proj_match_points_ransac_guidedProjMatchPointsRansacGuidedProjMatchPointsRansacGuidedproj_match_points_ransac_guidedT_proj_match_points_ransac_guided🔗

Short description🔗

proj_match_points_ransac_guidedProjMatchPointsRansacGuidedProjMatchPointsRansacGuidedproj_match_points_ransac_guidedT_proj_match_points_ransac_guided — Compute a projective transformation matrix between two images by finding correspondences between points based on a known approximation of the projective transformation matrix.

Signature🔗

proj_match_points_ransac_guided( image Image1, image Image2, point.x Rows1, point.y Cols1, point.x Rows2, point.y Cols2, string GrayMatchMethod, integer MaskSize, hom_mat2d HomMat2DGuide, real DistanceTolerance, number MatchThreshold, string EstimationMethod, real DistanceThreshold, integer RandSeed, out hom_mat2d HomMat2D, out integer Points1, out integer Points2 )

Description🔗

Given a set of coordinates of characteristic points \((\textrm{Cols1},\textrm{Rows1})\) and \((\textrm{Cols2},\textrm{Rows2})\) in both input images Image1image1image_1 and Image2image2image_2, and given a known approximation HomMat2DGuidehomMat2DGuidehom_mat_2dguide for the transformation matrix between Image1image1image_1 and Image2image2image_2, proj_match_points_ransac_guidedProjMatchPointsRansacGuided automatically determines corresponding points and the homogeneous projective transformation matrix HomMat2DhomMat2Dhom_mat_2d that best transforms the corresponding points from the different images into each other. The characteristic points can, for example, be extracted with points_foerstnerPointsFoerstner or points_harrisPointsHarris. The approximation HomMat2DGuidehomMat2DGuidehom_mat_2dguide can, for example, be calculated with proj_match_points_ransacProjMatchPointsRansac on lower resolution versions of Image1image1image_1 and Image2image2image_2.

The transformation is determined in two steps: First, gray value correlations of mask windows around the input points in the first and the second image are determined and an initial matching between them is generated using the similarity of the windows in both images.

The size of the mask windows is MaskSize x MaskSize. Three metrics for the correlation can be selected. If GrayMatchMethodgrayMatchMethodgray_match_method has the value 'ssd'"ssd", the sum of the squared gray value differences is used, 'sad'"sad" means the sum of absolute differences, and 'ncc'"ncc" is the normalized cross correlation. For details please refer to binocular_disparityBinocularDisparity. The metric is minimized ('ssd'"ssd", 'sad'"sad") or maximized ('ncc'"ncc") over all possible point pairs. A thus found matching is only accepted if the value of the metric is below the value of MatchThresholdmatchThresholdmatch_threshold ('ssd'"ssd", 'sad'"sad") or above that value ('ncc'"ncc").

To increase the algorithm’s performance, the search area for the matching operations is limited based on the approximate transformation HomMat2DGuidehomMat2DGuidehom_mat_2dguide. Only points within a distance of DistanceTolerancedistanceTolerancedistance_tolerance around the transformed a point in Image2image2image_2 of a point in Image1image1image_1 via HomMat2DGuidehomMat2DGuidehom_mat_2dguide are considered for the matching.

Once the initial matching is complete, a randomized search algorithm (RANSAC) is used to determine the transformation matrix HomMat2DhomMat2Dhom_mat_2d. It tries to find the matrix that is consistent with a maximum number of correspondences. For a point to be accepted, its distance from the coordinates predicted by the transformation must not exceed the threshold DistanceThresholddistanceThresholddistance_threshold.

Once a choice has been made, the matrix is further optimized using all consistent points. For this optimization, the EstimationMethodestimationMethodestimation_method can be chosen to either be the slow but mathematically optimal 'gold_standard'"gold_standard" method or the faster 'normalized_dlt'"normalized_dlt". Here, the algorithms of vector_to_proj_hom_mat2dVectorToProjHomMat2d are used.

Point pairs that still violate the consistency condition for the final transformation are dropped, the matched points are returned as control values. Points1points1points_1 contains the indices of the matched input points from the first image, Points2points2points_2 contains the indices of the corresponding points in the second image.

The parameter RandSeedrandSeedrand_seed can be used to control the randomized nature of the RANSAC algorithm, and hence to obtain reproducible results. If RandSeedrandSeedrand_seed is set to a positive number, the operator yields the same result on every call with the same parameters because the internally used random number generator is initialized with the seed value. If RandSeedrandSeedrand_seed \(=\) 00, the random number generator is initialized with the current time. Hence, the results may not be reproducible in this case. The value set for the HALCON system variable 'seed_rand'"seed_rand" (see set_systemSetSystem) does not affect the results of proj_match_points_ransac_guidedProjMatchPointsRansacGuided.

Execution information🔗

Execution information

Multithreading type: reentrant (runs in parallel with non-exclusive operators).
Multithreading scope: global (may be called from any thread).
Processed without parallelization.

Parameters🔗

Image1image1image_1 (input_object) singlechannelimage → object (byte / uint2)HObject (byte / uint2)HImage (byte / uint2)HObject (byte / uint2)Hobject (byte / uint2)

Input image 1.

Image2image2image_2 (input_object) singlechannelimage → object (byte / uint2)HObject (byte / uint2)HImage (byte / uint2)HObject (byte / uint2)Hobject (byte / uint2)

Input image 2.

Rows1rows1rows_1 (input_control) point.x-array → (real / integer)HTuple (double / Hlong)HTuple (double / int / long)Sequence[Union[float, int]]Htuple (double / Hlong)

Row coordinates of characteristic points in image 1.

Cols1cols1cols_1 (input_control) point.y-array → (real / integer)HTuple (double / Hlong)HTuple (double / int / long)Sequence[Union[float, int]]Htuple (double / Hlong)

Column coordinates of characteristic points in image 1.

Rows2rows2rows_2 (input_control) point.x-array → (real / integer)HTuple (double / Hlong)HTuple (double / int / long)Sequence[Union[float, int]]Htuple (double / Hlong)

Row coordinates of characteristic points in image 2.

Cols2cols2cols_2 (input_control) point.y-array → (real / integer)HTuple (double / Hlong)HTuple (double / int / long)Sequence[Union[float, int]]Htuple (double / Hlong)

Column coordinates of characteristic points in image 2.

GrayMatchMethodgrayMatchMethodgray_match_method (input_control) string → (string)HTuple (HString)HTuple (string)strHtuple (char*)

Gray value comparison metric.

Default: 'ssd'"ssd"
List of values: 'ncc', 'sad', 'ssd'"ncc", "sad", "ssd"

MaskSizemaskSizemask_size (input_control) integer → (integer)HTuple (Hlong)HTuple (int / long)intHtuple (Hlong)

Size of gray value masks.

Default: 1010
Value range: MaskSize ≤ 90

HomMat2DGuidehomMat2DGuidehom_mat_2dguide (input_control) hom_mat2d → (real)HTuple (double)HHomMat2D, HTuple (double)Sequence[float]Htuple (double)

Approximation of the Homogeneous projective transformation matrix between the two images.

DistanceTolerancedistanceTolerancedistance_tolerance (input_control) real → (real)HTuple (double)HTuple (double)floatHtuple (double)

Tolerance for the matching search window.

Default: 20.020.0
Suggested values: 0.2, 0.5, 1.0, 2.0, 3.0, 5.0, 10.0, 20.0, 50.0

MatchThresholdmatchThresholdmatch_threshold (input_control) number → (integer / real)HTuple (Hlong / double)HTuple (int / long / double)Union[int, float]Htuple (Hlong / double)

Threshold for gray value matching.

Default: 1010
Suggested values: 10, 20, 50, 100, 0.9, 0.710, 20, 50, 100, 0.9, 0.7

EstimationMethodestimationMethodestimation_method (input_control) string → (string)HTuple (HString)HTuple (string)strHtuple (char*)

Transformation matrix estimation algorithm.

Default: 'normalized_dlt'"normalized_dlt"
List of values: 'gold_standard', 'normalized_dlt'"gold_standard", "normalized_dlt"

DistanceThresholddistanceThresholddistance_threshold (input_control) real → (real)HTuple (double)HTuple (double)floatHtuple (double)

Threshold for transformation consistency check.

Default: 0.20.2

RandSeedrandSeedrand_seed (input_control) integer → (integer)HTuple (Hlong)HTuple (int / long)intHtuple (Hlong)

Seed for the random number generator.

Default: 00

HomMat2DhomMat2Dhom_mat_2d (output_control) hom_mat2d → (real)HTuple (double)HHomMat2D, HTuple (double)Sequence[float]Htuple (double)

Homogeneous projective transformation matrix.

Points1points1points_1 (output_control) integer-array → (integer)HTuple (Hlong)HTuple (int / long)Sequence[int]Htuple (Hlong)

Indices of matched input points in image 1.

Points2points2points_2 (output_control) integer-array → (integer)HTuple (Hlong)HTuple (int / long)Sequence[int]Htuple (Hlong)

Indices of matched input points in image 2.

Example🔗

(HDevelop)


zoom_image_factor (Image1, Image1Zoomed, 0.5, 0.5, 'constant')
zoom_image_factor (Image2, Image2Zoomed, 0.5, 0.5, 'constant')
points_foerstner (Image1Zoomed, 1, 2, 3, 200, 0.3, 'gauss', 'false', \
                  Rows1, Cols1, _, _, _, _, _, _, _, _)
points_foerstner (Image2Zoomed, 1, 2, 3, 200, 0.3, 'gauss', 'false', \
                  Rows2, Cols2, _, _, _, _, _, _, _, _)
get_image_pointer1 (Image1Zoomed, Pointer, Type, Width, Height)
proj_match_points_ransac (Image1Zoomed, Image2Zoomed, Rows1, Cols1, \
                          Rows2, Cols2, 'ncc', 10, 0, 0, \
                          Height, Width, 0, 0.5, 'gold_standard', \
                          5, 0, HomMat2D, Points1, Points2)
hom_mat2d_scale_local (HomMat2D, 0.5, 0.5, HomMat2DGuide)
hom_mat2d_scale (HomMat2DGuide, 2, 2, 0, 0, HomMat2DGuide)
points_foerstner (Image1, 1, 2, 3, 200, 0.3, 'gauss', 'false', \
                  Rows1, Cols1, _, _, _, _, _, _, _, _)
points_foerstner (Image2, 1, 2, 3, 200, 0.3, 'gauss', 'false', \
                  Rows2, Cols2, _, _, _, _, _, _, _, _)
proj_match_points_ransac_guided (Image1, Image2, Rows1, Cols1, \
                                 Rows2, Cols2, 'ncc', 10, \
                                 HomMat2DGuide, 40, 0.5, \
                                 'gold_standard', 10, 0, HomMat2D, \
                                 Points1, Points2)

Combinations with other operators🔗

Combinations

Possible predecessors

points_foerstnerPointsFoerstner, points_harrisPointsHarris

Possible successors

projective_trans_imageProjectiveTransImage, projective_trans_image_sizeProjectiveTransImageSize, projective_trans_regionProjectiveTransRegion, projective_trans_contour_xldProjectiveTransContourXld, projective_trans_point_2dProjectiveTransPoint2d, projective_trans_pixelProjectiveTransPixel

Alternatives

hom_vector_to_proj_hom_mat2dHomVectorToProjHomMat2d, vector_to_proj_hom_mat2dVectorToProjHomMat2d

References🔗

Richard Hartley, Andrew Zisserman: ``Multiple View Geometry in Computer Vision’‘; Cambridge University Press, Cambridge; 2000.

Olivier Faugeras, Quang-Tuan Luong: ``The Geometry of Multiple Images: The Laws That Govern the Formation of Multiple Images of a Scene and Some of Their Applications’‘; MIT Press, Cambridge, MA; 2001.

Module🔗

Matching