|
Vision and Mobile Robotics Laboratory | Software
|
|
| Home | Members | Projects | Publications | Software | Videos | Internal |
DisplayRecognition is used to display the results of RecognizeMesh. It reads a recognition file and displays the result in 3-D, or 2-D if a co-registered intensity image is available for the scene. A recognition file is created by RecognizeMesh and has the following format:
Scene: /IUS/usrp0/aej/sai1/scene/scene092
library: duckie20
Lib meshDir: /IUS/sai1/aej/library/mesh/model/
stackDir: /IUS/sai1/aej/library/spinImage/a60w10h10b0.5/
matchAlg: MA4
modelDim: 20
libraryDim: 20
numMatches: 4
model: MrPH
trans: -1.92373 -6.0716 1.17473 1.7341 -5.47583 -7.25301
model: bunny
trans: -4.11641 -1.65256 0.474002 10.2353 -6.12627 10.9399
model: faucet
trans: 5.36356 -0.437702 2.32639 29.7154 -17.957 157.149
model: ysplit
trans: 6.21276 -4.67035 5.82431 100.786 -38.192 117.438
ReconStats
num scene points: 2266
num correspondences: 2672
num_trans: 125 correspondence
time: 1089
grouping time: 49
verify time: 105
DisplayRecognition.c contains all of the functionality for creating and displaying recognition results. It can display results in 3-D or 2-D.
By typing 'DisplayRecognition -' the following options (format description default) are printed:
Usage: DisplayRecognition (See DisplayRecognition.html for complete usage)
%S set recognition file [required]
-analyze %S select model to analyze [off]
-plane display plane result [off]
-gray display plane result [off]
-write %S save result to VRML file [off]
%S set recognition file [required]
Indicate the filename of the recognition file to display.
-analyze %S select model to analyze [off]
Turn on analysis mode which can be used to interactivley determine segment the scene to determine how much clutter and occlusion exist in the scene (not documented).
-plane display plane result [off]
Display the recognition result in planar format. this requires that a scene tiff image and projection matrix exist in the same directory (with the same prefix) as the scene mesh. Recognized models are shown projected onto the scene image. (Requires the image format library (IFL)).
-gray display plane result [off]
Change the display to grayscale for generation of figures for bW hardcopy.
-write %S save result to VRML file [off]
Write out the recognition result to the indicated VRML file.
To create a 3-D color display of the recognition result in scene092.duckie20Lib.MA4.rf.
DisplayRecognition scene092.duckie20Lib.MA4
To create a 2-D grayscale display of the recognition result in scene092.duckie20Lib.MA4.rf.
DisplayRecognition scene092.duckie20Lib.MA4 -plane -gray