CN107595388B - Near-infrared binocular vision stereo matching method based on reflective ball mark points - Google Patents
Near-infrared binocular vision stereo matching method based on reflective ball mark points Download PDFInfo
- Publication number
- CN107595388B CN107595388B CN201710646652.6A CN201710646652A CN107595388B CN 107595388 B CN107595388 B CN 107595388B CN 201710646652 A CN201710646652 A CN 201710646652A CN 107595388 B CN107595388 B CN 107595388B
- Authority
- CN
- China
- Prior art keywords
- points
- image
- point
- camera
- matching
- 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.)
- Active
Links
Images
Landscapes
- Length Measuring Devices By Optical Means (AREA)
Abstract
The invention discloses a near-infrared binocular vision stereo matching method based on reflective ball mark points, which utilizes the three-dimensional geometric dimension characteristics of the reflective ball mark points to remove pseudo matching point pairs of the near-infrared binocular vision mark points so as to obtain a fast and robust stereo matching result, thereby accurately calculating the spatial three-dimensional coordinates of the mark points and further realizing the real-time and robust tracking and positioning of the mark points. The invention can efficiently and accurately find the corresponding relation of the mark points on the image points of the left camera and the right camera, and effectively eliminate the false matching, thereby eliminating the interference of the false three-dimensional space point on the optical positioning and improving the positioning precision and the robustness of the near-infrared optical positioning system.
Description
Technical Field
The invention relates to the field of computer vision and computer-assisted surgery research, in particular to a near-infrared binocular vision stereo matching method based on reflective ball mark points.
Background
In recent decades, image-guided therapeutic systems based on optical positioning have been widely used in neurosurgery, oromaxillofacial surgery, orthopedic surgery, tumor radiotherapy, and even in soft tissue and organ surgery. The accurate positioning of the target point is the most critical core content of the image-guided therapy system applied to radiotherapy, surgical operation and other therapies, and the positioning accuracy of the target point directly influences the curative effect of the therapy.
For the image-guided therapy system based on optical positioning, the positioning accuracy of the target point mainly depends on the positioning accuracy of the near-infrared optical positioning system. Currently, mainstream image guided therapy systems, such as the stereovision navigation system of the medton force company and the vector vision Sky system of the Brain company, all use the Polaris optical positioning system of the NDI company of canada as an aligner for tracking the reflective ball markers attached to the surgical tool and the patient. The Polaris system can track both active and passive emissive markers. The active light-emitting type mark point controls the light-emitting sequence of the light-emitting mark points through a system control unit, so that only one mark point is tracked by the Polaris system at a certain moment, and the problem of stereo matching does not exist. However, when an active light-emitting type mark point (usually a near-infrared light-emitting diode) is used, a maximum of 32 mark points can be tracked within 1 second, and in the tracking process, if the mark point is in a moving state, a tracking error is easily generated. In addition, each active light-emitting type mark point is connected with a wire, so that the active light-emitting type mark point is very inconvenient to use. When the passive light-emitting type mark points (generally, the reflective ball mark points) are used, the Polaris system can track 50 mark points at most at the same time, and the passive light-emitting type mark points are very convenient to use because no conducting wire is connected with the mark points. However, the passive light-emitting type mark points are easy to generate a stereo matching problem in the using process, when more than two mark points and two camera sensors share a polar plane, more than 1 false three-dimensional space point appears, and the false three-dimensional space points are generated in a three-dimensional space to influence the final positioning precision. The reason for this is that the stereo matching algorithm of the Polaris system has defects, which causes the stereo matching problem.
Similarly, if an industrial camera is used for self-constructing a near-infrared binocular vision positioning system, the system still has the stereo matching problem generated when more than 2 mark points share a polar plane, so that a plurality of false three-dimensional space points are generated.
Therefore, the design of the robust and stable near-infrared binocular vision stereo matching method has great significance for improving the positioning stability and the positioning precision of the optical positioning system.
Disclosure of Invention
The invention aims to solve the problem that when more than 2 marker points are on a common polar plane in the process of tracking the marker points of the reflective ball by a near-infrared binocular positioning system, the image points of the marker points on a left camera and a right camera are not uniquely matched to generate a pseudo-matching point pair, so that more than 1 pseudo three-dimensional space point appears, and provides a near-infrared binocular vision stereo matching method based on the marker points of the reflective ball.
In order to achieve the purpose, the technical scheme provided by the invention is as follows: a near-infrared binocular vision stereo matching method based on reflective ball marking points is characterized in that pseudo-matching point pairs of the near-infrared binocular vision marking points are eliminated by utilizing the three-dimensional geometric dimension characteristics of the reflective ball marking points to obtain a fast and robust stereo matching result, so that the spatial three-dimensional coordinates of the marking points are accurately calculated, and real-time and robust tracking and positioning of the marking points are realized; the method for eliminating the pseudo-matching point pairs of the mark points by utilizing the three-dimensional geometric dimension characteristics of the mark points of the reflective balls comprises the following steps:
1) calculating the sub-pixel coordinates of the geometric center of the imaging spots of the marking points of the left camera and the right camera, and recording the coordinates as the coordinates of the image points of the marking points;
2) drawing a corresponding polar line of a left camera image point on a right camera image according to an epipolar geometry principle of binocular vision;
3) finding out all image points falling on polar lines on the image of the right camera;
4) if only one image point is located on the polar line in the image of the right camera, calculating the three-dimensional space coordinate of the point according to the triangulation principle; if there are multiple pixels on the epipolar line, the following rules are followed: firstly, calculating three-dimensional space coordinate points formed by a left camera image point corresponding to the polar line and a plurality of image points on the polar line according to a triangulation principle, then constructing a sphere which takes the three-dimensional space coordinate points as a sphere center and takes the radius obtained by the actual measurement of a light reflecting sphere as a radius, then extracting the outline of an imaging spot of the mark point on the left camera, then calculating a conical surface formed by the outline and the optical center of the left camera, and finally judging whether the conical surface is tangent with the constructed sphere, if so, the conical surface is a correct matching point, and if not, the conical surface is a false matching point.
In the step 1), imaging spots of all the marking points are obtained through image processing methods such as threshold segmentation, connected domain extraction and the like, then the geometric center sub-pixel coordinates of the imaging spots of the marking points are calculated by adopting a weighted average method, and the coordinates are used as the coordinates of the image points marked as the marking points.
In step 2), obtaining internal and external parameters of the left camera and the right camera and a transformation matrix between coordinate systems of the left camera and the right camera from a calibration result of the binocular camera system, and drawing polar lines corresponding to image points of the left camera on an image of the right camera according to an epipolar geometry principle of binocular vision.
In step 3), a distance threshold is set, the distance between the image points of all the mark points on the right camera image and the epipolar line is calculated, if the distance between the image points and the epipolar line is smaller than the threshold, the image points are considered to be located on the epipolar line, otherwise, the image points do not fall on the epipolar line, and therefore all the image points falling on the epipolar line on the right camera image are found.
In step 4), judging according to the number of the image points falling on the epipolar line on the image of the right camera: a. if only 1 image point is located on the image polar line of the right camera, the matching point pair is a correct matching point pair, and the three-dimensional space point coordinate corresponding to the matching point pair is directly calculated according to the triangulation principle; b. if the number of the image points on the image polar line of the right camera is more than 2, eliminating the pseudo-matching point pairs according to the three-dimensional geometric dimension characteristics of the reflective ball marking points, wherein the elimination calculation process of the pseudo-matching points is as follows:
4-1) recording the left camera image point corresponding to the polar line as pliThe optical centers of the left and right cameras are olAnd orLet the number of pixels on the epipolar line in the right camera image be N, and the corresponding point be prjJ 1,2, …, N, pliAnd prjPerforming one-to-one pairing to form N combinations, and calculating two light rays o according to the triangulation principlelpliAnd orprjPoint of intersection PijI.e. is pliAnd prjCorresponding three-dimensional space coordinate points;
4-2) recording the sphere radius of the reflective sphere mark point as R, and constructing a virtual sphere V with the radius of R by taking the three-dimensional space coordinate point obtained by calculation in 4-1) as the sphere centerij；
4-3) processing the imaging spot of the ith mark point obtained by the image processing method on the left camera image to obtain the contour line of the imaging spot, and then performing circle fitting on the contour line to obtain a circle CiCalculate its optical center o with the left cameralFormed conical surface Sli；
4-4) setting a threshold t, calculating the conical surface SliWhether or not to contact with the virtual sphere VijTangent, if tangent, it is the image point p on the left camera planeliWith the image point p in the plane of the right camerarjAnd if the matching point pair is not tangent, the matching point pair is a false matching point pair.
Compared with the prior art, the invention has the following advantages and beneficial effects:
the near-infrared binocular vision stereo matching method based on the reflective ball mark points can efficiently and accurately find the corresponding relation of the mark points on the image points of the left camera and the right camera, and effectively eliminate false matching, thereby eliminating the interference of false three-dimensional space points on optical positioning and improving the positioning precision and robustness of a near-infrared optical positioning system.
Drawings
FIG. 1 shows the extraction result of the geometric center of the imaging spot of the left and right camera image mark points in the method of the present invention.
FIG. 2 is a diagram showing the epipolar lines in the right camera image corresponding to an image point a in the left camera image according to the method of the present invention.
Fig. 3 is a schematic diagram of calculating the coordinate of the matching point pair of the image points of the left camera and the right camera in three-dimensional space.
FIG. 4 shows the calculation of the conical surface SlaSchematic diagram of tangent relation with virtual sphere.
Detailed Description
The present invention will be further described with reference to the following specific examples.
The near-infrared binocular vision stereo matching method based on the reflective ball mark points can further improve the positioning robustness and precision of the near-infrared optical positioning system. The method comprises the steps of recognizing pseudo-matching point pairs of mark points by utilizing the three-dimensional geometric dimension characteristics of the mark points of a reflective ball, firstly calculating the sub-pixel coordinates of the geometric centers of imaging spots of the mark points of a left camera and a right camera, recording the coordinates as the coordinates of the image points of the mark points, then drawing corresponding polar lines of the image points of the left camera on the image of the right camera according to the epipolar geometric principle of binocular vision, then calculating all the mark point image points on the image of the right camera, wherein if only one image point on one polar line in the image of the right camera is provided, the three-dimensional space coordinates of the point are calculated according to the triangulation principle; if a plurality of image points exist on the polar line, calculating a three-dimensional space coordinate point formed by the image point of the left camera corresponding to the polar line and the plurality of image points on the polar line according to a triangulation principle, then constructing a sphere taking the three-dimensional space coordinate points as a spherical center and the actual radius of the light reflecting sphere as a radius, then extracting the outline of the imaging spot of the mark point on the image of the left camera, then calculating a conical surface formed by the outline and the optical center of the left camera, and finally calculating whether the conical surface is tangent with the constructed sphere, if the conical surface is tangent, the conical surface is a correct matching point, and if the conical surface is not tangent, the conical surface is a false matching point. The method comprises the following specific steps:
firstly, acquiring left and right images simultaneously shot by a left camera and a right camera, and calculating the sub-pixel coordinates of the geometric center of an imaging spot of a marking point of the left camera and the right camera. Respectively obtaining imaging spots of all marking points on the left camera image and the right camera image by image processing methods such as threshold segmentation, connected domain acquisition and the like to obtain a result shown as (a) in figure 1, then calculating the geometric center sub-pixel coordinates of the imaging spots of the marking points by adopting a weighted average method, wherein the gray value of m pixel points in the area of the imaging spots of the marking points is greater than a threshold t, and the gray value of the ith pixel point is set as viPixel coordinate is qi＝[qix,qiy]TThen the sub-pixel coordinate of the center of the region is
In fig. 1, (b) is indicated by a center dot, and the sub-pixel coordinates are expressed as the pixel coordinates of the marker.
Obtaining internal and external parameters of a left camera and a right camera and a transformation matrix between coordinate systems of the left camera and the right camera from a calibration result of a binocular camera system, taking an image point a of any one marked point on an image as an example, drawing an epipolar line m corresponding to the image point a on the image of the left camera on the image of the right camera according to an epipolar geometry principle of binocular visionraAs shown in fig. 2.
Set distanceCalculating all mark point image points and polar lines m on the right camera image from the threshold value kraIf the image point is far from the polar line mraIs less than a threshold k, the image point is considered to be located at epipolar line mraOn the contrary, the polar line m is not fallenraIn the above, the m falling on the polar line is calculatedraThe number of the upper image point is N, then the image point is recorded in the polar line mraHas a sub-pixel coordinate of prj(j＝1,2,…,N)。
Then, the judgment is carried out according to the number of the image points falling on the polar line on the image of the right camera: 1) if N is equal to 1, the polar line mraA matching point pair consisting of the corresponding image point of the left camera and the image point on the polar line of the right camera is a correct matching point pair, and the coordinate of the three-dimensional space point corresponding to the matching point pair is directly calculated according to the triangulation principle; 2) if N is larger than or equal to 2, eliminating the pseudo-matching point pairs according to the three-dimensional geometric size characteristics of the reflective ball marking points. Memory line mraThe corresponding sub-pixel coordinate of the left camera image point a is plaThe optical centers of the left and right cameras are olAnd orA 1 is to plaAnd prjOne-to-one pairing is carried out to form N combinations, and the triangulation principle is used to calculate two rays olplaAnd orprjPoint of intersection paj(j ═ 1,2, …, N) from o in the system coordinate systemlPoint to plaVector of points
Comprises the following steps:
similarly, the slave o is under the system coordinate systemrPoint to prjVector of pointsComprises the following steps:
wherein t isla、trjIs an arbitrary value, Bla＝[qlax,qlay,1]T，Brj＝[qrjx,qrjy,1]T，RlgaAnd TlgaIs a rotational-translation matrix transformed from the left camera coordinate system to the system coordinates, and RrgjAnd TrgjIs a rotation and translation matrix transformed from the coordinate system of the right camera to the system coordinate, and an equation set is solved
Then the three-dimensional space point P can be obtainedaj(j ═ 1,2, …, N) coordinates under the system, as shown in fig. 3.
Actually measuring the sphere radius of the reflective sphere mark point, recording the sphere radius as R, and obtaining a three-dimensional space coordinate point P through the previous calculationaj(j-1, 2, …, N) as the center of sphere, a plurality of virtual spheres V with radius R are constructedaj(j＝1,2,…,N)；
Processing the imaging spots of the mark points a obtained by processing the left camera image by methods such as threshold value and the like, identifying the contour line of the marking points a by adopting a Candy edge identifier, and then performing circle fitting on the contour line to obtain a circle ClaCalculate its optical center o with the left cameralFormed conical surface SlaAs shown in fig. 4.
Setting a threshold t, calculating the conical surface SlaWhether or not to contact with the virtual sphere VajTangent, if tangent, it is the image point p on the left camera planelaWith the image point p in the plane of the right camerarjAnd if the matching point pair is not tangent, the matching point pair is a false matching point pair.
Similarly, matching points on the imaging surface of the right camera corresponding to the other marking point image points on the left camera image can be calculated.
In conclusion, the method can efficiently and accurately find the corresponding relation of the mark points on the image points of the left camera and the right camera, and effectively eliminate the false matching, thereby eliminating the interference of the false three-dimensional space point on the optical positioning, improving the positioning precision and the robustness of the near-infrared optical positioning system, and being worthy of popularization.
The above-mentioned embodiments are merely preferred embodiments of the present invention, and the scope of the present invention is not limited thereto, so that the changes in the shape and principle of the present invention should be covered within the protection scope of the present invention.
Claims (5)
1. A near-infrared binocular vision stereo matching method based on reflective ball marking points is characterized in that pseudo-matching point pairs of the near-infrared binocular vision marking points are eliminated by utilizing the three-dimensional geometric dimension characteristics of the reflective ball marking points to obtain a fast and robust stereo matching result, so that the spatial three-dimensional coordinates of the marking points are accurately calculated, and real-time and robust tracking and positioning of the marking points are realized; the method is characterized in that: the method for eliminating the pseudo-matching point pairs of the mark points by utilizing the three-dimensional geometric dimension characteristics of the mark points of the reflective ball comprises the following steps:
1) calculating the sub-pixel coordinates of the geometric center of the imaging spots of the marking points of the left camera and the right camera, and recording the coordinates as the coordinates of the image points of the marking points;
2) drawing a corresponding polar line of a left camera image point on a right camera image according to an epipolar geometry principle of binocular vision;
3) finding out all image points falling on polar lines on the image of the right camera;
4) if only one image point is located on the polar line in the image of the right camera, calculating the three-dimensional space coordinate of the point according to the triangulation principle; if there are multiple pixels on the epipolar line, the following rules are followed: firstly, calculating three-dimensional space coordinate points formed by a left camera image point corresponding to the polar line and a plurality of image points on the polar line according to a triangulation principle, then constructing a sphere which takes the three-dimensional space coordinate points as a sphere center and takes the radius obtained by the actual measurement of a light reflecting sphere as a radius, then extracting the outline of an imaging spot of the mark point on the left camera, then calculating a conical surface formed by the outline and the optical center of the left camera, and finally judging whether the conical surface is tangent with the constructed sphere, if so, the conical surface is a correct matching point, and if not, the conical surface is a false matching point.
2. The near-infrared binocular vision stereo matching method based on the reflective ball marker points as claimed in claim 1, wherein the method comprises the following steps: in the step 1), imaging spots of all the marking points are obtained through threshold segmentation and connected domain extraction, then the geometric center sub-pixel coordinates of the imaging spots of the marking points are calculated by adopting a weighted average method, and the coordinates are used as the coordinates of the image points marked as the marking points.
3. The near-infrared binocular vision stereo matching method based on the reflective ball marker points as claimed in claim 1, wherein the method comprises the following steps: in step 2), obtaining internal and external parameters of the left camera and the right camera and a transformation matrix between coordinate systems of the left camera and the right camera from a calibration result of the binocular camera system, and drawing polar lines corresponding to image points of the left camera on an image of the right camera according to an epipolar geometry principle of binocular vision.
4. The near-infrared binocular vision stereo matching method based on the reflective ball marker points as claimed in claim 1, wherein the method comprises the following steps: in step 3), a distance threshold is set, the distance between the image points of all the mark points on the right camera image and the epipolar line is calculated, if the distance between the image points and the epipolar line is smaller than the threshold, the image points are considered to be located on the epipolar line, otherwise, the image points do not fall on the epipolar line, and therefore all the image points falling on the epipolar line on the right camera image are found.
5. The near-infrared binocular vision stereo matching method based on the reflective ball marker points as claimed in claim 1, wherein the method comprises the following steps: in step 4), judging according to the number of the image points falling on the epipolar line on the image of the right camera: a. if only 1 image point is located on the image polar line of the right camera, the matching point pair is a correct matching point pair, and the three-dimensional space point coordinate corresponding to the matching point pair is directly calculated according to the triangulation principle; b. if the number of the image points on the image polar line of the right camera is more than 2, eliminating the pseudo-matching point pairs according to the three-dimensional geometric dimension characteristics of the reflective ball marking points, wherein the elimination calculation process of the pseudo-matching points is as follows:
4-1) recording the left side corresponding to the polar lineThe camera image point is pliThe optical centers of the left and right cameras are olAnd orLet the number of pixels on the epipolar line in the right camera image be N, and the corresponding point be prjJ 1,2, …, N, pliAnd prjPerforming one-to-one pairing to form N combinations, and calculating two light rays o according to the triangulation principlelpliAnd orprjPoint of intersection PijI.e. is pliAnd prjCorresponding three-dimensional space coordinate points;
4-2) recording the sphere radius of the reflective sphere mark point as R, and constructing a virtual sphere V with the radius of R by taking the three-dimensional space coordinate point obtained by calculation in 4-1) as the sphere centerij；
4-3) processing the imaging spot of the ith mark point obtained by the image processing method on the left camera image to obtain the contour line of the imaging spot, and then performing circle fitting on the contour line to obtain a circle CiCalculate its optical center o with the left cameralFormed conical surface Sli；
4-4) setting a threshold t, calculating the conical surface SliWhether or not to contact with the virtual sphere VijTangent, if tangent, it is the image point p on the left camera planeliWith the image point p in the plane of the right camerarjAnd if the matching point pair is not tangent, the matching point pair is a false matching point pair.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710646652.6A CN107595388B (en) | 2017-08-01 | 2017-08-01 | Near-infrared binocular vision stereo matching method based on reflective ball mark points |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710646652.6A CN107595388B (en) | 2017-08-01 | 2017-08-01 | Near-infrared binocular vision stereo matching method based on reflective ball mark points |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107595388A CN107595388A (en) | 2018-01-19 |
| CN107595388B true CN107595388B (en) | 2020-02-18 |
Family
ID=61064031
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710646652.6A Active CN107595388B (en) | 2017-08-01 | 2017-08-01 | Near-infrared binocular vision stereo matching method based on reflective ball mark points |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107595388B (en) |
Families Citing this family (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109410277B (en) * | 2018-11-15 | 2022-03-08 | 广州科莱瑞迪医疗器材股份有限公司 | Virtual mark point filtering method and system |
| CN109538208A (en) * | 2018-12-21 | 2019-03-29 | 冀中能源峰峰集团有限公司 | A kind of compound positioning system of cutting head of roadheader and method |
| CN110812710B (en) * | 2019-10-22 | 2021-08-13 | 苏州雷泰智能科技有限公司 | Accelerator frame rotation angle measuring system and method and radiotherapy equipment |
| CN111028284A (en) * | 2019-10-31 | 2020-04-17 | 浙江未来技术研究院(嘉兴) | Binocular vision stereo matching method and device based on homonymous mark points |
| CN110969659B (en) * | 2019-10-31 | 2024-03-15 | 浙江清华长三角研究院 | Spatial positioning device and method for passive mark point |
| CN113208731B (en) * | 2021-04-23 | 2023-02-10 | 上海大学 | Binocular vision system-based hand and eye calibration method for surgical puncture robot |
| CN113499137B (en) * | 2021-07-07 | 2022-07-12 | 南开大学 | Surgical robot navigation positioning system and measurement visual angle multi-target optimization method |
| CN116047412B (en) * | 2023-03-28 | 2023-06-02 | 深圳惟德精准医疗科技有限公司 | Artifact coordinate removal method of marker and related device |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103297798A (en) * | 2012-03-02 | 2013-09-11 | 包崇兵 | Three-dimensional reconstruction method of scene spots in binocular stereo visual system |
| CN103968829A (en) * | 2014-05-13 | 2014-08-06 | 清华大学 | Three-dimensional space orientation tracking method and three-dimensional space orientation tracking system based on virtual marker |
| CN104224320A (en) * | 2013-06-19 | 2014-12-24 | 上海优益基医疗器械有限公司 | Surgical navigation-based wireless surface registering tool and implementation method thereof |
| CN106580470A (en) * | 2016-10-18 | 2017-04-26 | 南京医科大学附属口腔医院 | System and method for head positioning on basis of binocular vision |
-
2017
- 2017-08-01 CN CN201710646652.6A patent/CN107595388B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103297798A (en) * | 2012-03-02 | 2013-09-11 | 包崇兵 | Three-dimensional reconstruction method of scene spots in binocular stereo visual system |
| CN104224320A (en) * | 2013-06-19 | 2014-12-24 | 上海优益基医疗器械有限公司 | Surgical navigation-based wireless surface registering tool and implementation method thereof |
| CN103968829A (en) * | 2014-05-13 | 2014-08-06 | 清华大学 | Three-dimensional space orientation tracking method and three-dimensional space orientation tracking system based on virtual marker |
| CN106580470A (en) * | 2016-10-18 | 2017-04-26 | 南京医科大学附属口腔医院 | System and method for head positioning on basis of binocular vision |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107595388A (en) | 2018-01-19 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107595388B (en) | Near-infrared binocular vision stereo matching method based on reflective ball mark points | |
| US7561733B2 (en) | Patient registration with video image assistance | |
| US9622824B2 (en) | Method for automatically identifying instruments during medical navigation | |
| US11904182B2 (en) | Research and development of augmented reality in radiotherapy | |
| CN111388092B (en) | Positioning tracking piece, registration method, storage medium and electronic equipment | |
| Deacon et al. | The Pathfinder image-guided surgical robot | |
| CN108489398B (en) | Method for measuring three-dimensional coordinates by laser and monocular vision under wide-angle scene | |
| Martin-Gomez et al. | STTAR: surgical tool tracking using off-the-shelf augmented reality head-mounted displays | |
| CN106955122A (en) | Head 3 D detection method based on single camera module in radiotherapy | |
| CN112168357B (en) | System and method for constructing spatial positioning model of C-arm machine | |
| CN102252661B (en) | Globoid three-dimensional positioning method for machine vision | |
| CN113793389B (en) | Virtual-real fusion calibration method and device for augmented reality system | |
| CN114821031A (en) | Intraoperative image matching method, device and system based on C-arm machine | |
| CN112215871A (en) | Moving target tracking method and device based on robot vision | |
| CN112184653A (en) | Binocular endoscope-based focus three-dimensional size measuring and displaying method | |
| Lee et al. | Fast‐MICP for frameless image‐guided surgery | |
| Li et al. | A vision-based navigation system with markerless image registration and position-sensing localization for oral and maxillofacial surgery | |
| Liu et al. | Cognitive tracking of surgical instruments based on stereo vision and depth sensing | |
| CN106504257B (en) | A kind of radiotherapy head position attitude measuring and calculation method | |
| CN109364385A (en) | Tumor-localizing method based on binocular vision | |
| CN113963057B (en) | Imaging geometric relation calibration method and device, electronic equipment and storage medium | |
| CN109410277B (en) | Virtual mark point filtering method and system | |
| Lin et al. | Optimization model for the distribution of fiducial markers in liver intervention | |
| CN115471559B (en) | Head dynamic positioning and tracking method and system | |
| CN116993794B (en) | Virtual-real registration method and device for augmented reality surgery assisted navigation |
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 |