match_fundamental_matrix_ransac — Compute the fundamental matrix for a pair of stereo images by automatically finding correspondences between image points.
match_fundamental_matrix_ransac(Image1, Image2 : : Rows1, Cols1, Rows2, Cols2, GrayMatchMethod, MaskSize, RowMove, ColMove, RowTolerance, ColTolerance, Rotation, MatchThreshold, EstimationMethod, DistanceThreshold, RandSeed : FMatrix, CovFMat, Error, Points1, Points2)
Given a set of coordinates of characteristic points (Rows1,Cols1) and (Rows2,Cols2) in the stereo images Image1 and Image2, match_fundamental_matrix_ransac automatically finds the correspondences between the characteristic points and determines the geometry of the stereo setup. For unknown cameras the geometry of the stereo setup is represented by the fundamental matrix FMatrix and all corresponding points have to fulfill the epipolar constraint, namely:
T / Cols2 \ / Cols1 \ | Rows2 | * FMatrix * | Rows1 | = 0 . \ 1 / \ 1 /
Note the column/row ordering in the point coordinates: because the fundamental matrix encodes the projective relation between two stereo images embedded in 3D space, the x/y notation has to be compliant with the camera coordinate system. So, (x,y) coordinates correspond to (column,row) pairs.
The matching process is based on characteristic points, which can be extracted with point operators like points_foerstner or points_harris. The matching itself is carried out 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. Then, the RANSAC algorithm is applied to find the fundamental matrix that maximizes the number of correspondences under the epipolar constraint.
The size of the mask windows is MaskSize x MaskSize. Three metrics for the correlation can be selected. If GrayMatchMethod has the value 'ssd', the sum of the squared gray value differences is used, 'sad' means the sum of absolute differences, and 'ncc' is the normalized cross correlation. For details please refer to binocular_disparity. The metric is minimized ('ssd', 'sad') or maximized ('ncc') over all possible point pairs. A thus found matching is only accepted if the value of the metric is below the value of MatchThreshold ('ssd', 'sad') or above that value ('ncc').
To increase the speed of the algorithm the search area for the matchings can be limited. Only points within a window of 2*RowTolerance x 2*ColTolerance points are considered. The offset of the center of the search window in the second image with respect to the position of the current point in the first image is given by RowMove and ColMove.
If the second camera is rotated around the optical axis with respect to the first camera the parameter Rotation may contain an estimate for the rotation angle or an angle interval in radians. A good guess will increase the quality of the gray value matching. If the actual rotation differs too much from the specified estimate the matching will typically fail. In this case, an angle interval should be specified and Rotation is a tuple with two elements. The larger the given interval the slower the operator is since the RANSAC algorithm is run over all angle increments within the interval.
After the initial matching is completed a randomized search algorithm (RANSAC) is used to determine the fundamental matrix FMatrix. It tries to find the matrix that is consistent with a maximum number of correspondences. For a point to be accepted, the distance to its corresponding epipolar line must not exceed the threshold DistanceThreshold.
The parameter EstimationMethod decides whether the relative orientation between the cameras is of a special type and which algorithm is to be applied for its computation. If EstimationMethod is either 'normalized_dlt' or 'gold_standard' the relative orientation is arbitrary. If left and right camera are identical and the relative orientation between them is a pure translation then choose EstimationMethod equal to 'trans_normalized_dlt' or 'trans_gold_standard'. The typical application for this special motion case is the scenario of a single fixed camera looking onto a moving conveyor belt. In order to get a unique solution in the correspondence problem the minimum required number of corresponding points is eight in the general case and three in the special, translational case.
The fundamental matrix is computed by a linear algorithm if 'normalized_dlt' or 'trans_normalized_dlt' is chosen. With 'gold_standard' or 'trans_gold_standard' the algorithm gives a statistically optimal result, and returns as well the covariance of the fundamental matrix CovFMat. Here, 'normalized_dlt' and 'gold_standard' stand for direct-linear-transformation and gold-standard-algorithm respectively.
The value Error indicates the overall quality of the estimation procedure and is the mean euclidian distance in pixels between the points and their corresponding epipolar lines.
Point pairs consistent with the mentioned constraints are considered to be in correspondences. Points1 contains the indices of the matched input points from the first image and Points2 contains the indices of the corresponding points in the second image.
The parameter RandSeed can be used to control the randomized nature of the RANSAC algorithm, and hence to obtain reproducible results. If RandSeed 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 RandSeed. If RandSeed = 0 the random number generator is initialized with the current time. In this case the results may not be reproducible.
Input image 1.
Input image 2.
Row coordinates of characteristic points in image 1.
Restriction: length(Rows1) >= 8 || length(Rows1) >= 3
Column coordinates of characteristic points in image 1.
Restriction: length(Cols1) == length(Rows1)
Row coordinates of characteristic points in image 2.
Restriction: length(Rows2) >= 8 || length(Rows2) >= 3
Column coordinates of characteristic points in image 2.
Restriction: length(Cols2) == length(Rows2)
Gray value comparison metric.
Default value: 'ssd'
List of values: 'ncc', 'sad', 'ssd'
Size of gray value masks.
Default value: 10
Typical range of values: 3 ≤ MaskSize ≤ 15
Restriction: MaskSize >= 1
Average row coordinate shift of corresponding points.
Default value: 0
Typical range of values: 0 ≤ RowMove ≤ 200
Average column coordinate shift of corresponding points.
Default value: 0
Typical range of values: 0 ≤ ColMove ≤ 200
Half height of matching search window.
Default value: 200
Typical range of values: 50 ≤ RowTolerance ≤ 200
Restriction: RowTolerance >= 1
Half width of matching search window.
Default value: 200
Typical range of values: 50 ≤ ColTolerance ≤ 200
Restriction: ColTolerance >= 1
Estimate of the relative orientation of the right image with respect to the left image.
Default value: 0.0
Suggested values: 0.0, 0.1, -0.1, 0.7854, 1.571, 3.142
Threshold for gray value matching.
Default value: 10
Suggested values: 10, 20, 50, 100, 0.9, 0.7
Algorithm for the computation of the fundamental matrix and for special camera orientations.
Default value: 'normalized_dlt'
List of values: 'gold_standard', 'normalized_dlt', 'trans_gold_standard', 'trans_normalized_dlt'
Maximal deviation of a point from its epipolar line.
Default value: 1
Typical range of values: 0.5 ≤ DistanceThreshold ≤ 5
Restriction: DistanceThreshold > 0
Seed for the random number generator.
Default value: 0
Computed fundamental matrix.
9x9 covariance matrix of the fundamental matrix.
Root-Mean-Square of the epipolar distance error.
Indices of matched input points in image 1.
Indices of matched input points in image 2.
match_essential_matrix_ransac, match_rel_pose_ransac, proj_match_points_ransac
Richard Hartley, Andrew Zisserman: “Multiple View Geometry in
Computer Vision”; Cambridge University Press, Cambridge; 2003.
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.