria
20
Shut Down after finish (0/1)
0
lion-BS
Lion_GM.wrl
Orientations file
Lion.data
Photos
Lion 001.jpg
Lion 003.jpg
Lion 005.jpg
Lion 006.jpg
Lion 007.jpg
Lion 008.jpg
Lion 009.jpg
Lion 010.jpg
Lion 023.jpg
Lion 038.jpg
Lion 016.jpg
Lion 036.jpg
Lion 039.jpg
Lion 026.jpg
--------------------------------------------------------------------
3) Run the batch file in the 3DImage software Click (F12). One important issue is that the computer processing time is reasonable, as it takes about twenty minutes processing time to produce the desired textured model file.
4) A VRML textured file (**.wrl) with the same name of the batch file will be generated in the same folder. The textured model file can be opened now in any standard VRML viewer. After that, gray view, walk in, and animation can be achieved. Figure 8 shows some snapshots from the textured model of the status with different views.
5) The efficiency of the software can be judged by investigating figure 9. In that figure, the status photos have been replaced with single color photos. These single color photos are then used for visualization. It is clear from the figure that, the algorithm selects correctly the most appropriate photo for each part from the model as what might be selected by any experienced operator.
The followings are some other options in the software:
- Accept two different types of orientation angles. The orientation matrix itself can be also used.
- It is possible to give priority to groups of photos in three levels, this means that the software will search for the best texture in the first group then in the second one then in the third. In case if there is no texture available in the three groups, the software will search the texture in the remaining photos.
- The software can also use the photos in order, which means that it will begin with the first one searching for all the available texture in the model then the following and so on. This option is important in case of there is a frontal image for a façade or a wall painting for example.
- The user might push the software to texture a certain part from the mesh model from a certain photo. The idea behind this; is to mask this part in the remaining photos with a specific colour “Black”. Accordingly, the software will detect this colour and consider it as occluded areas. In that case, the only choice for the algorithm is to use the photo without masking to texture this part, see figure.
Mask option result
- Another example to get use of the mask and priority options in order to texture a complete wall painting from one photo, the painting can be selected in that photo and then the remaining part of the photo can be masked. With giving the masked photo priority, the software will use this photo first to texture the entire painting and then use the other photos to complete the rest of the model, see (Abdelhafiz et al, 2009).