Tracking method of three-dimensional mark point sequence and system thereof
A tracking system and marking point technology, which is applied in the field of computer vision and pattern recognition, can solve the problems of manual repair data repair errors, difficult personnel, and inability to fully guarantee the correct identification of marking points, etc.
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
specific Embodiment approach 1
[0130] This specific embodiment is a method based on a single-object marker image.
[0131] This method can be used if there is at least one feature point on the rigid body where the feature point corresponding to the lost 3D marker point has a corresponding reconstructed and identified 3D marker point.
[0132] The rigid body where the feature points corresponding to the lost 3D marker points are located is the object rigid body.
[0133] Step S511 , searching for 2D marker points that have not been used to reconstruct the identified 3D marker points in all images of the current frame.
[0134] Step S512, for each two-dimensional marker point, calculate the three-dimensional back-projection ray of the two-dimensional marker point starting from the optical center of the camera corresponding to the image where the two-dimensional marker point is located.
[0135] The calculation method is if the camera matrix is P 3×4 =[H 3×3 ,p 4 ], then the straight line passing through...
specific Embodiment approach 2
[0148] This specific implementation is a method based on the translation vector of a marked point in a rigid body.
[0149] This method can be used if there is at least one feature point on the rigid body where the feature point corresponding to the lost 3D marker point has a corresponding reconstructed and recognized 3D marker point in the current frame.
[0150] Such as image 3 As shown in (b), corresponding to the feature point f in the previous frame 1 The identified 3D marker points are m 1 , corresponding to the feature point f j The identified 3D marker points are m j , the vector between the two 3D markers is feature point f 1 The coordinates of the corresponding identified 3D marker points in the current frame are Then the feature point f j The corresponding candidate 3D markers in the current frame are The candidate three-dimensional marker points are added to the set of candidate supplementary three-dimensional marker points.
[0151] The source of this...
specific Embodiment approach 3
[0152] This particular embodiment is a method based on rigid body transformation.
[0153] This method can be used if there are at least three feature points on the rigid body where the feature points corresponding to the missing 3D marker points have corresponding reconstructed and identified 3D marker points in the current frame.
[0154] Since the rigid body structure basically remains unchanged during the movement, if the marker point m is lost 1 At least three 3D markers on the rigid body have been reconstructed and identified, then the rigid body transformation R, t can be calculated by the coordinates of the identified 3D markers in the rigid body in the current frame and the coordinates of the previous frame, and the lost 3D mark point m 1 Candidate supplementary point results of can be calculated by the following formula
[0155] M 1 k = R M j k ...
PUM
Login to View More Abstract
Description
Claims
Application Information
Login to View More 