CN116863086A - Rigid body stable reconstruction method for optical motion capture system - Google Patents
Rigid body stable reconstruction method for optical motion capture system Download PDFInfo
- Publication number
- CN116863086A CN116863086A CN202311129240.7A CN202311129240A CN116863086A CN 116863086 A CN116863086 A CN 116863086A CN 202311129240 A CN202311129240 A CN 202311129240A CN 116863086 A CN116863086 A CN 116863086A
- Authority
- CN
- China
- Prior art keywords
- points
- point
- cameras
- candidate
- camera
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 92
- 230000033001 locomotion Effects 0.000 title claims abstract description 51
- 230000003287 optical effect Effects 0.000 title claims abstract description 48
- 239000003550 marker Substances 0.000 claims description 44
- 238000012937 correction Methods 0.000 claims description 10
- 238000012163 sequencing technique Methods 0.000 claims description 10
- 230000001174 ascending effect Effects 0.000 claims description 3
- 238000012217 deletion Methods 0.000 claims description 2
- 230000037430 deletion Effects 0.000 claims description 2
- 230000002159 abnormal effect Effects 0.000 abstract description 8
- 206010044565 Tremor Diseases 0.000 description 25
- 239000011159 matrix material Substances 0.000 description 16
- 101100459261 Cyprinus carpio mycb gene Proteins 0.000 description 15
- 101100459256 Cyprinus carpio myca gene Proteins 0.000 description 10
- 238000010586 diagram Methods 0.000 description 9
- 238000004364 calculation method Methods 0.000 description 3
- 238000003384 imaging method Methods 0.000 description 3
- 238000012216 screening Methods 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 2
- 230000000903 blocking effect Effects 0.000 description 2
- 230000008054 signal transmission Effects 0.000 description 2
- 230000004075 alteration Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000005021 gait Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 239000004576 sand Substances 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T17/00—Three dimensional [3D] modelling, e.g. data description of 3D objects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/10—Geometric CAD
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2111/00—Details relating to CAD techniques
- G06F2111/04—Constraint-based CAD
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T10/00—Road transport of goods or passengers
- Y02T10/10—Internal combustion engine [ICE] based vehicles
- Y02T10/40—Engine management systems
Abstract
The invention relates to the technical field of rigid body stable reconstruction of an optical motion capture system, and discloses a rigid body stable reconstruction method of the optical motion capture system, which comprises the following steps: 1) By using the method of eliminating overlapped polar lines by utilizing the included angles and the distances of the polar lines, the abnormal situation of polar line overlapping is identified and eliminated, and the problem of error points in rigid body reconstruction is reduced. 2) The multi-view characteristic point matching method realizes that the high-priority camera is firstly used as the reference camera to perform multi-view characteristic point matching, and the reference points which are preferentially matched can be aligned with all cameras of the same matching group, so that the problem that the matching group omits the cameras, and leakage points and repeated points appear during reconstruction is solved. And the integrity and stability of rigid body reconstruction are improved.
Description
Technical Field
The invention relates to the technical field of optical motion capture systems, in particular to a rigid body stable reconstruction method of an optical motion capture system.
Background
Optical motion capture has become an important research branch in motion capture technology by virtue of the advantages of high acquisition precision, real-time feedback and the like. Common optical motion capture is mostly based on computer vision principle, and theoretically, as long as a point in space is visible to two cameras at the same time, the position of the point in space at the same time can be determined according to the images and camera parameters shot by the two cameras at the same time, and when the cameras continuously shoot at a sufficiently high speed, the three-dimensional motion trail of the point can be obtained from the image sequence. After being identified by a computer, the three-dimensional space coordinate data can be applied to the fields of digital sand tables, animation production, gait analysis, biomechanics, human engineering and the like.
The optical motion capture system adopting the marking point often comprises functional units such as an optical marking point, a motion capture camera (for short, a motion capture camera), a signal transmission device, a data processing center and the like. When a moving object is stuck with a mark point and moves, a plurality of dynamic capturing cameras can shoot the moving object at the same time from different angles, acquire images of the mark point on the moving object, and preprocess the shot images to acquire two-dimensional position information of the mark point; meanwhile, the signal transmission equipment transmits the two-dimensional position information of the mark point to the data processing center in real time, and the data processing center calculates the three-dimensional space coordinate (short for three-dimensional reconstruction) of the mark point according to a preset algorithm (such as a stereoscopic vision matching algorithm and the like), so that the motion track of the moving object (consisting of a plurality of mark points and short for rigid bodies) is obtained, and the space positioning of the moving object is realized.
Multi-view feature point matching is the biggest challenge in achieving rigid body reconstruction, which refers to the process of searching multiple views for projected points of the same three-dimensional spatial point in these views, which are called matching points or corresponding points. For marker-based optical motion capture techniques, the goal of multi-view feature point matching is to assign the same marker points captured by the cameras at the same time to the same matching group. Due to the lack of image feature information support, multi-view feature point matching based on position information mostly depends on epipolar geometry theory. In the multi-view feature point matching process, errors caused by various image noises cause the phenomena of miss-matching and mismatching to be easily generated in the multi-view feature point matching process. The missing matching and the mismatching not only can reduce the three-dimensional reconstruction precision of the rigid body points, but also can cause the problems of missing points, wrong points, repeated points and trembling points in the rigid body reconstruction.
When the optical motion capturing system reconstructs the three-dimensional coordinates of the rigid points, the problems of missing points, misplacement points, repeated points and trembling points are often encountered. These problems can cause errors and deviations in the three-dimensional coordinate calculation of the rigid body, so that obvious errors exist in the attitude calculation of the rigid body, and the stability of the motion capture system is affected, and therefore, a stable reconstruction method of the rigid body of the optical motion capture system is provided to solve the problems.
Disclosure of Invention
Aiming at the defects of the prior art, the invention provides a rigid body stable reconstruction method of an optical motion capture system, which has the advantages of improving the stability of rigid body reconstruction and the like, and solves the problems of leakage points, error points, repeated points and shake points in the reconstruction of the optical motion capture system.
In order to achieve the purpose of improving the stability of rigid body reconstruction, the invention provides the following technical scheme: a rigid body stable reconstruction method of an optical motion capture system comprises the following steps:
1) Removing overlapping epipolar lines includes:
a1: computing polar equation 1 for all points of one camera in other cameras 1 、1 2 、...、1 i ；
A2: the overlapping of polar lines is compared in pairs, the included angle between polar line equations is calculated first, if the included angle is smaller than T angle Judging that the polar lines are approximately parallel, continuously judging the distance between the polar lines, calculating the minimum distance of the straight line on the image, and if the minimum distance is smaller than the threshold value T distanse Then the polar line equation is judged to overlap, wherein the threshold T of the minimum distance distanse The average diameter of the marker point on the image;
2) The feature point matching includes:
b1: sequencing the priority of the reference cameras and the reference points;
b2: sequencing the priority of the candidate cameras;
b3: iterative three-camera spatial registration;
the iterative three-camera spatial registration in B3 is characterized by comprising the following steps:
b3-1: firstly, sequentially selecting datum points of a datum camera according to a priority order, then selecting a cooperative camera and a cooperative point corresponding to the datum points, and finally selecting a camera to be matched and a point to be matched according to the priority order;
b3-2: three-camera space matching, firstly calculating the intersection point P of a datum point and a cooperation point in a camera polar line to be matched cross The method comprises the steps of carrying out a first treatment on the surface of the Then calculate the candidate point and P cross Setting a threshold value threshold1 finally, if d is smaller than or equal to threshold1, then matching successfully, wherein the datum point, the cooperation point and the point to be matched form a candidate matching group, and the cameras and the points in the candidate matching group do not participate in the subsequent matching process;
b3-3: repeating the processes from B3-1 to B3-2, traversing the cameras to be matched according to the priority of the candidate cameras, sequentially updating the cameras to be matched and the points to be matched according to the priority sequence, repeating the three-phase space matching process of S2, and adding the successfully matched cameras to be matched and the points to be matched into the candidate matching group until all the cameras to be matched are matched, so as to obtain a final matching group;
b3-4: repeating the processes from B3-1 to B3-3, traversing each datum point in the cameras according to the datum point priority to match, and obtaining a matching group of each datum point of the first priority datum camera;
b3-5: repeating the processes from B3-1 to B3-4, traversing each reference camera according to the priority of the reference camera to obtain a matching group of each reference camera;
b3-6: comparing the matched groups obtained in the step B3-5 in pairs, merging the two matched groups if the matched groups have intersections, judging as invalid matched groups if the number of points of the matched groups is less than 3, and deleting;
3) The three-dimensional point reconstruction includes:
c1: performing three-dimensional reconstruction by using the matching group;
c2: removing outliers in the reconstructed point set;
and C3: merging the repeated points in the reconstructed point set;
4) The correction of the positions of the shake points of the rigid body marker points comprises the following steps:
d1: recording the position information CW of the rigid body of the previous frame, wherein the position information CW comprises the positions W0, W1 and Wn of n marker points, and the position information CW of the rigid body of the current frame comprises the positions W0, W1 and Wn of n marker points;
d2: traversing each point in the CW, searching the nearest point in the CW, calculating the nearest distance d, taking the radius of the marker point as a threshold value threthold2, and judging that the marker point moves if d > threthold 2;
d3: if the marker point of the current frame part moves, judging that the phenomenon of the shaking point exists, and replacing the position of the current frame of the marker by the position of the nearest point of the previous frame.
Further, the reference camera and the reference point in the B1 are ranked in priority, which is characterized by comprising the following steps: b1-1: sequencing reference cameras according to the effective reference points, wherein the reference points are at least provided with candidate points in two cameras, marking the candidate points as effective reference points, calculating the effective reference points of each camera, and the higher the number of the points is, the higher the priority of the cameras is; if the number of the effective reference points is 0, deleting the effective reference points from the sequence, and if the number of the effective reference points of a plurality of cameras is the same, sorting the number of the candidate cameras, and B1-2: according to the sum of the number of the candidate cameras, sorting the reference cameras, wherein the reference cameras have candidate points on other cameras, the other cameras are the candidate cameras of the reference points, the higher the sum of the number of the candidate cameras of all the reference points in one reference camera is, the higher the priority of the reference cameras is, if the number of the effective reference points is the same in the sorting of B1-1, the sorting is carried out according to the number of the candidate cameras, and B1-3: sorting the reference points according to the number of the candidate cameras, sorting the reference points of the same camera according to the descending order of the number of the candidate cameras in order to ensure that the reference points matched firstly find the whole candidate points as far as possible, and deleting the reference points from the sequence if the number of the candidate cameras is 0.
Further, the priority ordering of the candidate cameras in the B2 is characterized by comprising the following specific steps: b2-1: sorting the candidate cameras according to the sum of the distances from the candidate points to the polar lines, calculating the sum of the distances from the candidate points of the candidate cameras of each datum point, sorting in ascending order to obtain the serial numbers of the candidate cameras, and B2-2: the camera with the highest priority in the candidate cameras is used as a cooperative camera, and other cameras are used as cameras to be matched.
Further, the three-dimensional reconstruction is performed in C1 by using the matching group, and is characterized by comprising the following steps:
c1-1: reconstructing the points in the matching group in pairs by using triangulation to obtain candidate three-dimensional points, re-projecting the three-dimensional points onto other cameras of the matching group, and calculating re-projection errors on the other cameras; if the error is larger than the diameter of the marker point, judging that the error point is reconstructed, and eliminating; finally, taking the average points of the pairwise reconstructed three-dimensional points as final three-dimensional reconstruction points;
c1-2: each matching group reconstructs a three-dimensional point, and all the three-dimensional points form a reconstruction point set.
Further, the C2 method for removing outliers in the reconstructed points is characterized by comprising the following steps: c2-1: the distances of the reconstruction point sets are assumed to form Gaussian distribution, and the average value mu and the standard deviation sigma of the distances from all the reconstruction points to other points are calculated, wherein C2-2: let std be the standard deviation multiple, when the average distance d from one point to the other is within the rangeAnd when the point is in the range, the point is reserved, and when the point is not in the range, the outlier is defined for deletion.
Further, the C3 merging the repeated points in the reconstructed point set is characterized by comprising the following steps: c1: calculate the average μ of the distances of all reconstructed points to other points, C2: let the proximity threshold be n, when the distance d between the nearest points of a point is smaller thanThe repeat point is determined and the average of the two points is used instead.
Compared with the prior art, the invention provides a rigid body stable reconstruction method of an optical motion capture system, which has the following beneficial effects:
1. according to the rigid body stable reconstruction method of the optical motion capture system, by using the method of eliminating overlapped polar lines by utilizing the included angles and the distances of the polar lines, abnormal situations of polar line overlapping are identified and eliminated, the problem of error points in rigid body reconstruction is reduced, the multi-view characteristic point matching method is used for realizing that a high-priority camera is used as a reference camera to perform multi-view characteristic point matching first, and the reference points which are matched preferentially can be aligned with all cameras of the same matching group, so that the problem of missing cameras of the matching group and repeated points in reconstruction is solved.
2. According to the rigid body stable reconstruction method of the optical motion capture system, outliers in the reconstruction points are identified and removed in the three-dimensional point reconstruction method, and the reconstructed error points can be identified and removed; the repeated points are combined, so that the repeated points can be effectively eliminated, the problem of the tremble points can be effectively identified by a position correction method of the tremble points of the rigid body marker points, the tremble points are corrected, and the problems of missing points, wrong points, repeated points and tremble points in the reconstruction of the optical motion capturing system are solved.
Drawings
FIG. 1 is a epipolar geometry basic model of a rigid body stable reconstruction method for an optical motion capture system according to the present invention;
FIG. 2 is a graph showing the correspondence between epipolar lines and matching points in a rigid body stable reconstruction method for an optical motion capture system according to the present invention;
FIG. 3 is a diagram illustrating a bipolar line constraint of a rigid body stable reconstruction method for an optical motion capture system according to the present invention;
FIG. 4 is a schematic diagram of matching feature points of a three-phase machine of a rigid body stable reconstruction method of an optical motion capture system according to the present invention;
FIG. 5 is a distance matrix diagram of epipolar lines and candidate matching points of a rigid body stable reconstruction method of an optical motion capture system according to the present invention;
FIG. 6 is a distance matrix diagram of a rigid body stable reconstruction method for an optical motion capture system according to the present invention for removing overlapping polar lines;
FIG. 7 is a matrix diagram of matching between the reference points and the candidate points of a rigid body stable reconstruction method of an optical motion capture system according to the present invention;
FIG. 8 is a distance matrix diagram of candidate points of a rigid body stable reconstruction method of an optical motion capture system according to the present invention;
FIG. 9 is a diagram showing the matching priority ranking of the camera and the points of the rigid body stable reconstruction method of the optical motion capturing system according to the present invention;
FIG. 10 is a schematic diagram of a matching set of two-dimensional view points of a rigid body stable reconstruction method of an optical motion capture system according to the present invention;
fig. 11 is a schematic diagram of a three-dimensional reconstruction point result of a rigid body stable reconstruction method for an optical motion capturing system according to the present invention.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and fully with reference to the accompanying drawings, in which it is evident that the embodiments described are only some, but not all embodiments of the invention. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
Example 1
A rigid body stable reconstruction method of an optical motion capture system comprises the following steps:
1) Epipolar geometry is a special geometrical relationship based on a pinhole model and existing only between two views, which is independent of scene structure and depends only on camera internal and external parameters. Epipolar geometry is commonly used to search for matching points when multi-view feature points match, which reduces the search area from an entire image to a straight line;
2) A fundamental model of epipolar geometry. Wherein O is 1 And O 2 Representing the optical centers of the two cameras respectively, the connecting line of the two cameras representing the base line, the pole being the intersection point of the base line and the image plane, denoted as e1 and e2 respectively, the pole plane being a three-dimensional point P and the optical center O 1 And O 2 Defined plane, denoted pi, l 1 And l 2 Is a polar line, which is a straight line intersecting the polar plane and the image plane, a polar line l 1 Is the optical center O of the camera 2 2 Projection of the line connecting P and P on the imaging plane of another camera, polar l 1 Then in contrast, the point in the camera 1 can find a corresponding pole line on the imaging plane of the camera 2, and the camera 2 imagesThe point on the surface matched with the point is necessarily positioned on the polar line, so that the search area of the matched point is reduced to a straight line when the multi-view feature points are matched;
3) Under ideal conditions, the corresponding relation between epipolar lines and projection points is constrained by epipolar lines, so that the projection point P of a space point on the imaging plane of the camera 1 is known 1 1 、P 2 1 、P 3 1 I.e. it can be determined that it is in the polar line l of the camera 2 1 、l 2 、1 3 And P 1 2 、P 2 2 、P 3 2 Is as good as 1 1 、1 2 、1 3 On-line, there are often mismatching situations, one of which is that due to errors in the internal and external parameters of the camera, the observation point is often near the epipolar line, such asAnd->Is far away from the polar line, so that the polar line is easy to be missed, thereby causing the missed-matching phenomenon, which is the case of two polar lines and observation points in many-to-one way, such as P 2 2 At the same time at l 2 And 1 3 Two lines, three, one-to-many lines and observation points, e.g. P 1 2 、P 2 2 At the same time at l 1 Fourthly, when the plane formed by the rigid points is parallel to the polar plane, the situation that a plurality of polar lines and a plurality of observation points are overlapped can occur;
4) To achieve multi-view feature point matching when the viewpoint is near the epipolar line, a bipolar line constraint is used to match the viewpoint to the epipolar line for two-dimensional viewpoint P matching in camera 1 and camera 2 1 1 And P 1 2 Ideally, P 1 2 Should be located at the polar l of the camera 2 1 On the contrary, under the influence of errors, the point in the actual situationTypically just near the pole line and possibly even far from the pole line. Double-pieceEpipolar constraints can be used to determine a search area of candidate matching points, defined by two parallel lines at a distance θ from the epipolar line, only two-dimensional observation points located within the area can be selected as candidate matching points. θ is the threshold of the two-dimensional viewpoint-to-epipolar distance, i.e., the threshold of bipolar-line constraint, as epipolar P in FIG. 4 1 1 Corresponding polar line l in camera 2 1 ，P 1 2 And P 2 2 Are all located in the search area, thus P 1 2 And P 2 2 Are all P 1 1 Is a candidate matching point of (a);
5) Screening candidate matching points by utilizing bi-polar line constraint, recording the candidate matching relation between all points in a matching matrix, wherein a matching matrix creation algorithm selects a certain point in a single camera as a datum point to start a candidate matching point screening process, firstly calculating polar equations of the point in other cameras, then sequentially calculating distances from two-dimensional observation points in each camera to corresponding polar lines, after traversing all the two-dimensional observation points of the camera, recording the distances from all the points of other cameras to the datum point polar lines through a blocking matrix, sequentially carrying out the above operation on all the points of all the cameras to obtain a distance matrix (shown in a figure 6) formed by the blocking matrix, wherein the distance from all the points to the polar lines in all the cameras is stored in the matrix;
6) In order to avoid overlapping of a plurality of polar lines and mismatching of multi-view feature point matching in fig. 3, the invention discloses a polar line eliminating method using included angles and distances between polar lines, comprising:
a1: computing polar equation 1 for all points of one camera in other cameras 1 、1 2 、...、1 i ；
A2: the overlapping of the epipolar lines is compared pairwise. And calculating an included angle between polar lines, and if the included angle is smaller than 1 degree, judging that polar lines are approximately parallel, and continuously judging the distance between the polar lines. And calculating the minimum distance of the straight line on the image, and judging that the polar line equations overlap if the minimum distance is smaller than 7. Wherein the threshold value of the minimum distance is the average diameter of the marker point;
a3: for overlapping epipolar lines, the distance value of two epipolar lines is changed to 99 in the distance matrix.
The distance threshold 10 is set, the matching value is marked as 1 when the threshold screening is passed, the matching value is marked as 0 when the threshold is not passed, a matching matrix with only 0 and 1 is obtained, and candidate matching relations among all two-dimensional observation points of all cameras are recorded, so that the method can be used for subsequent multi-camera matching.
The method for eliminating overlapped polar lines by utilizing the included angles and the distances of the polar lines identifies and eliminates the abnormal situation of polar line overlapping, reduces the problem of error points in rigid body reconstruction, and realizes that the multi-view characteristic point matching is carried out by taking a high-priority camera as a reference camera. The datum points which are preferentially matched can be aligned with all cameras of the same matching group, so that the problems of missing points and repeated points in reconstruction caused by missing cameras of the matching group are reduced, and in the three-dimensional point reconstruction method, outliers in the reconstruction point set are identified and removed, and the reconstructed error points can be identified and removed; the repeated points are combined, so that the repeated points can be effectively eliminated, the problem of the tremble points can be effectively identified by a position correction method for the tremble points of the rigid body marker points, and the tremble points are corrected.
Example two
A rigid body stable reconstruction method of an optical motion capture system comprises the following steps:
1) Three visual angle characteristic point matching realized by using three-phase machine space matching, the situation that one epipolar line corresponds to a plurality of observation points occurs to the epipolar geometry, the problem of mismatching and thus mismatching is easily caused, and P is the problem that 1 2 And P 2 2 Are all at 1 1 So P is within the neighborhood of 1 2 And P 2 2 Are all P 1 1 Is a candidate matching point of (a);
2) As shown in fig. 4 and 5, in the three-phase space matching process, the camera 1 that starts searching is called a reference camera; point P for generating epipolar line in reference camera 1 1 、P 2 1 、P 3 1 Referred to as a fiducial; both camera 2 and camera 3 become candidate cameras, points in the candidate cameras being referred to as candidate points; specifically, the camera 2 is calledFor the camera to be matched, P 1 2 、P 2 2 、P 3 2 Called points to be matched; the camera 3 is called a cooperative camera, P 2 3 Called collaboration points;
3) To solve this problem, a camera 3, P is introduced 1 1 Finding the candidate point P in the camera 3 2 3 Calculation of P by epipolar constraint 2 3 Polar line L of camera 2 i As shown in fig. 5. In the camera 2 image, l 1 And L is equal to i Generating an intersection point P cross 。P 1 1 Candidate matching point P of (2) 1 2 Within the search radius around the intersection point, the observation points in the three cameras are successfully matched, P 1 2 、P 2 2 、P 3 2 The three observation points form a candidate matching group;
4) When the number of cameras is greater than three, in order to find P 1 1 At the matching points of all cameras, P needs to be used on the basis of the candidate matching groups 1 1 And P 2 3 And continuing to perform three-camera space matching in the rest cameras until all cameras complete space matching, finally recording the point in a matching group at the matching points of all cameras, and completing matching of one point, wherein all points in the matching group are not participated in the subsequent matching process.
Wherein the above procedure is repeated for two-dimensional image points in the camera 1. When all the image points in the reference camera are processed in the same manner, another camera is selected as the reference camera, and the same process is repeated as with the remaining unmatched two-dimensional observation points.
The method for eliminating overlapped polar lines by utilizing the included angles and the distances of the polar lines identifies and eliminates the abnormal situation of polar line overlapping, reduces the problem of error points in rigid body reconstruction, and realizes that the multi-view characteristic point matching is carried out by taking a high-priority camera as a reference camera. The datum points which are preferentially matched can be aligned with all cameras of the same matching group, so that the problems of missing points and repeated points in reconstruction caused by missing cameras of the matching group are reduced, and in the three-dimensional point reconstruction method, outliers in the reconstruction point set are identified and removed, and the reconstructed error points can be identified and removed; the repeated points are combined, so that the repeated points can be effectively eliminated, the problem of the tremble points can be effectively identified by a position correction method for the tremble points of the rigid body marker points, and the tremble points are corrected.
Example III
A rigid body stable reconstruction method of an optical motion capture system comprises the following steps:
1) The feature point matching includes:
b1: sequencing the priority of the reference cameras and the reference points;
b2: sequencing the priority of the candidate cameras;
b3: iterative three-camera spatial registration;
the priority ordering of the reference cameras and the reference points in the step B1 is characterized by comprising the following steps:
b1-1: sequencing reference cameras according to the effective reference points, wherein the reference points are at least provided with candidate points in two cameras, marking the candidate points as effective reference points, calculating the effective reference points of each camera, and the higher the number of the points is, the higher the priority of the cameras is; deleting the effective reference points from the sequence if the number of the effective reference points is 0, and sorting the number of the candidate cameras if the number of the effective reference points of the plurality of cameras is the same;
counting the effective reference points of cam0, cam1, cam2, cam3 and cam4 from top to bottom to be 4, 0 and 4 respectively, so that the cameras are sequenced cam0, cam1, cam2, cam4 and cam3; because the effective reference point number of cam3 is 0, cam3 is deleted; the final camera sequences are cam0, cam1, cam2 and cam4, and the effective reference points of cam0, cam1, cam2 and cam4 are all 4, so that further sequencing is needed;
b1-2: sorting the reference cameras according to the sum of the numbers of the candidate cameras, wherein the reference points have candidate points on other cameras, the other cameras are the candidate cameras of the reference points, the higher the sum of the numbers of the candidate cameras of all the reference points in one reference camera is, the higher the priority of the reference cameras is, and if the effective reference points are the same in the sorting of B1-1, the sorting is performed according to the number of the candidate cameras;
counting from top to bottom, cam0 has 5 points, the candidate camera numbers of p0, p1, p2 and p3 are respectively 3, 4 and 4, and the sum of the candidate camera numbers of the cameras cam0 is 14. Similarly, the sum of candidate cameras of cam1, cam2, cam3, cam4 is 9, 0, 11, respectively. In S1, the effective reference points of cam0, cam1, cam2 and cam4 are the same, the sum of the candidate camera numbers is used for descending order, and finally the effective reference points are sequenced into cam0, cam4, cam1 and cam2;
b1-3: sorting the reference points according to the number of the candidate cameras, sorting the reference points of the same camera according to the descending order of the number of the candidate cameras in order to ensure that the reference points matched firstly find the whole candidate points as far as possible, and deleting the reference points from the sequence if the number of the candidate cameras is 0;
cam0 has 5 points, and the candidate camera numbers of p0, p1, p2, p3 and p4 are 3, 4 and 4 respectively, so the candidate points are ordered as p2, p3, p0 and p1. Similarly, the reference points of cam4, cam1, cam2 are ordered. The reference camera and reference point serial numbers finally output are shown in fig. 10, and the priorities of the reference camera and the reference point are arranged from top to bottom;
the matching matrix records candidate matching relations among all two-dimensional observation points of all cameras, the matching matrix value is 1, the point is the candidate point, the distance matrix records the distance from the candidate point to the datum point polar line, and the smaller the distance is, the larger the probability that the datum point and the candidate point are the same group of points is. Taking the distance of the candidate point with the matching matrix value of 1 to form the figure 9;
b2-1: sorting the candidate cameras according to the sum of the distances from the candidate points to the polar lines, calculating the sum of the distances from the candidate points of the candidate cameras of each datum point, and sorting in ascending order to obtain the serial numbers of the candidate cameras;
b2-2: the camera with the highest priority in the candidate cameras is used as a cooperative camera, and other cameras are used as cameras to be matched;
the cam0 point p0 has three candidate cameras, the candidate point of the candidate camera cam2 is p0, and the distance is 0; candidate points of the candidate camera cam3 are p1, the distance is 2, candidate points of the candidate camera cam4 are p0 and p2, the distances are 7 and 2 respectively, and the sum of the distances is 9, so that the candidate cameras of the cam0 point p0 are cam2, cam3 and cam4 in sequence, wherein cam2 is used as a cooperative camera, and cam3 and cam4 are used as cameras to be matched. Similarly, the candidate cameras of other reference points can be ranked, the complete ranking result is shown in fig. 10, and the sequence of the candidate cameras is arranged from top to bottom according to the priority;
and B3, iterative three-camera spatial registration, which is characterized by comprising the following steps:
b3-1: the reference points of the reference cameras are sequentially selected according to the priority order. And selecting the cooperative camera and the cooperative point corresponding to the reference point, and finally selecting the camera to be matched and the point to be matched according to the priority. As shown in fig. 10, a cam0 point p2 is selected, then a assistant camera cam1 and a cooperation point cam1 point p1 are selected, and finally a camera cam2 to be matched and a point cam2 to be matched point p3 are selected;
b3-2: three-camera space matching, firstly calculating the intersection point P of a datum point and a cooperation point in a camera polar line to be matched cross The method comprises the steps of carrying out a first treatment on the surface of the Then calculate the candidate point and P cross Setting a threshold value threshold1 finally, if d is smaller than or equal to threshold1, then matching successfully, wherein the datum point, the cooperation point and the point to be matched form a candidate matching group, and the cameras and the points in the candidate matching group do not participate in the subsequent matching process;
if a plurality of points to be matched exist in the camera to be matched, selecting the point to be matched closest to the point to be matched to add into a candidate matching group, wherein two candidate points p0 and p3 exist in the camera to be matched cam4 of the cam0 point p2, and in the matching process, comparing the distances between the p0 and p3, and adding the p3 with smaller distance into the matching group;
b3-3: repeating the processes from B3-1 to B3-2, traversing the cameras to be matched according to the priority of the candidate cameras, sequentially updating the cameras to be matched and the points to be matched according to the priority sequence, repeating the three-phase space matching process of S2, and adding the successfully matched cameras to be matched and the points to be matched into the candidate matching group until all the cameras to be matched are matched, so as to obtain a final matching group;
b3-4: repeating the processes from B3-1 to B3-3, traversing each datum point in the cameras according to the datum point priority to match, and obtaining a matching group of each datum point of the first priority datum camera;
b3-5: repeating the processes from B3-1 to B3-4, traversing each reference camera according to the priority of the reference camera to obtain a matching group of each reference camera;
b3-6: and comparing the matched groups obtained in the step B3-5 in pairs, merging the two matched groups if the matched groups have intersection, and judging as invalid matched groups if the matched groups have fewer points than 3, and deleting.
The method for eliminating overlapped polar lines by utilizing the included angles and the distances of the polar lines identifies and eliminates the abnormal situation of polar line overlapping, reduces the problem of error points in rigid body reconstruction, and realizes that the multi-view characteristic point matching is carried out by taking a high-priority camera as a reference camera. The datum points which are preferentially matched can be aligned with all cameras of the same matching group, so that the problems of missing points and repeated points in reconstruction caused by missing cameras of the matching group are reduced, and in the three-dimensional point reconstruction method, outliers in the reconstruction point set are identified and removed, and the reconstructed error points can be identified and removed; the repeated points are combined, so that the repeated points can be effectively eliminated, the problem of the tremble points can be effectively identified by a position correction method for the tremble points of the rigid body marker points, and the tremble points are corrected.
Example IV
A rigid body stable reconstruction method of an optical motion capture system comprises the following steps:
1) The three-dimensional point reconstruction includes:
c1: performing three-dimensional reconstruction by using the matching group;
c2: removing outliers in the reconstructed point set;
and C3: merging the repeated points in the reconstructed point set;
and C1, performing three-dimensional reconstruction by using the matching group, wherein the method comprises the following steps of:
c1-1: and carrying out pairwise reconstruction on the points in the matching group by using triangulation to obtain candidate three-dimensional points, re-projecting the three-dimensional points onto other cameras of the matching group, and calculating re-projection errors on the other cameras. If the error is larger than the diameter of the marker point, judging to reconstruct the error point, and eliminating. Finally, taking the average points of the pairwise reconstructed three-dimensional points as final three-dimensional reconstruction points;
three points cam0 point p3, cam2 point p2 and cam3 point p2 are arranged in the matching group; by means ofThree-dimensional point T1 is obtained by reconstructing cam0 point p3 and cam2 point p2 through triangulation, two-dimensional point T1 is obtained by re-projecting T1 onto cam3, and the distance between T1 and cam3 point p2 is calculated and recorded as re-projection error R error If R is error And (7) removing the mark points with the diameter being 7 larger than the diameter of the mark points. Two-by-two reconstruction to obtain three-dimensional points T1, T2 and T3, and calculating an average point W of the three-dimensional points as a final reconstruction point;
c1-2: reconstructing a three-dimensional point by each matching group, wherein all the three-dimensional points form a reconstruction point set;
removing outliers in the reconstructed points in C2, wherein the method comprises the following steps of:
c2-1: the distance assumption of the reconstruction point set is formed into Gaussian distribution, and the average value mu and the standard deviation sigma of the distances from all reconstruction points to other points are calculated;
c2-2: let std be the standard deviation multiple, when the average distance d from one point to the other is within the rangeWhen the point is in the range, defining an outlier for deleting if the point is not in the range;
and C3, merging repeated points in the reconstructed point set, and is characterized by comprising the following steps of:
c1: calculating the average value mu of the distances from all the reconstruction points to other points;
c2: let the proximity threshold be n, when the distance d between the nearest points of a point is smaller thanThe repeat point is determined and the average of the two points is used instead.
Finally, 4 three-dimensional points, W0 (357, 527, 1094), W1 (295, 491, 1129), W2 (372, 495, 1192), W3 (322, 418, 1144), are reconstructed, as shown in FIG. 10.
The method for eliminating overlapped polar lines by utilizing the included angles and the distances of the polar lines identifies and eliminates the abnormal situation of polar line overlapping, reduces the problem of error points in rigid body reconstruction, and realizes that the multi-view characteristic point matching is carried out by taking a high-priority camera as a reference camera. The datum points which are preferentially matched can be aligned with all cameras of the same matching group, so that the problems of missing points and repeated points in reconstruction caused by missing cameras of the matching group are reduced, and in the three-dimensional point reconstruction method, outliers in the reconstruction point set are identified and removed, and the reconstructed error points can be identified and removed; the repeated points are combined, so that the repeated points can be effectively eliminated, the problem of the tremble points can be effectively identified by a position correction method for the tremble points of the rigid body marker points, and the tremble points are corrected.
Example five
A rigid body stable reconstruction method of an optical motion capture system comprises the following steps:
1) In consecutive frames, the same marker jumps back and forth at two positions. A rigid body is composed of a plurality of marker points, the phenomenon of shaking points has randomness, and most of the marker points in the rigid body shake, so that the shaking of all the marker points is rarely generated. If the rigid body moves or rotates, all markers are displaced. Through the rule, whether all points in the rigid body move or not can be judged to judge the problem that the marker points in the rigid body jump points, and the positions of the shaking points are corrected by using the positions of the marker points of the previous frame:
2) Correcting the position of the shake point of the rigid body marker point comprises;
d1: recording the position information CW of the rigid body of the previous frame, wherein the position information CW comprises the positions W0, W1 and Wn of n marker points, and the position information CW of the rigid body of the current frame comprises the positions W0, W1 and Wn of n marker points;
d2: traversing each point in the CW, searching the nearest point in the CW, calculating the nearest distance d, taking the radius of the marker point as a threshold value threthold2, and judging that the marker point moves if d > threthold 2;
d3: if the marker point of the current frame part moves, judging that the phenomenon of the shaking point exists, and replacing the position of the current frame of the marker by the position of the nearest point of the previous frame.
Assuming that a rigid body consists of four marker points, the radius of the marker points on an image is 3, the coordinates of the previous frame are W0 (0, 0), W1 (100, 0), W2 (200,0,0) and W3 (300,0,0), and the coordinates of the current frame are W0 (0,10,0), W1 (100,1,0), W2 (200,2,0) and W3 (300,1,0); the nearest points of W0, W1, W2 and W3 are respectively W0, W1, W2 and W3, and the distances are respectively 10, 1, 2 and 1; since only W0 has a movement distance greater than 3, it is determined that the state is moved, and W1, W2, W3 are determined that the state is not moved. Therefore, W0 in the rigid body is judged to be a shake point, and the coordinate of the shake point is replaced by the coordinate of W0 to realize the coordinate correction of the shake point.
The method for eliminating overlapped polar lines by utilizing the included angles and the distances of the polar lines identifies and eliminates the abnormal situation of polar line overlapping, reduces the problem of error points in rigid body reconstruction, and realizes that the multi-view characteristic point matching is carried out by taking a high-priority camera as a reference camera. The datum points which are preferentially matched can be aligned with all cameras of the same matching group, so that the problems of missing points and repeated points in reconstruction caused by missing cameras of the matching group are reduced, and in the three-dimensional point reconstruction method, outliers in the reconstruction point set are identified and removed, and the reconstructed error points can be identified and removed; the repeated points are combined, so that the repeated points can be effectively eliminated, the problem of the tremble points can be effectively identified by a position correction method for the tremble points of the rigid body marker points, and the tremble points are corrected.
The beneficial effects of the invention are as follows:
according to the rigid body stable reconstruction method of the optical motion capture system, by using the method of eliminating overlapped polar lines by utilizing the included angles and the distances of the polar lines, abnormal situations of polar line overlapping are identified and eliminated, the problem of error points in rigid body reconstruction is reduced, the multi-view characteristic point matching method is used for realizing that a high-priority camera is used as a reference camera to perform multi-view characteristic point matching first, and the reference points which are matched preferentially can be aligned with all cameras of the same matching group, so that the problem of missing cameras of the matching group and repeated points in reconstruction is solved.
According to the rigid body stable reconstruction method of the optical motion capture system, outliers in the reconstruction points are identified and removed in the three-dimensional point reconstruction method, and the reconstructed error points can be identified and removed; the repeated points are combined, so that the repeated points can be effectively eliminated, the problem of the tremble points can be effectively identified by a position correction method of the tremble points of the rigid body marker points, the tremble points are corrected, and the problems of missing points, wrong points, repeated points and tremble points in the reconstruction of the optical motion capturing system are solved.
Although embodiments of the present invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made therein without departing from the principles and spirit of the invention, the scope of which is defined in the appended claims and their equivalents.
Claims (6)
1. The rigid body stable reconstruction method of the optical motion capture system is characterized by comprising the following steps of:
1) Removing overlapping epipolar lines includes:
a1: computing polar equation 1 for all points of one camera in other cameras 1 、1 2 、...、1 i ；
A2: the overlapping of polar lines is compared in pairs, the included angle between polar line equations is calculated first, if the included angle is smaller than T angle Judging that the polar lines are approximately parallel, continuously judging the distance between the polar lines, calculating the minimum distance of the straight line on the image, and if the minimum distance is smaller than the threshold value T distanse Then the polar line equation is judged to overlap, wherein the threshold T of the minimum distance distanse The average diameter of the marker point on the image;
2) The feature point matching includes:
b1: sequencing the priority of the reference cameras and the reference points;
b2: sequencing the priority of the candidate cameras;
b3: iterative three-camera spatial registration;
the iterative three-camera spatial registration in B3 is characterized by comprising the following steps:
b3-1: firstly, sequentially selecting datum points of a datum camera according to a priority order, then selecting a cooperative camera and a cooperative point corresponding to the datum points, and finally selecting a camera to be matched and a point to be matched according to the priority order;
b3-2: three-camera space matching, firstly calculating the intersection point P of a datum point and a cooperation point in a camera polar line to be matched cross The method comprises the steps of carrying out a first treatment on the surface of the Then calculate the candidate point and P cross Finally, a threshold value threshold1 is set,if d is smaller than or equal to threshold1, the matching is successful, the datum point, the cooperation point and the point to be matched form a candidate matching group, and the cameras and the points in the candidate matching group do not participate in the subsequent matching process;
b3-3: repeating the processes from B3-1 to B3-2, traversing the cameras to be matched according to the priority of the candidate cameras, sequentially updating the cameras to be matched and the points to be matched according to the priority sequence, repeating the three-phase space matching process of S2, and adding the successfully matched cameras to be matched and the points to be matched into the candidate matching group until all the cameras to be matched are matched, so as to obtain a final matching group;
b3-4: repeating the processes from B3-1 to B3-3, traversing each datum point in the cameras according to the datum point priority to match, and obtaining a matching group of each datum point of the first priority datum camera;
b3-5: repeating the processes from B3-1 to B3-4, traversing each reference camera according to the priority of the reference camera to obtain a matching group of each reference camera;
b3-6: comparing the matched groups obtained in the step B3-5 in pairs, merging the two matched groups if the matched groups have intersections, judging as invalid matched groups if the number of points of the matched groups is less than 3, and deleting;
3) The three-dimensional point reconstruction includes:
c1: performing three-dimensional reconstruction by using the matching group;
c2: removing outliers in the reconstructed point set;
and C3: merging the repeated points in the reconstructed point set;
4) The correction of the positions of the shake points of the rigid body marker points comprises the following steps:
d1: recording the position information CW of the rigid body of the previous frame, wherein the position information CW comprises the positions W0, W1 and Wn of n marker points, and the position information CW of the rigid body of the current frame comprises the positions W0, W1 and Wn of n marker points;
d2: traversing each point in the CW, searching the nearest point in the CW, calculating the nearest distance d, taking the radius of the marker point as a threshold value threthold2, and judging that the marker point moves if d > threthold 2;
d3: if the marker point of the current frame part moves, judging that the phenomenon of the shaking point exists, and replacing the position of the current frame of the marker by the position of the nearest point of the previous frame.
2. The method for reconstructing rigid body stability of optical motion capture system according to claim 1, wherein said B1 reference camera and reference point priority ranking comprises the steps of: b1-1: sequencing reference cameras according to the effective reference points, wherein the reference points are at least provided with candidate points in two cameras, marking the candidate points as effective reference points, calculating the effective reference points of each camera, and the higher the number of the points is, the higher the priority of the cameras is; if the number of the effective reference points is 0, deleting the effective reference points from the sequence, and if the number of the effective reference points of a plurality of cameras is the same, sorting the number of the candidate cameras, and B1-2: according to the sum of the number of the candidate cameras, sorting the reference cameras, wherein the reference cameras have candidate points on other cameras, the other cameras are the candidate cameras of the reference points, the higher the sum of the number of the candidate cameras of all the reference points in one reference camera is, the higher the priority of the reference cameras is, if the number of the effective reference points is the same in the sorting of B1-1, the sorting is carried out according to the number of the candidate cameras, and B1-3: sorting the reference points according to the number of the candidate cameras, sorting the reference points of the same camera according to the descending order of the number of the candidate cameras in order to ensure that the reference points matched firstly find the whole candidate points as far as possible, and deleting the reference points from the sequence if the number of the candidate cameras is 0.
3. The method for reconstructing rigid body stability of optical motion capture system according to claim 1, wherein the ranking of the priority of candidate cameras in B2 is characterized by comprising the following specific steps: b2-1: sorting the candidate cameras according to the sum of the distances from the candidate points to the polar lines, calculating the sum of the distances from the candidate points of the candidate cameras of each datum point, sorting in ascending order to obtain the serial numbers of the candidate cameras, and B2-2: the camera with the highest priority in the candidate cameras is used as a cooperative camera, and other cameras are used as cameras to be matched.
4. The method for reconstructing rigid body stability of optical motion capturing system according to claim 1, wherein the three-dimensional reconstruction is performed by using the matching group in the C1, comprising the steps of:
c1-1: reconstructing the points in the matching group in pairs by using triangulation to obtain candidate three-dimensional points, re-projecting the three-dimensional points onto other cameras of the matching group, and calculating re-projection errors on the other cameras; if the error is larger than the diameter of the marker point, judging that the error point is reconstructed, and eliminating; finally, taking the average points of the pairwise reconstructed three-dimensional points as final three-dimensional reconstruction points;
c1-2: each matching group reconstructs a three-dimensional point, and all the three-dimensional points form a reconstruction point set.
5. The method for reconstructing rigid body stability of optical motion capture system according to claim 1, wherein said C2 removing outliers in reconstructed points comprises the steps of: c2-1: the distances of the reconstruction point sets are assumed to form Gaussian distribution, and the average value mu and the standard deviation sigma of the distances from all the reconstruction points to other points are calculated, wherein C2-2: let std be the standard deviation multiple, when the average distance d from one point to the other is within the rangeAnd when the point is in the range, the point is reserved, and when the point is not in the range, the outlier is defined for deletion.
6. The method for reconstructing rigid body stability of optical motion capturing system according to claim 1, wherein the C3 combining the repeated points in the reconstruction point set is characterized by comprising the following steps: c1: calculate the average μ of the distances of all reconstructed points to other points, C2: let the proximity threshold be n, when the distance d between the nearest points of a point is smaller thanThe repeat point is determined and the average of the two points is used instead.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311129240.7A CN116863086B (en) | 2023-09-04 | 2023-09-04 | Rigid body stable reconstruction method for optical motion capture system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311129240.7A CN116863086B (en) | 2023-09-04 | 2023-09-04 | Rigid body stable reconstruction method for optical motion capture system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116863086A true CN116863086A (en) | 2023-10-10 |
| CN116863086B CN116863086B (en) | 2023-11-24 |
Family
ID=88223787
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311129240.7A Active CN116863086B (en) | 2023-09-04 | 2023-09-04 | Rigid body stable reconstruction method for optical motion capture system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116863086B (en) |
Citations (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6516099B1 (en) * | 1997-08-05 | 2003-02-04 | Canon Kabushiki Kaisha | Image processing apparatus |
| JP2007025863A (en) * | 2005-07-13 | 2007-02-01 | Advanced Telecommunication Research Institute International | Photographing system, photographing method, and image processing program |
| KR100755450B1 (en) * | 2006-07-04 | 2007-09-04 | 중앙대학교 산학협력단 | 3d reconstruction apparatus and method using the planar homography |
| US20120154393A1 (en) * | 2010-12-21 | 2012-06-21 | Electronics And Telecommunications Research Institute | Apparatus and method for creating animation by capturing movements of non-rigid objects |
| JP2017053739A (en) * | 2015-09-10 | 2017-03-16 | オプティトラック・ジャパン株式会社 | Motion capture method and motion capture system |
| US20170220037A1 (en) * | 2016-02-03 | 2017-08-03 | Sony Corporation | System and method for utilization of multiple-camera network to capture static and/or motion scenes |
| CN110634149A (en) * | 2018-06-22 | 2019-12-31 | 湖南大学 | Non-rigid target characteristic point matching method for optical motion capture system |
| CN112067233A (en) * | 2020-09-02 | 2020-12-11 | 中国航天空气动力技术研究院 | Six-degree-of-freedom motion capture method for wind tunnel model |
| CN112085848A (en) * | 2020-08-21 | 2020-12-15 | 深圳市瑞立视多媒体科技有限公司 | Method and equipment for optimally selecting rigid body mark points and optical motion capture system |
| WO2021088481A1 (en) * | 2019-11-08 | 2021-05-14 | 南京理工大学 | High-precision dynamic real-time 360-degree omnibearing point cloud acquisition method based on fringe projection |
| WO2022111525A1 (en) * | 2020-11-30 | 2022-06-02 | 魔珐(上海)信息科技有限公司 | Posture capturing method and apparatus, electronic device, and storage medium |
| CN116612243A (en) * | 2023-07-21 | 2023-08-18 | 武汉国遥新天地信息技术有限公司 | Method for inhibiting and processing abnormal points of three-dimensional track of optical motion capture system |
| CN116681827A (en) * | 2023-05-31 | 2023-09-01 | 北京航空航天大学 | Defect-free three-dimensional point cloud reconstruction method and device based on multi-monitoring camera and point cloud fusion |
-
2023
- 2023-09-04 CN CN202311129240.7A patent/CN116863086B/en active Active
Patent Citations (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6516099B1 (en) * | 1997-08-05 | 2003-02-04 | Canon Kabushiki Kaisha | Image processing apparatus |
| JP2007025863A (en) * | 2005-07-13 | 2007-02-01 | Advanced Telecommunication Research Institute International | Photographing system, photographing method, and image processing program |
| KR100755450B1 (en) * | 2006-07-04 | 2007-09-04 | 중앙대학교 산학협력단 | 3d reconstruction apparatus and method using the planar homography |
| US20120154393A1 (en) * | 2010-12-21 | 2012-06-21 | Electronics And Telecommunications Research Institute | Apparatus and method for creating animation by capturing movements of non-rigid objects |
| JP2017053739A (en) * | 2015-09-10 | 2017-03-16 | オプティトラック・ジャパン株式会社 | Motion capture method and motion capture system |
| US20170220037A1 (en) * | 2016-02-03 | 2017-08-03 | Sony Corporation | System and method for utilization of multiple-camera network to capture static and/or motion scenes |
| CN110634149A (en) * | 2018-06-22 | 2019-12-31 | 湖南大学 | Non-rigid target characteristic point matching method for optical motion capture system |
| WO2021088481A1 (en) * | 2019-11-08 | 2021-05-14 | 南京理工大学 | High-precision dynamic real-time 360-degree omnibearing point cloud acquisition method based on fringe projection |
| CN112085848A (en) * | 2020-08-21 | 2020-12-15 | 深圳市瑞立视多媒体科技有限公司 | Method and equipment for optimally selecting rigid body mark points and optical motion capture system |
| CN112067233A (en) * | 2020-09-02 | 2020-12-11 | 中国航天空气动力技术研究院 | Six-degree-of-freedom motion capture method for wind tunnel model |
| WO2022111525A1 (en) * | 2020-11-30 | 2022-06-02 | 魔珐(上海)信息科技有限公司 | Posture capturing method and apparatus, electronic device, and storage medium |
| CN116681827A (en) * | 2023-05-31 | 2023-09-01 | 北京航空航天大学 | Defect-free three-dimensional point cloud reconstruction method and device based on multi-monitoring camera and point cloud fusion |
| CN116612243A (en) * | 2023-07-21 | 2023-08-18 | 武汉国遥新天地信息技术有限公司 | Method for inhibiting and processing abnormal points of three-dimensional track of optical motion capture system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116863086B (en) | 2023-11-24 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109345620B (en) | Improved object point cloud splicing method for ICP (inductively coupled plasma) to-be-measured object by fusing fast point feature histogram | |
| CN109816703B (en) | Point cloud registration method based on camera calibration and ICP algorithm | |
| JP5739409B2 (en) | Method for determining the relative position of a first image device and a second image device and these devices | |
| US5475422A (en) | Method and apparatus for reconstructing three-dimensional objects | |
| CN106600627A (en) | Rigid body motion capturing method and system based on mark point | |
| CN110487216A (en) | A kind of fringe projection 3-D scanning method based on convolutional neural networks | |
| US20020164067A1 (en) | Nearest neighbor edge selection from feature tracking | |
| CN106485207B (en) | A kind of Fingertip Detection and system based on binocular vision image | |
| WO2011145285A1 (en) | Image processing device, image processing method and program | |
| CN110281231B (en) | Three-dimensional vision grabbing method for mobile robot for unmanned FDM additive manufacturing | |
| CN111080776B (en) | Human body action three-dimensional data acquisition and reproduction processing method and system | |
| JP6097903B2 (en) | Three-dimensional shape acquisition apparatus, processing method, and program | |
| CN1512455A (en) | Object three-dimensional model quick obtaining method based on active vision | |
| CN111189415A (en) | Multifunctional three-dimensional measurement reconstruction system and method based on line structured light | |
| Qingshan et al. | Point Cloud Registration Algorithm Based on Combination of NDT and PLICP | |
| CN116863086B (en) | Rigid body stable reconstruction method for optical motion capture system | |
| CN109215118B (en) | Incremental motion structure recovery optimization method based on image sequence | |
| CN112164119B (en) | Calibration method for multi-camera system placed in surrounding mode and suitable for narrow space | |
| Zheng et al. | Rail sensor: A mobile lidar system for 3d archiving the bas-reliefs in angkor wat | |
| CN111023994B (en) | Grating three-dimensional scanning method and system based on multiple measurement | |
| CN112258494A (en) | Focal position determination method and device and electronic equipment | |
| CN109859313B (en) | 3D point cloud data acquisition method and device, and 3D data generation method and system | |
| CN116862984A (en) | Space pose estimation method of camera | |
| CN114820798A (en) | Calibrator matching method and device | |
| JP2017094482A (en) | Robot control system and robot control method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |