find_calib_descriptor_modelT_find_calib_descriptor_modelfind_calib_descriptor_modelFindCalibDescriptorModelFindCalibDescriptorModel (Operator)

Name

find_calib_descriptor_modelT_find_calib_descriptor_modelfind_calib_descriptor_modelFindCalibDescriptorModelFindCalibDescriptorModel — Find the best matches of a calibrated descriptor model in an image and return their 3D pose.

Signature

Description

The operator find_calib_descriptor_modelfind_calib_descriptor_modelfind_calib_descriptor_modelFindCalibDescriptorModelFindCalibDescriptorModel finds the best NumMatchesNumMatchesNumMatchesNumMatchesnumMatches matches of the descriptor model ModelIDModelIDModelIDModelIDmodelID in ImageImageImageImageimage. The descriptor model must have been created previously by calling create_calib_descriptor_modelcreate_calib_descriptor_modelcreate_calib_descriptor_modelCreateCalibDescriptorModelCreateCalibDescriptorModel or read_descriptor_modelread_descriptor_modelread_descriptor_modelReadDescriptorModelReadDescriptorModel.

The operator returns the 3D PosePosePosePosepose (in camera coordinate system) of the template object found in the search ImageImageImageImageimage. As two different cameras can be used for the model generation and the online search, the camera parameters of the camera used to create ImageImageImageImageimage have to be set in CamParamCamParamCamParamCamParamcamParam.

The detection process is the same as for find_uncalib_descriptor_modelfind_uncalib_descriptor_modelfind_uncalib_descriptor_modelFindUncalibDescriptorModelFindUncalibDescriptorModel. The setting of DetectorParamNameDetectorParamNameDetectorParamNameDetectorParamNamedetectorParamName, DetectorParamValueDetectorParamValueDetectorParamValueDetectorParamValuedetectorParamValue, DescriptorParamNameDescriptorParamNameDescriptorParamNameDescriptorParamNamedescriptorParamName, DescriptorParamValueDescriptorParamValueDescriptorParamValueDescriptorParamValuedescriptorParamValue, MinScoreMinScoreMinScoreMinScoreminScore, NumMatchesNumMatchesNumMatchesNumMatchesnumMatches and ScoreTypeScoreTypeScoreTypeScoreTypescoreType are explained with find_uncalib_descriptor_modelfind_uncalib_descriptor_modelfind_uncalib_descriptor_modelFindUncalibDescriptorModelFindUncalibDescriptorModel. After the object was detected, and using the world (from reference plane) coordinates of the matched model points (explained with create_calib_descriptor_modelcreate_calib_descriptor_modelcreate_calib_descriptor_modelCreateCalibDescriptorModelCreateCalibDescriptorModel), the pose of the object is computed similarly to applying vector_to_posevector_to_posevector_to_poseVectorToPoseVectorToPose operator.

The point correspondences for each object can be queried with get_descriptor_model_pointsget_descriptor_model_pointsget_descriptor_model_pointsGetDescriptorModelPointsGetDescriptorModelPoints.

Attention

A descriptor model 'ModelID'"ModelID""ModelID""ModelID""ModelID" cannot be shared between two or more user's threads. Different descriptor models can be used independently and safely in different threads.

Parallelization

• Multithreading type: reentrant (runs in parallel with non-exclusive operators).
• Multithreading scope: global (may be called from any thread).
• Processed without parallelization.

Parameters

Example (HDevelop)

* Determination of a reference pose:
* Two images of the object are needed: One shows the object (Image)
* and the other one shows the object with a calibration table put
* onto it (ImageCal)

find_caltab (ImageCal,Caltab,CaltabName,SizeGauss,MarkThresh,MinDiamMarks)
find_marks_and_pose (ImageCal,Caltab,CaltabName,CamParam,StartThresh, \
                     DeltaThresh,MinThresh,Alpha,MinContLength, \
                     MaxDiamMarks,RCoord,CCoord, \
                     InitialPoseForCalibrationPlate)
caltab_points (CaltabName,X,Y,Z)
camera_calibration (X,Y,Z,RCoord,CCoord,CamParam, \
                    InitialPoseForCalibrationPlate,'pose', \
                    CamParamUnchanged,PoseReference,Errors)

* The thickness of the calibration table itself (0.75 mm) has to be
* reduced:
set_origin_pose (PoseReference,0,0,0.00075,PoseReference)

* Create descriptor model:
create_calib_descriptor_model (Image,CamParam,PoseReference, \
                               'harris',[],[],[],[],42,ModelID)

get_descriptor_model_params (ModelID,DetectorType, \
                             DetectorParamName,DetectorParamValue, \
                             DescriptorParamName,DescriptorParamValue)
write_descriptor_model (ModelID,'simple_example.dsm')
clear_descriptor_model (ModelID)

read_descriptor_model ('simple_example.dsm',ModelID)
find_calib_descriptor_model (SearchImage,ModelID,[],[],[],[],0.2,1, \
                             CamParam,['num_points','inlier_ratio'], \
                             Pose,Score)
clear_descriptor_model (ModelID)

Possible Predecessors

create_calib_descriptor_modelcreate_calib_descriptor_modelcreate_calib_descriptor_modelCreateCalibDescriptorModelCreateCalibDescriptorModel, read_descriptor_modelread_descriptor_modelread_descriptor_modelReadDescriptorModelReadDescriptorModel

See also

vector_to_posevector_to_posevector_to_poseVectorToPoseVectorToPose, points_lepetitpoints_lepetitpoints_lepetitPointsLepetitPointsLepetit, create_calib_descriptor_modelcreate_calib_descriptor_modelcreate_calib_descriptor_modelCreateCalibDescriptorModelCreateCalibDescriptorModel

Module

Matching