org.bytedeco.chilitags

Class Chilitags3Dd

java.lang.Object

org.bytedeco.javacpp.Pointer

org.bytedeco.chilitags.Chilitags3Dd

All Implemented Interfaces:

AutoCloseable

@Name(value="chilitags::Chilitags3D_<double>")
 @NoOffset
 @Properties(inherit=chilitags.class)
public class Chilitags3Dd
extends Pointer

Nested Class Summary

Nested classes/interfaces inherited from class org.bytedeco.javacpp.Pointer

Pointer.CustomDeallocator, Pointer.Deallocator, Pointer.NativeDeallocator, Pointer.ReferenceCounter

Field Summary

Fields inherited from class org.bytedeco.javacpp.Pointer

address, capacity, limit, position

Constructor Summary

Constructors

Constructor and Description
Chilitags3Dd()
Chilitags3Dd(long size) Native array allocator.
Chilitags3Dd(Pointer p) Pointer cast constructor.
Chilitags3Dd(org.bytedeco.opencv.opencv_core.Size cameraResolution) Creates an object ready to find the 3D pose of chilitags.

Method Summary

Modifier and Type	Method and Description
void	enableFilter(boolean enabled) \brief Enables/disables Kalman filtering on tag pose (enabled by default)
TagPoseMapd	estimate(org.bytedeco.opencv.opencv_core.Mat inputImage) This is a convenience variant of estimate() which also takes care of the detection.
TagPoseMapd	estimate(TagCornerMap tags)
org.bytedeco.opencv.opencv_core.Mat	getCameraMatrix() Returns the camera matrix used for the pose estimation.
Chilitags	getChilitags() Accessor to the underlying (2D) Chilitags detection.
org.bytedeco.opencv.opencv_core.Mat	getDistortionCoeffs() Returns the distortion coefficients used for the pose estimation.
Chilitags3Dd	getPointer(long i)
Chilitags3Dd	position(long position)
org.bytedeco.opencv.opencv_core.Size	readCalibration(BytePointer filename) For accurate results, Chilitags3D can be provided the calibration data of the camera detecting the chilitags.
org.bytedeco.opencv.opencv_core.Size	readCalibration(String filename)
boolean	readTagConfiguration(BytePointer filenameOrString)
boolean	readTagConfiguration(BytePointer filenameOrString, boolean omitOtherTags, boolean readFromString) Chilitags3D can also detect rigid assemblies of tags.
boolean	readTagConfiguration(String filenameOrString)
boolean	readTagConfiguration(String filenameOrString, boolean omitOtherTags, boolean readFromString)
void	setCalibration(org.bytedeco.opencv.opencv_core.GpuMat newCameraMatrix, org.bytedeco.opencv.opencv_core.GpuMat newDistCoeffs)
void	setCalibration(org.bytedeco.opencv.opencv_core.Mat newCameraMatrix, org.bytedeco.opencv.opencv_core.Mat newDistCoeffs) For accurate results, Chilitags3D can be provided the calibration data of the camera detecting the chilitags.
void	setCalibration(org.bytedeco.opencv.opencv_core.UMat newCameraMatrix, org.bytedeco.opencv.opencv_core.UMat newDistCoeffs)
void	setDefaultTagSize(double defaultSize) Sets the default size of tags (used to compute their 3D pose) when not explicitly specified with read3DConfiguration().
void	setFilterObservationNoiseCovariance(org.bytedeco.opencv.opencv_core.Mat covariance) \brief Sets the observation noise covariance matrix a.k.a R for the Kalman filter The observation (done by image processing on the camera image) is described by (x,y,z,qw,qx,qy,qz) where x,y,z is the tag position and qw,qx,qy,qz is the tag orientation in quaternion representation.
void	setFilterProcessNoiseCovariance(org.bytedeco.opencv.opencv_core.Mat covariance) \brief Sets the process noise covariance matrix a.k.a Q for the Kalman filter The state is described by (x,y,z,qw,qx,qy,qz) where x,y,z is the tag position and qw,qx,qy,qz is the tag orientation in quaternion representation.
void	setPersistence(double persistence) \brief Sets the persistence of tags against being discarded when not observed (10 by default)

Methods inherited from class org.bytedeco.javacpp.Pointer

address, asBuffer, asByteBuffer, availablePhysicalBytes, calloc, capacity, capacity, close, deallocate, deallocate, deallocateReferences, deallocator, deallocator, equals, fill, formatBytes, free, getDirectBufferAddress, getPointer, getPointer, getPointer, hashCode, interruptDeallocatorThread, isNull, isNull, limit, limit, malloc, maxBytes, maxPhysicalBytes, memchr, memcmp, memcpy, memmove, memset, offsetAddress, offsetof, offsetof, parseBytes, physicalBytes, physicalBytesInaccurate, position, put, realloc, referenceCount, releaseReference, retainReference, setNull, sizeof, sizeof, toString, totalBytes, totalCount, totalPhysicalBytes, withDeallocator, zero

Methods inherited from class java.lang.Object

clone, finalize, getClass, notify, notifyAll, wait, wait, wait

Constructor Detail

Chilitags3Dd

public Chilitags3Dd(Pointer p)

Pointer cast constructor. Invokes Pointer(Pointer).

Chilitags3Dd

public Chilitags3Dd(long size)

Native array allocator. Access with Pointer.position(long).

Chilitags3Dd

public Chilitags3Dd(@ByVal(nullValue="cv::Size(640, 480)")
                    org.bytedeco.opencv.opencv_core.Size cameraResolution)

Creates an object ready to find the 3D pose of chilitags.

By default, Chilitags3D assumes an arbitrary, but reasonnable focal length (700), and expects the dimensions of the captured images. In this configuration, the depth estimation makes sense, but it is not accurate. In order to correctly estimate the 3D pose, the intrinsic calibration parameters of your camera needs to be provided. To do so, use the readCalibration() or setCalibration() methods.

Chilitags3D also assumes by default that the 3D pose of every detected tag is expected, and that every tag is independent from the others, and that they are 20 millimetres wide. The method read3DConfiguration() can be used to specify which tags are of interest, and how they are arranged on a rigid object, and how big they are.

To first detect th tags in the image, Chilitags3D creates a Chilitags instance, which can be accessed through the getChilitags() accessors. This Chilitags instance is set to have a persistence of 0, because Chilitags3D uses a more advanced Kalman filter. See enableFilter() and setPersistence() for more details.

You can also create yourself a separate instance of Chilitagsfor the 2D detection of tags and use it by calling Chilitags3D::estimate(const TagCornerMap &tags) with the output of Chilitags::find(const cv::Mat &inputImage) instead of calling directly Chilitags3D::estimate(const cv::Mat &inputImage).

Parameters:

cameraResolution - Resolution of the camera used as input (640x480 by default). This parameter is only used to provide meaningful pose estimation. Input images of different resolution can be provided to the detection anyway. This parameter is overridden by readCalibration() or setCalibration().

Chilitags3Dd

public Chilitags3Dd()

Method Detail

position

public Chilitags3Dd position(long position)

Overrides:

position in class Pointer

getPointer

public Chilitags3Dd getPointer(long i)

Overrides:

getPointer in class Pointer

getChilitags

@ByRef
public Chilitags getChilitags()

Accessor to the underlying (2D) Chilitags detection.

estimate

@ByVal
 @Cast(value="chilitags::Chilitags3D_<double>::TagPoseMap*")
public TagPoseMapd estimate(@Const @ByRef
                                                                                                  TagCornerMap tags)

Parameters:

tags - a list of tags, as returned by Chilitags::find().

camDeltaR - Rotation from the previous camera frame to the current camera frame, i.e rotation of the current camera frame in the last camera frame. Quaternion format (scalar, vx, vy, vz).

camDeltaX - Translation from the previous camera frame to the current camera frame, i.e position of the current camera frame in the last camera frame.

Returns:

a mapping of the detected objects to their transformation matrices. Transformation matrices are row-major and follow the standard convention to code the rotation and translation parameters in homogeneous coordinates:

    { r11 , r12 , r13 , tx
      r21 , r22 , r23 , ty
      r31 , r32 , r33 , tz
        0 ,   0 ,   0 ,  1 }
    

estimate

@ByVal
 @Cast(value="chilitags::Chilitags3D_<double>::TagPoseMap*")
public TagPoseMapd estimate(@Const @ByRef
                                                                                                  org.bytedeco.opencv.opencv_core.Mat inputImage)

This is a convenience variant of estimate() which also takes care of the detection.

Parameters:

inputImage - the image to feed to Chilitags::find().

detectionTrigger - specifies how to combine tracking and full detection. Tracking is drastically faster, but it can at best return tags previously found; it won't find new ones, but can lose some. See Chilitags::DetectionTrigger for a description of the possible values.

camDeltaR - Rotation from the previous camera frame to the current camera frame, i.e rotation of the current camera frame in the last camera frame. Quaternion format (scalar, vx, vy, vz).

camDeltaX - Translation from the previous camera frame to the current camera frame, i.e position of the current camera frame in the last camera frame.

Returns:

a mapping of the detected objects to their transformation matrices. Transformation matrices are row-major and follow the standard convention to code the rotation and translation parameters in homogeneous coordinates:

    { r11 , r12 , r13 , tx
      r21 , r22 , r23 , ty
      r31 , r32 , r33 , tz
        0 ,   0 ,   0 ,  1 }
    

readTagConfiguration

@Cast(value="bool")
public boolean readTagConfiguration(@StdString
                                                         BytePointer filenameOrString,
                                                         @Cast(value="bool")
                                                         boolean omitOtherTags,
                                                         @Cast(value="bool")
                                                         boolean readFromString)

Chilitags3D can also detect rigid assemblies of tags. This allows for a more precise estimation of the object holding the tag, and for a graceful degradation of the estimation, should some of the tag be misdetected or occluded.

Parameters:

filenameOrString - The name of the YAML configuration file (or the whole file itself as a string) describing rigid clusters of tags. The library is distributed with a sample configuration file documenting the expected format.

omitOtherTags - If true, ignore the tags that are not explicitly listed in the configuration file. If false (default), Chilitags3D::estimate() estimates the 3D pose of all the detected tags. You can set the size of tags not described in the configuration file with setDefaultTagSize().

readFromString - If true, will read tag configuration directly from the given string. If false (default) will open the file with the given name and try to read the configuration from there.

Returns:

Whether reading the configuration was successful

readTagConfiguration

@Cast(value="bool")
public boolean readTagConfiguration(@StdString
                                                         BytePointer filenameOrString)

readTagConfiguration

@Cast(value="bool")
public boolean readTagConfiguration(@StdString
                                                         String filenameOrString,
                                                         @Cast(value="bool")
                                                         boolean omitOtherTags,
                                                         @Cast(value="bool")
                                                         boolean readFromString)

readTagConfiguration

@Cast(value="bool")
public boolean readTagConfiguration(@StdString
                                                         String filenameOrString)

setDefaultTagSize

public void setDefaultTagSize(double defaultSize)

Sets the default size of tags (used to compute their 3D pose) when not explicitly specified with read3DConfiguration(). To be accurate, the unit must match the unit used for the camera calibration (usually, millimetres).

Note that is assumes all the tags have the same size. If tags have different size, you may want to list them in the configuration file (see read3DConfiguration()).

The default value of the default tag size is 20 millimetres.

enableFilter

public void enableFilter(@Cast(value="bool")
                         boolean enabled)

\brief Enables/disables Kalman filtering on tag pose (enabled by default)

Parameters:

enabled - Whether to enable Kalman filtering

setPersistence

public void setPersistence(double persistence)

\brief Sets the persistence of tags against being discarded when not observed (10 by default)

Parameters:

persistence - Persistence value, roughly correponds to number of frames

setFilterProcessNoiseCovariance

public void setFilterProcessNoiseCovariance(@Const @ByRef
                                            org.bytedeco.opencv.opencv_core.Mat covariance)

\brief Sets the process noise covariance matrix a.k.a Q for the Kalman filter The state is described by (x,y,z,qw,qx,qy,qz) where x,y,z is the tag position and qw,qx,qy,qz is the tag orientation in quaternion representation. Therefore, the input matrix should correspond to cov((x,y,z,qw,qx,qy,qz)). When IMU data is present, the process moves the tag according to camera movement. When IMU data is not present, the process tries to keep the tag still. Smaller values in the process covariance matrix causes the tags to resist motion more. By default, this matrix has 1e-3, 1e-3, 1e-3, 1e-4, 1e-4, 1e-4, 1e-4 on its diagonal.

Parameters:

covariance - 7x7 covariance matrix

setFilterObservationNoiseCovariance

public void setFilterObservationNoiseCovariance(@Const @ByRef
                                                org.bytedeco.opencv.opencv_core.Mat covariance)

\brief Sets the observation noise covariance matrix a.k.a R for the Kalman filter The observation (done by image processing on the camera image) is described by (x,y,z,qw,qx,qy,qz) where x,y,z is the tag position and qw,qx,qy,qz is the tag orientation in quaternion representation. Therefore, the input matrix should correspond to cov((x,y,z,qw,qx,qy,qz)). Larger values in the observation noise covariance matrix indicates noisier measurements and causes observations to affect the state less, making tags more resistant to motion. By default, this matrix has 1e-3, 1e-3, 1e-1, 1e-3, 1e-2, 1e-2, 1e-5 on its diagonal.

Parameters:

covariance - 7x7 covariance matrix

setCalibration

public void setCalibration(@ByVal
                           org.bytedeco.opencv.opencv_core.Mat newCameraMatrix,
                           @ByVal
                           org.bytedeco.opencv.opencv_core.Mat newDistCoeffs)

For accurate results, Chilitags3D can be provided the calibration data of the camera detecting the chilitags. See https://docs.opencv.org/modules/calib3d/doc/camera_calibration_and_3d_reconstruction.html for background on this topic.

Note that this method can be called as often as needed with a new calibration configuration (for instance if the user switched to another camera).

Parameters:

newCameraMatrix - the 3x3 matrix of the camera intrinsics (see https://en.wikipedia.org/wiki/Camera_resectioning#Intrinsic_parameters).

newDistCoeffs - a vector containing the distortion coefficients.

setCalibration

public void setCalibration(@ByVal
                           org.bytedeco.opencv.opencv_core.UMat newCameraMatrix,
                           @ByVal
                           org.bytedeco.opencv.opencv_core.UMat newDistCoeffs)

setCalibration

public void setCalibration(@ByVal
                           org.bytedeco.opencv.opencv_core.GpuMat newCameraMatrix,
                           @ByVal
                           org.bytedeco.opencv.opencv_core.GpuMat newDistCoeffs)

readCalibration

@ByVal
public org.bytedeco.opencv.opencv_core.Size readCalibration(@StdString
                                                                    BytePointer filename)

For accurate results, Chilitags3D can be provided the calibration data of the camera detecting the chilitags. See http://docs.opencv.org/modules/calib3d/doc/camera_calibration_and_3d_reconstruction.html for background on this topic.

Note that this method can be called as often as needed with a new calibration configuration (for instance if the user switched to another camera).

This method is similar to setCalibration, but reads the camera calibration information directly from a file, as generated by OpenCV's 'calibration' sample.

Parameters:

filename - the path to a file containing the calibration data

Returns:

the size of the images used to generate the calibration data.

readCalibration

@ByVal
public org.bytedeco.opencv.opencv_core.Size readCalibration(@StdString
                                                                    String filename)

getCameraMatrix

@Const
 @ByRef
public org.bytedeco.opencv.opencv_core.Mat getCameraMatrix()

Returns the camera matrix used for the pose estimation.

getDistortionCoeffs

@Const
 @ByRef
public org.bytedeco.opencv.opencv_core.Mat getDistortionCoeffs()

Returns the distortion coefficients used for the pose estimation.