OpenCV  3.4.2
Open Source Computer Vision
Classes | Functions
RGB-Depth Processing

Classes

class  cv::linemod::ColorGradient
 Modality that computes quantized gradient orientations from a color image. More...
 
class  cv::rgbd::DepthCleaner
 
class  cv::linemod::DepthNormal
 Modality that computes quantized surface normals from a dense depth map. More...
 
class  cv::linemod::Detector
 Object detector using the LINE template matching algorithm with any set of modalities. More...
 
struct  cv::linemod::Feature
 Discriminant feature described by its location and label. More...
 
class  cv::rgbd::ICPOdometry
 
struct  cv::linemod::Match
 Represents a successful template match. More...
 
class  cv::linemod::Modality
 Interface for modalities that plug into the LINE template matching representation. More...
 
class  cv::rgbd::Odometry
 
struct  cv::rgbd::OdometryFrame
 
class  cv::linemod::QuantizedPyramid
 Represents a modality operating over an image pyramid. More...
 
struct  cv::rgbd::RgbdFrame
 
class  cv::rgbd::RgbdICPOdometry
 
class  cv::rgbd::RgbdNormals
 
class  cv::rgbd::RgbdOdometry
 
class  cv::rgbd::RgbdPlane
 
struct  cv::linemod::Template
 

Functions

 cv::linemod::QuantizedPyramid::Candidate::Candidate (int x, int y, int label, float score)
 
 cv::linemod::Feature::Feature (int x, int y, int label)
 
 cv::linemod::Match::Match (int x, int y, float similarity, const String &class_id, int template_id)
 
void cv::linemod::colormap (const Mat &quantized, Mat &dst)
 Debug function to colormap a quantized image for viewing. More...
 
void cv::rgbd::depthTo3d (InputArray depth, InputArray K, OutputArray points3d, InputArray mask=noArray())
 
void cv::rgbd::depthTo3dSparse (InputArray depth, InputArray in_K, InputArray in_points, OutputArray points3d)
 
Ptr< linemod::Detectorcv::linemod::getDefaultLINE ()
 Factory function for detector using LINE algorithm with color gradients. More...
 
Ptr< linemod::Detectorcv::linemod::getDefaultLINEMOD ()
 Factory function for detector using LINE-MOD algorithm with color gradients and depth normals. More...
 
bool cv::rgbd::isValidDepth (const float &depth)
 
bool cv::rgbd::isValidDepth (const double &depth)
 
bool cv::rgbd::isValidDepth (const short int &depth)
 
bool cv::rgbd::isValidDepth (const unsigned short int &depth)
 
bool cv::rgbd::isValidDepth (const int &depth)
 
bool cv::rgbd::isValidDepth (const unsigned int &depth)
 
void cv::rgbd::registerDepth (InputArray unregisteredCameraMatrix, InputArray registeredCameraMatrix, InputArray registeredDistCoeffs, InputArray Rt, InputArray unregisteredDepth, const Size &outputImagePlaneSize, OutputArray registeredDepth, bool depthDilation=false)
 
void cv::rgbd::rescaleDepth (InputArray in, int depth, OutputArray out)
 
void cv::rgbd::warpFrame (const Mat &image, const Mat &depth, const Mat &mask, const Mat &Rt, const Mat &cameraMatrix, const Mat &distCoeff, OutputArray warpedImage, OutputArray warpedDepth=noArray(), OutputArray warpedMask=noArray())
 

Detailed Description

Function Documentation

§ Candidate()

cv::linemod::QuantizedPyramid::Candidate::Candidate ( int  x,
int  y,
int  label,
float  score 
)
inline

§ Feature()

cv::linemod::Feature::Feature ( int  x,
int  y,
int  label 
)
inline

§ Match()

cv::linemod::Match::Match ( int  x,
int  y,
float  similarity,
const String class_id,
int  template_id 
)
inline

§ colormap()

void cv::linemod::colormap ( const Mat quantized,
Mat dst 
)
Python:
dst=cv.linemod.colormap(quantized[, dst])

Debug function to colormap a quantized image for viewing.

§ depthTo3d()

void cv::rgbd::depthTo3d ( InputArray  depth,
InputArray  K,
OutputArray  points3d,
InputArray  mask = noArray() 
)
Python:
points3d=cv.rgbd.depthTo3d(depth, K[, points3d[, mask]])

Converts a depth image to an organized set of 3d points. The coordinate system is x pointing left, y down and z away from the camera

Parameters
depththe depth image (if given as short int CV_U, it is assumed to be the depth in millimeters (as done with the Microsoft Kinect), otherwise, if given as CV_32F or CV_64F, it is assumed in meters)
KThe calibration matrix
points3dthe resulting 3d points. They are of depth the same as depth if it is CV_32F or CV_64F, and the depth of K if depth is of depth CV_U
maskthe mask of the points to consider (can be empty)

§ depthTo3dSparse()

void cv::rgbd::depthTo3dSparse ( InputArray  depth,
InputArray  in_K,
InputArray  in_points,
OutputArray  points3d 
)
Python:
points3d=cv.rgbd.depthTo3dSparse(depth, in_K, in_points[, points3d])
Parameters
depththe depth image
in_K
in_pointsthe list of xy coordinates
points3dthe resulting 3d points

§ getDefaultLINE()

Ptr<linemod::Detector> cv::linemod::getDefaultLINE ( )
Python:
retval=cv.linemod.getDefaultLINE()

Factory function for detector using LINE algorithm with color gradients.

Default parameter settings suitable for VGA images.

§ getDefaultLINEMOD()

Ptr<linemod::Detector> cv::linemod::getDefaultLINEMOD ( )
Python:
retval=cv.linemod.getDefaultLINEMOD()

Factory function for detector using LINE-MOD algorithm with color gradients and depth normals.

Default parameter settings suitable for VGA images.

§ isValidDepth() [1/6]

bool cv::rgbd::isValidDepth ( const float &  depth)
inline

Checks if the value is a valid depth. For CV_16U or CV_16S, the convention is to be invalid if it is a limit. For a float/double, we just check if it is a NaN

Parameters
depththe depth to check for validity

§ isValidDepth() [2/6]

bool cv::rgbd::isValidDepth ( const double &  depth)
inline

§ isValidDepth() [3/6]

bool cv::rgbd::isValidDepth ( const short int &  depth)
inline

§ isValidDepth() [4/6]

bool cv::rgbd::isValidDepth ( const unsigned short int &  depth)
inline

§ isValidDepth() [5/6]

bool cv::rgbd::isValidDepth ( const int &  depth)
inline

§ isValidDepth() [6/6]

bool cv::rgbd::isValidDepth ( const unsigned int &  depth)
inline

§ registerDepth()

void cv::rgbd::registerDepth ( InputArray  unregisteredCameraMatrix,
InputArray  registeredCameraMatrix,
InputArray  registeredDistCoeffs,
InputArray  Rt,
InputArray  unregisteredDepth,
const Size outputImagePlaneSize,
OutputArray  registeredDepth,
bool  depthDilation = false 
)
Python:
registeredDepth=cv.rgbd.registerDepth(unregisteredCameraMatrix, registeredCameraMatrix, registeredDistCoeffs, Rt, unregisteredDepth, outputImagePlaneSize[, registeredDepth[, depthDilation]])

Registers depth data to an external camera Registration is performed by creating a depth cloud, transforming the cloud by the rigid body transformation between the cameras, and then projecting the transformed points into the RGB camera.

uv_rgb = K_rgb * [R | t] * z * inv(K_ir) * uv_ir

Currently does not check for negative depth values.

Parameters
unregisteredCameraMatrixthe camera matrix of the depth camera
registeredCameraMatrixthe camera matrix of the external camera
registeredDistCoeffsthe distortion coefficients of the external camera
Rtthe rigid body transform between the cameras. Transforms points from depth camera frame to external camera frame.
unregisteredDepththe input depth data
outputImagePlaneSizethe image plane dimensions of the external camera (width, height)
registeredDepththe result of transforming the depth into the external camera
depthDilationwhether or not the depth is dilated to avoid holes and occlusion errors (optional)

§ rescaleDepth()

void cv::rgbd::rescaleDepth ( InputArray  in,
int  depth,
OutputArray  out 
)
Python:
out=cv.rgbd.rescaleDepth(in, depth[, out])

If the input image is of type CV_16UC1 (like the Kinect one), the image is converted to floats, divided by 1000 to get a depth in meters, and the values 0 are converted to std::numeric_limits<float>::quiet_NaN() Otherwise, the image is simply converted to floats

Parameters
inthe depth image (if given as short int CV_U, it is assumed to be the depth in millimeters (as done with the Microsoft Kinect), it is assumed in meters)
depththe desired output depth (floats or double)
outThe rescaled float depth image

§ warpFrame()

void cv::rgbd::warpFrame ( const Mat image,
const Mat depth,
const Mat mask,
const Mat Rt,
const Mat cameraMatrix,
const Mat distCoeff,
OutputArray  warpedImage,
OutputArray  warpedDepth = noArray(),
OutputArray  warpedMask = noArray() 
)
Python:
warpedImage, warpedDepth, warpedMask=cv.rgbd.warpFrame(image, depth, mask, Rt, cameraMatrix, distCoeff[, warpedImage[, warpedDepth[, warpedMask]]])

Warp the image: compute 3d points from the depth, transform them using given transformation, then project color point cloud to an image plane. This function can be used to visualize results of the Odometry algorithm.

Parameters
imageThe image (of CV_8UC1 or CV_8UC3 type)
depthThe depth (of type used in depthTo3d fuction)
maskThe mask of used pixels (of CV_8UC1), it can be empty
RtThe transformation that will be applied to the 3d points computed from the depth
cameraMatrixCamera matrix
distCoeffDistortion coefficients
warpedImageThe warped image.
warpedDepthThe warped depth.
warpedMaskThe warped mask.