#include <PointFeatureTracker.h>
Public Member Functions | |
virtual void | addPredictedState (const ft_state_t &predicted_state, const double &predicted_state_timestamp_ns) |
Add a new prediction about the state generated in the higher level (as prediction about the next measurement) More... | |
virtual void | correctState () |
Corrects the predicted state(s). First we track the features using both the prediction with the system model (copy of the previous feature locations) and if there is a prediction from the higher level (RBT) we track features using it too. We then decide which prediction is the "most likely" based on the quality measurement that the feature tracker returns and use the new location to update the belief state. More... | |
virtual cv_bridge::CvImagePtr | getDepthEdgesImg () |
virtual cv_bridge::CvImagePtr | getDepthImg () |
virtual cv_bridge::CvImagePtr | getDetectingMaskImg () |
virtual cv_bridge::CvImagePtr | getPredictedAndLastFeaturesImg () |
virtual cv_bridge::CvImagePtr | getPredictingMaskImg () |
virtual cv_bridge::CvImagePtr | getRGBImg () |
virtual ft_state_t | getState () const |
virtual cv_bridge::CvImagePtr | getTrackedFeaturesImg () |
virtual cv_bridge::CvImagePtr | getTrackedFeaturesWithPredictionMaskImg () |
virtual cv_bridge::CvImagePtr | getTrackingMaskImg () |
PointFeatureTracker (double loop_period_ns, bool using_pc=false, std::string ft_ns=std::string("")) | |
virtual void | predictMeasurement () |
Second step when updating the filter. The next measurement is predicted from the predicted next state (part of the RecursiveEstimatorInterface) More... | |
virtual void | predictState (double time_interval_ns) |
First step when updating the filter. The next state is predicted from current state and system model (part of the RecursiveEstimatorInterace) In the FeatureTracker the system model is the most simple one: no change in the state. So the predicted state is just the current state. More... | |
virtual void | setCameraInfoMsg (const sensor_msgs::CameraInfo *camera_info) |
virtual void | setDynamicReconfigureValues (feature_tracker::FeatureTrackerDynReconfConfig &config) |
Set the messages that have to be estimated and published. More... | |
virtual void | setFullRGBDPC (PointCloudPCL::ConstPtr full_rgb_pc) |
virtual void | setOcclusionMaskImg (cv_bridge::CvImagePtr occ_mask_img) |
virtual void | setSelfOcclusionPositive (bool so_positive) |
Set if the self occlusion image will be black in the background and white for the occluded area (so_positive = true, new) or white in the background and black in the occluded area (so_positive = false, old) More... | |
virtual | ~PointFeatureTracker () |
Public Member Functions inherited from omip::FeatureTracker | |
FeatureTracker (double loop_period_ns) | |
Default constructor. More... | |
virtual | ~FeatureTracker () |
Default destructor (empty) More... | |
Public Member Functions inherited from omip::RecursiveEstimatorFilterInterface< ft_state_t, ft_measurement_t > | |
virtual void | addPredictedState (const StateType &predicted_state, const double &predicted_state_timestamp_ns)=0 |
virtual MeasurementType | getPredictedMeasurement () const |
RecursiveEstimatorFilterInterface (double loop_period_ns) | |
virtual void | setMeasurement (const MeasurementType &acquired_measurement, const double &measurement_timestamp) |
virtual | ~RecursiveEstimatorFilterInterface () |
Protected Member Functions | |
int | _Compare3DCoords () |
void | _DetectNewFeatures () |
void | _EstimateDetectionMask () |
void | _InitializeVariables () |
void | _ProcessDepthImg () |
void | _ProcessOcclusionMaskImg () |
void | _ProcessPredictedLocationsPC () |
void | _ProcessRGBImg () |
void | _ReadParameters () |
int | _Retrieve3DCoords (std::vector< cv::Point2f > &points_in_2d, std::vector< bool > &points_status, std::vector< cv::Point3d > &points_in_3d) |
void | _TrackFeatures () |
Class PointFeatureTracker It implements the interface FeatureTracker Tracks features in the RGB stream and returns a PCL point cloud with their new 3D coordinates
Definition at line 63 of file PointFeatureTracker.h.
PointFeatureTracker::PointFeatureTracker | ( | double | loop_period_ns, |
bool | using_pc = false , |
||
std::string | ft_ns = std::string("") |
||
) |
Constructor
Definition at line 33 of file PointFeatureTracker.cpp.
|
virtual |
Destructor
Definition at line 71 of file PointFeatureTracker.cpp.
|
protected |
Compare current and previous 3D coordinates of the tracked Features to reject jumping ones. This limits the maximum velocity of a pixel in 3D space. 3D locations is too large) its status changes to false
Definition at line 966 of file PointFeatureTracker.cpp.
|
protected |
Definition at line 666 of file PointFeatureTracker.cpp.
|
protected |
Definition at line 655 of file PointFeatureTracker.cpp.
|
protected |
Initialize matrices and variables used during the execution
Definition at line 599 of file PointFeatureTracker.cpp.
|
protected |
Process the previously received depth img
Definition at line 229 of file PointFeatureTracker.cpp.
|
protected |
Process the previously received occlusion mask img
Definition at line 379 of file PointFeatureTracker.cpp.
|
protected |
Process the previously received predicted locations pointcloud
|
protected |
Process the previously received rgb img
Definition at line 372 of file PointFeatureTracker.cpp.
|
protected |
Read parameters from ROS
Definition at line 552 of file PointFeatureTracker.cpp.
|
protected |
Retrieve the 3D coordinates of a set of 2D Features
points_in_2d | - 2D locations of the Features in image plane |
points_ids | - ID of the Features |
points_status | - Status of the Features: If the 3D coordinates of a Feature cannot be retrieved, its status changes to false |
points_in_3d | - Returning vector with the 3D coordinates and the Id of the Features |
Definition at line 884 of file PointFeatureTracker.cpp.
|
protected |
Track the features that were detected previously and detect new features if needed
Definition at line 726 of file PointFeatureTracker.cpp.
|
virtual |
Add a new prediction about the state generated in the higher level (as prediction about the next measurement)
predicted_state | Predicted next state |
Definition at line 212 of file PointFeatureTracker.cpp.
|
virtual |
Corrects the predicted state(s). First we track the features using both the prediction with the system model (copy of the previous feature locations) and if there is a prediction from the higher level (RBT) we track features using it too. We then decide which prediction is the "most likely" based on the quality measurement that the feature tracker returns and use the new location to update the belief state.
Implements omip::FeatureTracker.
Definition at line 173 of file PointFeatureTracker.cpp.
|
virtual |
Get the image with the depth edges, used to reject Features that are on edges
Reimplemented from omip::FeatureTracker.
Definition at line 475 of file PointFeatureTracker.cpp.
|
virtual |
Reimplemented from omip::FeatureTracker.
Definition at line 458 of file PointFeatureTracker.cpp.
|
virtual |
Get the mask used to detect new features. Combines edges, max depth and mask of current features (to not detect them)
Reimplemented from omip::FeatureTracker.
Definition at line 499 of file PointFeatureTracker.cpp.
|
virtual |
Get the image with the last feature locations and the predicted feature locations
Reimplemented from omip::FeatureTracker.
Definition at line 507 of file PointFeatureTracker.cpp.
|
virtual |
Get the image with the depth edges, used to reject Features that are on edges
Reimplemented from omip::FeatureTracker.
Definition at line 483 of file PointFeatureTracker.cpp.
|
virtual |
Reimplemented from omip::FeatureTracker.
Definition at line 453 of file PointFeatureTracker.cpp.
|
virtual |
Get a point cloud with the 3D locations and the unique Feature Id (L value) of the last tracked features
Implements omip::FeatureTracker.
Definition at line 416 of file PointFeatureTracker.cpp.
|
virtual |
Get the last tracked Features as marks in the RGB image
Reimplemented from omip::FeatureTracker.
Definition at line 421 of file PointFeatureTracker.cpp.
|
virtual |
Get a pointer to a RGB image with markers where the Features have been tracked in the last step and only showing the RGB area where the features are searched
Reimplemented from omip::FeatureTracker.
Definition at line 463 of file PointFeatureTracker.cpp.
|
virtual |
Get the mask used to track features. Combines edges, predictions mask and max depth
Reimplemented from omip::FeatureTracker.
Definition at line 491 of file PointFeatureTracker.cpp.
|
virtual |
Second step when updating the filter. The next measurement is predicted from the predicted next state (part of the RecursiveEstimatorInterface)
In the FeatureTracker it is not clear what are the measurements:
Implements omip::RecursiveEstimatorFilterInterface< ft_state_t, ft_measurement_t >.
Definition at line 88 of file PointFeatureTracker.cpp.
|
virtual |
First step when updating the filter. The next state is predicted from current state and system model (part of the RecursiveEstimatorInterace) In the FeatureTracker the system model is the most simple one: no change in the state. So the predicted state is just the current state.
Implements omip::FeatureTracker.
Definition at line 81 of file PointFeatureTracker.cpp.
|
virtual |
Set the camera parameters to be used to project the 3D predicted Feature locations to the image plane and convert them into image coordinates
camera_info | - Pointer to the CameraInfo message containing the camera parameters |
Reimplemented from omip::FeatureTracker.
Definition at line 391 of file PointFeatureTracker.cpp.
|
virtual |
Set the messages that have to be estimated and published.
config | Value from Dynamic Reconfigure |
Implements omip::FeatureTracker.
Definition at line 76 of file PointFeatureTracker.cpp.
|
virtual |
Set a new RGBD point cloud to use for tracking in 3D. The RGBD point cloud is used to retrieve the 3D coordinates of the features tracked in 2D
full_rgb_pc | - Shared pointer to the new RGBD point cloud |
Reimplemented from omip::FeatureTracker.
Definition at line 218 of file PointFeatureTracker.cpp.
|
virtual |
Set a new occlusion mask image for tracking in 3D. The occlusion mask predicts the area of the RGB image that will be occluded (where it will be visible) the arm of the robot when it executes an interaction. Features on this area will be rejected.
occ_mask_img | - Shared pointer to the new occlusion mask image |
Reimplemented from omip::FeatureTracker.
Definition at line 223 of file PointFeatureTracker.cpp.
|
virtual |
Set if the self occlusion image will be black in the background and white for the occluded area (so_positive = true, new) or white in the background and black in the occluded area (so_positive = false, old)
so_positive | Flag value |
Definition at line 547 of file PointFeatureTracker.cpp.
|
protected |
Definition at line 293 of file PointFeatureTracker.h.
|
protected |
Definition at line 276 of file PointFeatureTracker.h.
|
protected |
Definition at line 291 of file PointFeatureTracker.h.
|
protected |
Definition at line 266 of file PointFeatureTracker.h.
|
protected |
Definition at line 275 of file PointFeatureTracker.h.
|
protected |
Definition at line 296 of file PointFeatureTracker.h.
|
protected |
Definition at line 298 of file PointFeatureTracker.h.
|
protected |
Definition at line 297 of file PointFeatureTracker.h.
|
protected |
Definition at line 282 of file PointFeatureTracker.h.
|
protected |
Definition at line 281 of file PointFeatureTracker.h.
|
protected |
Definition at line 272 of file PointFeatureTracker.h.
|
protected |
Definition at line 284 of file PointFeatureTracker.h.
|
protected |
Definition at line 276 of file PointFeatureTracker.h.
|
protected |
Definition at line 273 of file PointFeatureTracker.h.
|
protected |
Definition at line 272 of file PointFeatureTracker.h.
|
protected |
Definition at line 275 of file PointFeatureTracker.h.
|
protected |
Definition at line 274 of file PointFeatureTracker.h.
|
protected |
Definition at line 276 of file PointFeatureTracker.h.
|
protected |
Definition at line 276 of file PointFeatureTracker.h.
|
protected |
Definition at line 296 of file PointFeatureTracker.h.
|
protected |
Definition at line 296 of file PointFeatureTracker.h.
|
protected |
Definition at line 303 of file PointFeatureTracker.h.
|
protected |
Definition at line 283 of file PointFeatureTracker.h.
|
protected |
Definition at line 275 of file PointFeatureTracker.h.
|
protected |
Definition at line 275 of file PointFeatureTracker.h.
|
protected |
Definition at line 300 of file PointFeatureTracker.h.
|
protected |
Definition at line 305 of file PointFeatureTracker.h.
|
protected |
Definition at line 291 of file PointFeatureTracker.h.
|
protected |
Definition at line 297 of file PointFeatureTracker.h.
|
protected |
Definition at line 301 of file PointFeatureTracker.h.
|
protected |
Definition at line 278 of file PointFeatureTracker.h.
|
protected |
Definition at line 298 of file PointFeatureTracker.h.
|
protected |
Definition at line 298 of file PointFeatureTracker.h.
|
protected |
Definition at line 298 of file PointFeatureTracker.h.
|
protected |
Definition at line 297 of file PointFeatureTracker.h.
|
protected |
Definition at line 298 of file PointFeatureTracker.h.
|
protected |
Definition at line 298 of file PointFeatureTracker.h.
|
protected |
Definition at line 297 of file PointFeatureTracker.h.
|
protected |
Definition at line 298 of file PointFeatureTracker.h.
|
protected |
Definition at line 265 of file PointFeatureTracker.h.
|
protected |
Definition at line 297 of file PointFeatureTracker.h.
|
protected |
Definition at line 272 of file PointFeatureTracker.h.
|
protected |
Definition at line 291 of file PointFeatureTracker.h.
|
protected |
Definition at line 276 of file PointFeatureTracker.h.
|
protected |
Definition at line 274 of file PointFeatureTracker.h.
|
protected |
Definition at line 275 of file PointFeatureTracker.h.
|
protected |
Definition at line 280 of file PointFeatureTracker.h.
|
protected |
Definition at line 291 of file PointFeatureTracker.h.
|
protected |
Definition at line 291 of file PointFeatureTracker.h.
|
protected |
Definition at line 287 of file PointFeatureTracker.h.
|
protected |
Definition at line 288 of file PointFeatureTracker.h.
|
protected |
Definition at line 273 of file PointFeatureTracker.h.
|
protected |
Definition at line 282 of file PointFeatureTracker.h.
|
protected |
Definition at line 272 of file PointFeatureTracker.h.
|
protected |
Definition at line 276 of file PointFeatureTracker.h.
|
protected |
Definition at line 284 of file PointFeatureTracker.h.
|
protected |
Definition at line 281 of file PointFeatureTracker.h.
|
protected |
Definition at line 287 of file PointFeatureTracker.h.
|
protected |
Definition at line 297 of file PointFeatureTracker.h.
|
protected |
Definition at line 292 of file PointFeatureTracker.h.
|
protected |
Definition at line 292 of file PointFeatureTracker.h.
|
protected |
Definition at line 292 of file PointFeatureTracker.h.
|
protected |
Definition at line 269 of file PointFeatureTracker.h.
|
protected |
Definition at line 285 of file PointFeatureTracker.h.
|
protected |
Definition at line 272 of file PointFeatureTracker.h.
|
protected |
Definition at line 298 of file PointFeatureTracker.h.
|
protected |
Definition at line 307 of file PointFeatureTracker.h.
|
protected |
Definition at line 273 of file PointFeatureTracker.h.
|
protected |
Definition at line 273 of file PointFeatureTracker.h.
|
protected |
Definition at line 274 of file PointFeatureTracker.h.
|
protected |
Definition at line 293 of file PointFeatureTracker.h.
|
protected |
Definition at line 291 of file PointFeatureTracker.h.