relativeCameraPose

Compute relative rotation and translation between camera poses

collapse all in page

Syntax

[relativeOrientation,relativeLocation] = relativeCameraPose(M,cameraParams,inlierPoints1,inlierPoints2)
[relativeOrientation,relativeLocation] = relativeCameraPose(M,cameraParams1,cameraParams2,inlierPoints1,inlierPoints2)
[relativeOrientation,relativeLocation,validPointsFraction] = relativeCameraPose(M, ___)

Description

[relativeOrientation,relativeLocation] = relativeCameraPose(M,cameraParams,inlierPoints1,inlierPoints2) returns the orientation and location of a calibrated camera relative to its previous pose. The two poses are related by M, which must be a fundamental, essential, affine2d, or projective2d matrix. The function computes the camera location up to scale and returns relativeLocation as a unit vector.

[relativeOrientation,relativeLocation] = relativeCameraPose(M,cameraParams1,cameraParams2,inlierPoints1,inlierPoints2) returns the orientation and location of the second camera relative to the first one.

[relativeOrientation,relativeLocation,validPointsFraction] = relativeCameraPose(M, ___) additionally returns the fraction of the inlier points that project in front of both cameras.

Input Arguments

collapse all

Fundamental, essential matrix, or a homography matrix, specified as a 3-by-3 matrix, an affine2d object, or a projective2d object containing a homography matrix. You can obtain the 3-by-3 matrix using one of the following functions:

Data Types: single | double

Camera parameters, specified as a cameraParameters or cameraIntrinsics object. You can return the cameraParameters object using the estimateCameraParameters function. The cameraParameters object contains the intrinsic, extrinsic, and lens distortion parameters of a camera.

Camera parameters for camera 1, specified as a cameraParameters or cameraIntrinsics object. You can return the cameraParameters object using the estimateCameraParameters function. The cameraParameters object contains the intrinsic, extrinsic, and lens distortion parameters of a camera.

Camera parameters for camera 2, specified as a cameraParameters or cameraIntrinsics object. You can return the cameraParameters object using the estimateCameraParameters function. The cameraParameters object contains the intrinsic, extrinsic, and lens distortion parameters of a camera.

Coordinates of corresponding points in view 1, specified as an M-by-2 matrix of M number of [x,y] coordinates, or as a SURFPoints, MSERRegions, or cornerPoints object. You can obtain these points using the estimateFundamentalMatrix function or the estimateEssentialMatrix.

Coordinates of corresponding points in view 2, specified as an M-by-2 matrix of M number of [x,y] coordinates, or as a SURFPoints, MSERRegions, or cornerPoints object. You can obtain these points using the estimateFundamentalMatrix function or the estimateEssentialMatrix.

Output Arguments

collapse all

Orientation of camera, returned as a 3-by-3 matrix. If you use only one camera, the matrix describes the orientation of the second camera pose relative to the first camera pose. If you use two cameras, the matrix describes the orientation of camera 2 relative to camera 1.

Data Types: single | double

Location of camera, returned as a 1-by-3 unit vector. If you use only one camera, the vector describes the location of the second camera pose relative to the first camera pose. If you use two cameras, the vector describes the location of camera 2 relative to camera 1.

Data Types: single | double

Fraction of valid inlier points that project in front of both cameras, returned as a scalar. If validPointsFraction is too small, e.g. less than 0.9, it can indicate that the fundamental matrix is incorrect.

Tips

  • You can compute the camera extrinsics, rotationMatrix and translationVector, corresponding to the camera pose, from relativeOrientation and relativeLocation: 

    [rotationMatrix,translationVector] = cameraPoseToExtrinsics(relativeOrientation,relativeLocation)
    The orientation of the previous camera pose is the identity matrix, eye(3), and its location is, [0,0,0].

  • You can then use rotationMatrix and translationVector as inputs to the cameraMatrix function.

  • You can compute four possible combinations of orientation and location from the input fundamental matrix. Three of the combinations are not physically realizable, because they project 3-D points behind one or both cameras. The relativeCameraPose function uses inlierPoints1 and inlierPoints2 to determine the realizable combination.

Extended Capabilities

See Also

| | | | | | | | |

Topics

Introduced in R2016b

Computer Vision Toolbox Documentation

Support