Generate transformation maps that describe the mapping of the images of a binocular camera pair to a common rectified image plane.

Given a pair of stereo images, rectification determines a transformation of each image plane in a way that pairs of conjugate epipolar lines become collinear and parallel to the horizontal image axes. The rectified epipolar images can be thought of as acquired by a new stereo rig, obtained by rotating the original cameras. The camera centers of this virtual rig are maintained whereas the image planes coincide, which means that the focal lengths are set equal, and the optical axes parallel.

To achieve the transformation map for epipolar images gen_binocular_rectification_map requires the internal camera parameters CamParam1 of camera 1 and CamParam2 of camera 2, as well as the relative pose RelPose defining a point transformation from camera 2 to camera 1. These parameters can be obtained, e.g., from the operator binocular_calibration.

The projection onto a common plane has many degrees of freedom which are implicitly restricted by selecting a certain method in Method (currently only one method available):

'geometric' specifies the orientation of the common image
plane by the cross product of the base line and the line of
intersection of the original image planes. The new focal lengths of
the rectified cameras are determined in such a way as the old prinzipal
points have the same distance to the new common image plane.

Similar to gen_image_to_world_plane_map the parameter Interpolation specifies whether bilinear interpolation ('bilinear') should be applied between the pixels in the input image or the gray value of the nearest neighboring pixel should be taken ('none'). The size and resolution of the maps and of the transformed images can be adjusted by the SubSampling parameter which applies a sub-sampling factor to the original images.

The mapping functions for the images of camera 1 and camera 2 are returned in the images Map1 and Map2. If Interpolation is set to 'none', both maps consist of one single-channel image which contains the linear coordinate of the pixel of the respective input image that is the nearest neighbor of the transformed coordinate. In case of bilinear interpolation, each map contains one five-channel image. The first channel contains for each pixel of the respective map the linear coordinate of the pixel in the respective input image that is in the upper left position with respect to the transformed coordinate. The remaining four channels of each map contain the weights of the four neighboring pixels of the transformed coordinates which are used for the bilinear interpolation. The mapping of the channel numbers to the neighboring pixels is as follows:

                +---+---+
                | 2 | 3 |
                +---+---+
                | 4 | 5 |
                +---+---+

In addition, gen_binocular_rectification_map returns the modified internal and external camera parameters of the rectified stereo rig. CamParamRect1 and CamParamRect2 contain the modified internal parameters of camera 1 and camera 2, respectively. The rotation of the rectified camera in relation to the original camera is specified by CamPoseRect1 and CamPoseRect2, respectively. Finally, RelPoseRect returns the modified relative pose of the rectified camera system 2 in relation to the rectified camera system 1 defining a translation in x only. Generally, the transformations are defined in a way that the rectified camera 1 is left of the rectified camera 2. This means that the optical center of camera 2 has a positive x coordinate of the rectified coordinate system of camera 1.

// ...
// read the internal and external stereo parameters
read_cam_par ('cam_left.dat', CamParam1)
read_cam_par ('cam_right.dat', CamParam2)
read_pose ('relpos.dat', RelPose)

//  the stereo images
gen_binocular_rectification_map (Map1, Map2, CamParam1, CamParam2, RelPose, 
                                 'geometric', 'bilinear', CamParRect1,
                                 CamParamRect2, Cam1PoseRect1, Cam2PoseRect2,
                                 RelPoseRect)

while 1
  grab_image_async (Image1, FGHandle1, -1)
  map_image (Image1, Map1, ImageMapped1)

  grab_image_async (Image2, FGHandle2, -1)
  map_image (Image2, Map2, ImageMapped2)

  binocular_disparity(ImageMapped1, ImageMapped2, Disparity, Score, 'sad',
                      11, 11, 20, -40, 20, 2, 25, 'left_right_check',
                      'interpolation')
endwhile

gen_binocular_rectification_map returns 2 (H_MSG_TRUE) if all parameter values are correct. If necessary, an exception handling is raised.

gen_binocular_rectification_map is reentrant and processed without parallelization.

binocular_calibration

map_image

gen_image_to_world_plane_map

map_image, gen_image_to_world_plane_map, contour_to_world_plane_xld, image_points_to_world_plane

3D Metrology