Registration method, device and equipment for target of surgical navigation positioning system

Abstract

The invention relates to the technical field of medical instruments and computer intelligence, in particular to a target registration method, device and equipment for a surgical navigation positioning system. The method comprises the following steps: identifying a target to obtain point cloud data of the target; obtaining a first feature description of at least one virtual point in the point cloud data; acquiring a second feature description of at least one mark point of the target; the second feature description of the mark point comprises at least one second element, and the at least one second element is used for representing relative position information between the mark point and at least one other mark point around the mark point; and matching the at least one virtual point with the at least one corresponding mark point to obtain at least one group of virtual points and mark points which are matched with each other. The technical problem that in the surgical navigation positioning technology, an optical positioning tracking method in the prior art is insufficient in real-time tracking precision keeping and point cloud data analysis speed is mainly solved.

Description

Technical Field

[0001] The present application relates to the field of medical devices and computer intelligence technology, and in particular to a registration method, device and equipment for a target of a surgical navigation and positioning system. Background Art

[0002] In recent years, with the advancement of medical imaging technology and the widespread application of minimally invasive surgery, the accuracy and real-time performance of surgical navigation and positioning systems have become key factors affecting surgical outcomes. With the maturity of high-definition medical imaging technology and the emphasis on refined and minimally invasive surgery in the medical field, the use of surgical navigation and positioning systems to assist doctors in surgery has become an inevitable trend in the development of the medical field. The core part of the surgical navigation and positioning system is the optical positioning and tracking technology, which provides accurate and real-time target positioning information for the navigation process during surgery, thereby ensuring the safety and reliability of medical surgery.

[0003] Existing surgical navigation and positioning technologies, especially those based on optical positioning and tracking, have improved the accuracy of surgical operations to a certain extent, but they still have some significant limitations. In actual applications, the traditional "coarse first, fine later" registration strategy may cause the registration results to gradually deviate over time due to factors such as human tissue deformation and respiratory movement, and it is impossible to always maintain high-precision real-time tracking. In addition, when processing large amounts of point cloud data, the existing technology has low efficiency in neighborhood feature analysis and matching, especially for the problem of positioning many targets in complex surgical environments, and it is difficult to efficiently and accurately identify and distinguish markers on different targets. Summary of the invention

[0004] The present application mainly solves the technical problem that the optical positioning and tracking method of the prior art in surgical navigation and positioning technology is insufficient in maintaining real-time tracking accuracy and point cloud data analysis speed.

[0005] An embodiment of the present application provides a registration method for a target of a surgical navigation positioning system, the registration method comprising:

[0006] Identify the target to obtain point cloud data of the target; the target includes at least one marking point, the point cloud data includes at least one virtual point, and at least one virtual point corresponds to at least one marking point of the target;

[0007] Acquire a first feature description of at least one virtual point in the point cloud data; the first feature description of the virtual point includes at least one first element, and the at least one first element is used to characterize the relative position information between the virtual point and at least one other virtual point around it;

[0008] Acquire a second feature description of at least one marking point of the target; the second feature description of the marking point includes at least one second element, and the at least one second element is used to characterize the relative position information between the marking point and at least one other marking point around it;

[0009] Match at least one virtual point with at least one corresponding marked point to obtain at least one group of mutually matching virtual points and marked points; wherein, when a first condition is satisfied between a first feature description of a virtual point and a second feature description of a corresponding marked point, the virtual point and the marked point corresponding to the virtual point are mutually matched; and the first condition is set based on the similarity between the first feature description and the second feature description.

[0010] In the above-mentioned registration method for the target of the surgical navigation and positioning system, as a preferred solution, the step of identifying the target to obtain point cloud data of the target includes:

[0011] Identify the target through the identification device to obtain the point cloud data of the target;

[0012] in:

[0013] The recognition device includes a three-dimensional scanner, and / or a visual sensor, and / or an RGB-D camera, and / or a projector and a camera;

[0014] When the identification device is a three-dimensional scanner, the target is scanned by the three-dimensional scanner to obtain point cloud data about the marking points on the target;

[0015] When the recognition device is a binocular vision sensor, the target is recognized by the vision sensor and image processing is performed to obtain point cloud data about the marking points on the target;

[0016] When the recognition device is an RGB-D camera, the target is photographed by the RGB-D camera to obtain a depth image of the target, and point cloud data about the marking points on the target are obtained according to the depth image;

[0017] When the recognition device is a projector and a camera, the projector transmits a grating image to the target, and the camera captures a reflected image of the target to obtain point cloud data about the marking points on the target.

[0018] In the above-mentioned registration method for the target of the surgical navigation and positioning system, as a preferred solution, the step of obtaining a first feature description of at least one virtual point in the point cloud data includes:

[0019] For any virtual point P in the point cloud data i , then:

[0020] Get the virtual point Pi at least one other virtual point within a preset distance;

[0021] According to the virtual point P i The virtual point P is obtained by combining the relative position information between the virtual point P and at least one other virtual point. i The first feature description P id ;

[0022] Where i = 1, ..., n; n is the number of virtual points in the point cloud data; the first neighborhood point set P id At least one of the first elements corresponds to at least one other virtual point within a preset distance.

[0023] In the above-mentioned registration method for the target of the surgical navigation positioning system, as a preferred solution, at the virtual point P i The neighborhood point set P id At least one other virtual point is in accordance with the virtual point P i The distances of the virtual points are sorted from near to far, or at least one other virtual point is sorted according to the distance from the virtual point P i The distances are sorted from far to near.

[0024] In the above-mentioned registration method for surgical navigation and positioning system targets, as a preferred solution, the step of obtaining a second feature description of at least one marker point of the target includes:

[0025] Acquire the real position information of at least one marking point on the target; the real position information includes the three-dimensional coordinates of the marking point;

[0026] For any target point P Ij , then:

[0027] According to the real position information of at least one marking point, calculate and obtain the position information of the marking point P Ij at least one other marked point within a preset distance;

[0028] According to the marking point P Ij The set of relative position information between the marker point P and at least one other marker point is obtained Ij The second feature description P Ijd ;

[0029] Where I represents the number of the target, P Ij Refers to the jth marked point of the Ith target; the second neighborhood point set P Ijd At least one of the second elements corresponds to at least one other marking point within a preset distance.

[0030] In the above-mentioned registration method for the target of the surgical navigation positioning system, as a preferred solution, at the marker point P Ij The neighborhood point set P Ijd At least one other marked point is in accordance with the marked point P Ij The distances of the points are sorted from near to far, or at least one other point is sorted according to the distance from the point P. Ij The distances are sorted from far to near.

[0031] In the above-mentioned registration method for the target of the surgical navigation positioning system, as a preferred solution, the matching of at least one virtual point with at least one corresponding marking point includes:

[0032] Compare at least one of the first elements of the first feature description of any virtual point with at least one of the second elements of the second feature description of the corresponding marking point at least once;

[0033] At each comparison, determine whether the difference between the first element and the second element is less than the error threshold: if so, increase the similarity between the first feature description and the second feature description; if not, do not increase the similarity between the first feature description and the second feature description;

[0034] After the first element of the first feature description and the second element of the second feature description are compared, determine whether the similarity between the first feature description and the second feature description exceeds a preset threshold: if so, the first condition is satisfied between the first feature description and the second feature description, and the virtual points and marked points respectively corresponding to the first feature description and the second feature description are matched with each other; if not, the first condition is not satisfied between the first feature description and the second feature description, and the virtual points and marked points respectively corresponding to the first feature description and the second feature description are not matched.

[0035] In the above-mentioned registration method for surgical navigation and positioning system targets, as a preferred solution, the first feature description P id is {l1,l2,…,l k}, the second feature description P Ijd is {d1,d2,…,d k};

[0036] in:

[0037] l1,l2,…,l k are all first elements, and the relative position information of the first elements includes the relative distance between two virtual points;

[0038] d1,d2,…,d k are both second elements, and the relative position information of the second elements includes the relative distance between the two marking points.

[0039] The present application also provides a registration device for a target of a surgical navigation positioning system, the registration device comprising:

[0040] A point cloud data acquisition module, wherein the point cloud data acquisition module is used to identify the target to acquire point cloud data of the target; the target includes at least one marking point, and the point cloud data includes at least one virtual point, and at least one virtual point corresponds to at least one marking point of the target;

[0041] A first feature description acquisition module, the first feature description acquisition module is used to acquire a first feature description of at least one virtual point in the point cloud data; the first feature description of the virtual point includes at least one first element, and the at least one first element is used to characterize the relative position information between the virtual point and at least one other virtual point around it;

[0042] A second feature description acquisition module, the second feature description acquisition module is used to acquire a second feature description of at least one marking point of the target; the second feature description of the marking point includes at least one second element, and the at least one second element is used to characterize the relative position information between the marking point and at least one other marking point around it;

[0043] A matching module, wherein the matching module is used to match at least one virtual point with at least one corresponding marking point to obtain at least one group of virtual points and marking points that match each other; wherein, when a first condition is satisfied between a first feature description of a virtual point and a second feature description of a corresponding marking point, the virtual point and the marking point corresponding to the virtual point are matched with each other; and the first condition is set based on the similarity between the first feature description and the second feature description.

[0044] The present application also provides a registration device, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the registration method as described above when executing the computer program.

[0045] According to the above-mentioned embodiments of the registration method, device and equipment for the target of the surgical navigation and positioning system, the feature point contour extraction and binocular stereo matching calculation are performed on the collected target image to obtain the spatial point cloud data of each marked point on the target during the operation; and the virtual points in the above-mentioned point cloud data set are calculated in turn to obtain the first feature description of the virtual point; the marked points on the target are similarly calculated to determine the second feature description of each marked point on the real target; the point cloud data is registered with the target according to the feature description of each point, thereby dividing the point cloud data corresponding to each target, so as to achieve the purpose of real-time positioning and tracking of each surgical target during the operation. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 Flow chart (I) of the automatic handling method for assembly parts provided in the embodiment of the present application;

[0047] Figure 2 Flow chart (II) of the automatic handling method for assembly parts provided in the embodiment of the present application;

[0048] Figure 3 Flow chart of the automatic handling method of assembly parts provided in the embodiment of the present application (III);

[0049] Figure 4 Flow chart (IV) of the automatic handling method for assembly parts provided in the embodiment of the present application;

[0050] Figure 5 A principle block diagram of the automatic handling device for assembly parts provided in an embodiment of the present application;

[0051] Figure 6 A schematic diagram of the structure of the automatic handling system for assembly parts provided in an embodiment of the present application;

[0052] Figure 7 A schematic diagram of the structure of the automatic assembly handling system identification device provided in the embodiment of the present application;

[0053] Figure 8 A schematic diagram of the structure of the target of the automatic assembly handling system provided in the embodiment of the present application;

[0054] Fig. 9 A schematic diagram of virtual points in the target point cloud data provided in an embodiment of the present application.

[0055] In the figure: 501, point cloud data acquisition module; 502, first feature description acquisition module; 503, second feature description acquisition module; 504, matching module; 6, controller; 7, recognition device; 701, shell; 702, lens; 703, camera; 704, fill light; 8, target; 801, target body; 802, marking point; 901, virtual point. DETAILED DESCRIPTION

[0056] The present application is further described in detail below by specific embodiments in conjunction with the accompanying drawings. Wherein similar elements in different embodiments adopt associated similar element numbers. In the following embodiments, many detailed descriptions are intended to enable the present application to be better understood. However, those skilled in the art can easily recognize that some of the features can be omitted in different situations, or can be replaced by other elements, materials, and methods. In some cases, some operations related to the present application are not shown or described in the specification, in order to avoid the core part of the present application being overwhelmed by too much description, and for those skilled in the art, it is not necessary to describe these related operations in detail, and they can fully understand the related operations based on the description in the specification and the general technical knowledge in the art.

[0057] In addition, the features, operations or characteristics described in the specification can be combined in any appropriate manner to form various implementations. At the same time, the steps or actions in the method description can also be interchanged or adjusted in a manner that is obvious to those skilled in the art. Therefore, the various sequences in the specification and the drawings are only for the purpose of clearly describing a certain embodiment and are not meant to be a required sequence, unless otherwise specified that a certain sequence must be followed.

[0058] The serial numbers of the components in this document, such as "first", "second", etc., are only used to distinguish the objects described and do not have any order or technical meaning. The "connection" and "coupling" mentioned in this application, unless otherwise specified, include direct and indirect connections (couplings).

[0059] Please refer to Figure 1 In order to solve the technical problem that the optical positioning tracking method of the prior art in surgical navigation positioning technology is insufficient in maintaining real-time tracking accuracy and point cloud data analysis speed, an embodiment of the present application provides a registration method for a surgical navigation positioning system target, and the registration method includes the following steps:

[0060] Step S101, identifying a target to obtain point cloud data of the target; the target includes at least one marking point, the point cloud data includes at least one virtual point, and at least one virtual point corresponds to at least one marking point of the target.

[0061] In some embodiments, since there are multiple marking points on the target, point cloud data can be obtained by identifying the target, and each virtual point in the point cloud data has a corresponding marking point.

[0062] In some embodiments, the process of identifying the target utilizes visual sensing technology to accurately scan or photograph the target of the target surgical navigation and positioning system, such as a three-dimensional scanner, a stereo camera, a lidar, etc. These visual sensors can capture the position information of the marking points on the target surface, and convert the original optical data into structured three-dimensional point cloud data through corresponding data processing methods, such as point cloud generation methods, stereo matching methods, and beam echo processing. In the generated point cloud data, the marking points on the target are accurately mapped to virtual points, each virtual point corresponds to an actual marking point on the target, and carries its precise coordinate information in three-dimensional space.

[0063] In some embodiments, the marking point on the target may be a preset optical mark, a physical mark, or a biological anatomical feature.

[0064] Step S102, obtaining a first feature description of at least one virtual point in the point cloud data; the first feature description of the virtual point includes at least one first element, and the at least one first element is used to characterize the relative position information between the virtual point and at least one other virtual point around it.

[0065] In some embodiments, for each virtual point obtained in the point cloud data, a first feature description is further extracted. The first feature description includes information reflecting the relative position relationship between the virtual point and other virtual points around it, including but not limited to the distance, angle, direction, topological connection relationship, curvature characteristics, local density, etc. between the virtual points.

[0066] In some embodiments, feature extraction may use point cloud analysis methods or statistical methods, such as kd-tree, RANSAC, PCA, MST, etc., and statistical methods such as histogram, covariance matrix, Gaussian mixture model, etc. Through feature extraction, the point cloud data is deeply mined, and the position and morphological characteristics of the virtual point in the overall structure of the point cloud data are quantified through the first feature description, providing a key basis for subsequent matching with the actual marking point features.

[0067] Step S103, obtaining a second feature description of at least one marking point of the target; the second feature description of the marking point includes at least one second element, and the at least one second element is used to characterize the relative position information between the marking point and at least one other marking point around it.

[0068] In some embodiments, when obtaining the second feature description of the marking point on the target, an expression similar to the first feature description of the virtual point is used, and the second feature description of the marking point also includes the relative position information between the marking point and other marking points around it.

[0069] In some embodiments, the relative position information included in the second feature description can be obtained through direct measurement, medical image analysis and / or pre-operative planning data.

[0070] In some embodiments, the second feature description may also include physical characteristics, biological characteristics or medical imaging characteristics of the marker point, which together constitute a multimodal feature description of the marker point, which helps to improve the recognition and matching accuracy of the identity of the real marker point during the registration process. Among them, physical characteristics include size, shape, material, etc., biological characteristics include tissue structure, hemodynamic parameters, etc., and medical imaging characteristics include CT / MRI texture characteristics or PET metabolic activity, etc.

[0071] Step S104, matching at least one virtual point with at least one corresponding marked point to obtain at least one group of mutually matching virtual points and marked points; wherein, when a first condition is satisfied between a first feature description of a virtual point and a second feature description of a corresponding marked point, the virtual point and the marked point corresponding to the virtual point are mutually matched; and the first condition is set based on the similarity between the first feature description and the second feature description.

[0072] In some embodiments, after the extraction of the feature descriptions of the virtual points and the marker points is completed, the virtual points need to be matched with the corresponding marker points. The matching process is based on the similarity between the first feature description and the second feature description. When the similarity between the two meets the preset first condition, the virtual point and the corresponding marker point are determined to be matched with each other. The similarity calculation may use various measurement methods, such as Euclidean distance, Mahalanobis distance, cosine similarity, information entropy, KL divergence, etc. The specific selection depends on the feature type and application scenario. In some embodiments, the first condition is usually to set a threshold or probability threshold. Only when the similarity exceeds the threshold, the two are considered to be matched successfully. Through this matching process, the system is able to match the virtual points with the real markers one by one, thereby realizing the precise alignment of the target in the surgical navigation and positioning system.

[0073] In the above-mentioned registration method for the target of the surgical navigation and positioning system, as a preferred embodiment, the target is identified to obtain point cloud data of the target, including: identifying the target by an identification device to obtain point cloud data of the target; wherein the identification device includes a three-dimensional scanner, and / or a visual sensor, and / or an RGB-D camera, and / or a projector and a camera.

[0074] When the identification device is a three-dimensional scanner, the target is scanned by the three-dimensional scanner to obtain point cloud data about the marking points on the target. The three-dimensional scanner accurately determines the three-dimensional coordinates of each point on the target surface by emitting laser or structured light and receiving reflected signals, calculating the round-trip time or phase difference of the light. The scanner scans the target in all directions along a preset path or rotates in all directions to form dense point cloud data, which contains the precise three-dimensional position information of the marking points on the target. During the scanning process, the processing unit inside the scanner will identify and distinguish the marking points from the background or other structures to ensure that the point cloud data contains a complete and accurate representation of the target marking points.

[0075] When the recognition device is a binocular vision sensor, the target is recognized by the vision sensor and image processing is performed to obtain point cloud data about the marking points on the target. The binocular vision sensor uses two synchronously working cameras (usually with a known baseline distance) to shoot the same target scene from different perspectives. The same-name points are found between the left and right images through stereo matching algorithms (such as SIFT, SURF, ORB, etc.), and the positions of these corresponding points in three-dimensional space are calculated using the principle of triangulation, thereby constructing point cloud data of the target and marking points. The binocular vision system may be combined with deep learning methods for feature detection and matching to improve the robustness and accuracy of marking point recognition in complex environments.

[0076] When the recognition device is an RGB-D camera, the target is photographed by the RGB-D camera to obtain a depth image of the target, and point cloud data about the marked points on the target is obtained based on the depth image. The RGB-D camera integrates an RGB camera and a depth sensor to simultaneously obtain the RGB image and depth information of the scene at one time. The depth sensor (such as TOF, structured light, active binocular and other technologies) measures the distance from each pixel to the camera, combines the color information, and generates point cloud data containing color and depth values. For the marked points on the target, the RGB-D camera can directly output their precise position and color attributes in three-dimensional space to form a point cloud representation.

[0077] When the recognition device is a projector and a camera, the projector transmits a grating image to the target, and the camera captures the reflected image of the target to obtain point cloud data about the marked point on the target. The projector projects a specific grating pattern (such as coded structured light, sinusoidal fringe pattern, etc.) onto the target surface, and the camera synchronously captures the pattern reflected back from the target. By analyzing the deformation, phase or coding information of the pattern, the depth information between the projected point and the camera is solved, and a three-dimensional point cloud of the target surface is constructed. The depth information of the marked point as part of the target surface is calculated and included in the generated point cloud data.

[0078] Please refer to Figure 2In the above-mentioned registration method for the target of the surgical navigation and positioning system, as a preferred solution, when obtaining the first feature description of at least one virtual point in the point cloud data, for any virtual point P in the point cloud data i , then perform the following steps:

[0079] Step S201: Get the virtual point P i at least one other virtual point within a preset distance.

[0080] Step S202: According to the virtual point P i The virtual point P is obtained by combining the relative position information between the virtual point P and at least one other virtual point. i The first feature description P id . Where i = 1,…,n; n is the number of virtual points in the point cloud data; the first neighborhood point set P id At least one of the first elements corresponds to at least one other virtual point within a preset distance.

[0081] In the above-mentioned registration method for the target of the surgical navigation positioning system, as a preferred solution, at the virtual point P i The neighborhood point set P id At least one other virtual point is in accordance with the virtual point P i The distances of the virtual points are sorted from near to far, or at least one other virtual point is sorted according to the distance from the virtual point P i The distances are sorted from far to near.

[0082] Please refer to Figure 3 In the above-mentioned registration method for the target of the surgical navigation and positioning system, as a preferred solution, the obtaining of the second feature description of at least one marker point of the target comprises the following steps:

[0083] Step S301, obtaining the real position information of at least one marking point on the target; the real position information includes the three-dimensional coordinates of the marking point.

[0084] For any target point P Ij , then:

[0085] Step S302: Calculate and obtain the real position information of at least one marking point P Ij at least one other marked point within a preset distance.

[0086] Step S303: According to the marking point P Ij The set of relative position information between the marker point P and at least one other marker point is obtained Ij The second feature description P Ijd .

[0087] Where I represents the number of the target, P Ij Refers to the jth marked point of the Ith target; the second neighborhood point set P Ijd At least one of the second elements corresponds to at least one other marking point within a preset distance.

[0088] In the above-mentioned registration method for the target of the surgical navigation positioning system, as a preferred solution, at the marker point P Ij The neighborhood point set P Ijd At least one other marked point is in accordance with the marked point P Ij The distances of the points are sorted from near to far, or at least one other point is sorted according to the distance from the point P. Ij The distances are sorted from far to near.

[0089] Please refer to Figure 4 In the above-mentioned registration method for the target of the surgical navigation positioning system, as a preferred solution, the matching of at least one virtual point with at least one corresponding marking point comprises the following steps:

[0090] Step S401: Compare at least one of the first elements of the first feature description of any virtual point with at least one of the second elements of the second feature description of its corresponding marking point at least once.

[0091] Step S402: at each comparison, determine whether the difference between the first element and the second element is less than an error threshold: if so, increase the similarity between the first feature description and the second feature description; if not, do not increase the similarity between the first feature description and the second feature description.

[0092] Step S403: after the first element of the first feature description and the second element of the second feature description are compared, determine whether the similarity between the first feature description and the second feature description exceeds a preset threshold: if so, the first condition is satisfied between the first feature description and the second feature description, and the virtual points and marked points respectively corresponding to the first feature description and the second feature description are matched; if not, the first condition is not satisfied between the first feature description and the second feature description, and the virtual points and marked points respectively corresponding to the first feature description and the second feature description are not matched.

[0093] In the above-mentioned registration method for surgical navigation and positioning system targets, as a preferred solution, the first feature description P id is {l1,l2,…,l k}, the second feature description P Ijd is {d1,d2,…,d k};

[0094] Where: l1,l2,…,lk All are the first elements, and the relative position information of the first elements includes the relative distance between two virtual points; d1, d2, …, d k All are the second elements, and the relative position information of the second elements includes the relative distance between two marked points.

[0095] In some embodiments, the spatial point cloud data P of each marked point on all the targets is obtained by using the feature point contour extraction and binocular stereo matching algorithms. The points P in the above point cloud dataset are sequentially i (i = 1, …, n; n is the number of points) for kd_tree calculation. According to the Euclidean distance from any point to P i , search for the nearest k-neighborhood point set P i of P id , and store the k-neighborhood point sets P d {P 1d , …, P nd} of all points. Take the k-neighborhood point set of each point as the feature description of this point. During the operation, the positioning and tracking of the target are as shown in Figure 6 . Calculate the corresponding k-neighborhood point set P Ijd (I: refers to the I-th target, j: refers to the j-th marked point of the i-th target) according to the true position information <x, y, z> of the marked points in each target, and use this as the feature description of this point. Compare the elements in the feature descriptions P id {l1, l2, …, l k} of the points in the point cloud with the elements in the feature descriptions P Ijd {d1, d2, …, d k} of the marked points of the target one by one. When the element l id of P i and the element d Ijd of P j satisfy |l i - d j | < h (h: error threshold), the similarity value p ic of the two-point feature descriptions increases by one; when |p ic - H| > 0, it indicates that the point P i in the point cloud and the marked point P Ij on the target are a pair of matching points. Find all the matching points in the above method in sequence. According to the target number information represented by P Ij , divide the matched point cloud data to obtain the point cloud data corresponding to each target, so as to realize the real-time tracking of the surgical target during the operation.

[0096] Please refer to Figure 5The present application also provides a registration device for a surgical navigation and positioning system target, the registration device includes a point cloud data acquisition module 501, a first feature description acquisition module 502, a second feature description acquisition module 503 and a matching module 504, which will be analyzed in detail below.

[0097] The point cloud data acquisition module 501 is used to identify the target to obtain the point cloud data of the target; the target includes at least one marking point, and the point cloud data includes at least one virtual point, and at least one virtual point corresponds to at least one marking point of the target.

[0098] In some embodiments, the point cloud data acquisition module 501 identifies the target through an identification device to obtain point cloud data of the target; wherein the identification device includes a three-dimensional scanner, and / or a visual sensor, and / or an RGB-D camera, and / or a projector and a camera; when the identification device is a three-dimensional scanner, the target is scanned by the three-dimensional scanner to obtain point cloud data about the marked points on the target; when the identification device is a binocular visual sensor, the target is identified by the visual sensor and image processing is performed to obtain point cloud data about the marked points on the target; when the identification device is an RGB-D camera, the target is photographed by the RGB-D camera to obtain a depth image of the target, and the point cloud data about the marked points on the target are obtained according to the depth image; when the identification device is a projector and a camera, a grating image is transmitted to the target by the projector, and the reflected image of the target is captured by the camera to obtain point cloud data about the marked points on the target.

[0099] The first feature description acquisition module 502 is used to obtain a first feature description of at least one virtual point in the point cloud data; the first feature description of the virtual point includes at least one first element, and at least one first element is used to characterize the relative position information between the virtual point and at least one other virtual point around it.

[0100] In some embodiments, when the first feature description acquisition module 502 acquires the first feature description of at least one virtual point in the point cloud data, for any virtual point P in the point cloud data i , then proceed as follows: Get the value corresponding to the virtual point P i At least one other virtual point within a preset distance; according to the virtual point P i The virtual point P is obtained by combining the relative position information of at least one other virtual point with the virtual point P. i The first feature description P id ; Where i = 1, ..., n; n is the number of virtual points in the point cloud data; the first neighborhood point set P idAt least one of the first elements corresponds to at least one other virtual point within a preset distance.

[0101] In some embodiments, when the first feature description acquisition module 502 acquires the first feature description of at least one virtual point in the point cloud data, at the virtual point P i The neighborhood point set P id At least one other virtual point is in accordance with the virtual point P i The distances of the virtual points are sorted from near to far, or at least one other virtual point is sorted according to the distance from the virtual point P i The distances are sorted from far to near.

[0102] The second feature description acquisition module 503 is used to acquire the second feature description of at least one marking point of the target; the second feature description of the marking point includes at least one second element, and at least one second element is used to characterize the relative position information between the marking point and at least one other marking point around it.

[0103] In some embodiments, when acquiring the second feature description of at least one marking point of the target, the second feature description acquisition module 503 acquires the real position information of at least one marking point on the target; the real position information includes the three-dimensional coordinates of the marking point. Ij , then: according to the real position information of at least one marking point, calculate and obtain the position information of the marking point P Ij At least one other marking point within a preset distance; according to the marking point P Ij The set of relative position information between the marker point P and at least one other marker point is obtained Ij The second feature description P Ijd ; Where I represents the number of the target, P Ij Refers to the jth marked point of the Ith target; the second neighborhood point set P Ijd At least one of the second elements corresponds to at least one other marking point within a preset distance.

[0104] In some embodiments, when the second feature description acquisition module 503 acquires the second feature description of at least one marking point of the target, at the marking point P Ij The neighborhood point set P Ijd At least one other marked point is in accordance with the marked point P Ij The distances of the points are sorted from near to far, or at least one other point is sorted according to the distance from the point P. Ij The distances are sorted from far to near.

[0105] A matching module 504, wherein the matching module 504 is used to match at least one virtual point with at least one corresponding marking point to obtain at least one group of mutually matching virtual points and marking points; wherein, when a first condition is satisfied between a first feature description of a virtual point and a second feature description of a corresponding marking point, the virtual point and the marking point corresponding to the virtual point are mutually matched; and the first condition is set based on the similarity between the first feature description and the second feature description.

[0106] In some embodiments, when matching at least one virtual point with at least one corresponding marked point, the matching module 504 compares at least one of the first elements of the first feature description of any virtual point with at least one of the second elements of the second feature description of the corresponding marked point at least once; in each comparison, it is determined whether the difference between the first element and the second element is less than the error threshold: if so, the similarity between the first feature description and the second feature description is increased; if not, the similarity between the first feature description and the second feature description is not increased; after the first element of the first feature description and the second element of the second feature description are compared, it is determined whether the similarity between the first feature description and the second feature description exceeds a preset threshold: if so, the first condition is satisfied between the first feature description and the second feature description, and the virtual points and marked points respectively corresponding to the first feature description and the second feature description are matched with each other; if not, the first condition is not satisfied between the first feature description and the second feature description, and the virtual points and marked points respectively corresponding to the first feature description and the second feature description are not matched.

[0107] The present application also provides a registration device, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the registration method as described above when executing the computer program.

[0108] Please refer to Figure 6 , Figure 7 , Figure 8 and Fig. 9The embodiment of the present application also provides a registration system for a target of a surgical navigation and positioning system, the registration system is used to register the point cloud data obtained by identifying a target 8, and the registration system includes a recognition device 7 and a controller 6; the recognition device 7 is provided with an image acquisition end and an image output end, the image acquisition end acquires image data about the target 8 by acquiring an image of a target 8; the image output end is connected to the controller 6, and the image data of the recognition device 7 is output to the controller 6 through the image output end; the controller 6 is used to process the image data to acquire point cloud data about the target 8, and to register the point cloud data so as to acquire at least one group of virtual points 901 and marking points 802 that match each other; the target 8 includes at least one marking point 802, and the point cloud data includes at least one virtual point 901, and at least one of the virtual points 901 corresponds to at least one marking point 802 of the target 8; after receiving the image data output by the recognition device 7, the controller 6 recognizes the target 8 to acquire the target 8 point cloud data, and obtain a first feature description of at least one virtual point 901 in the point cloud data, and obtain a second feature description of at least one marking point 802 of the target 8, and match at least one virtual point 901 with the corresponding at least one marking point 802 to obtain at least one group of virtual points 901 and marking points 802 that match each other; wherein the first feature description of the virtual point 901 includes at least one first element, and at least one of the first elements is used to characterize the relative position information between the virtual point 901 and at least one other virtual point 901 around it; the second feature description of the marking point 802 includes at least one second element, and at least one of the second elements is used to characterize the relative position information between the marking point 802 and at least one other marking point 802 around it; when a first condition is satisfied between the first feature description of a virtual point 901 and the second feature description of the corresponding marking point 802, the virtual point 901 and the corresponding marking point 802 are matched with each other; the first condition is set based on the similarity between the first feature description and the second feature description.

[0109] In some embodiments, the recognition device 7 includes a three-dimensional scanner, and / or a visual sensor, and / or an RGB-D camera 703 , and / or a projector and a camera 703 .

[0110] In some embodiments, when the recognition device 7 includes a three-dimensional scanner, the target 8 is scanned by the three-dimensional scanner to obtain point cloud data about the marking point 802 on the target 8; when the recognition device 7 includes a binocular vision sensor, the target 8 is recognized by the vision sensor and image processing is performed to obtain point cloud data about the marking point 802 on the target 8; when the recognition device 7 includes an RGB-D camera 703, the target 8 is photographed by the RGB-D camera 703 to obtain a depth image of the target 8, and the point cloud data about the marking point 802 on the target 8 is obtained based on the depth image; when the recognition device 7 includes a projector and a camera 703, a grating image is transmitted to the target 8 by the projector, and the reflected image of the target 8 is captured by the camera 703 to obtain point cloud data about the marking point 802 on the target 8.

[0111] Please refer to Figure 7 In some embodiments, the identification device 7 is an optical positioning tracking device, which includes a shell 701, a lens 702 and a camera 703. The lens 702 and the camera 703 are respectively arranged on both sides of the shell 701. The lens 702 is used to track the target 8, and the camera 703 is used to capture the image of the marking point 802 on the target 8 to obtain the image data of the target 8; the image acquisition end of the identification device 7 is the camera 703.

[0112] In some embodiments, the recognition device 7 further includes a fill light 704 , and the fill light 704 is used to illuminate the target 8 to increase the brightness of the target 8 .

[0113] Please refer to Figure 8 In some embodiments, the target 8 includes a target body 801 , and at least one marking point 802 is set on the target body 801 , and the marking point 802 is used to display prominence on the target body 801 .

[0114] The controller 6 executes the above-mentioned registration method for the target of the surgical navigation positioning system.

[0115] Those skilled in the art will appreciate that all or part of the functions of the various methods in the above-mentioned embodiments can be implemented by hardware or by computer programs. When all or part of the functions in the above-mentioned embodiments are implemented by computer programs, the program can be stored in a computer-readable storage medium, and the storage medium can include: read-only memory, random access memory, disk, optical disk, hard disk, etc., and the program is executed by a computer to implement the above-mentioned functions. For example, the program is stored in the memory of the device, and when the program in the memory is executed by the processor, all or part of the above-mentioned functions can be implemented. In addition, when all or part of the functions in the above-mentioned embodiments are implemented by computer programs, the program can also be stored in a storage medium such as a server, another computer, disk, optical disk, flash disk or mobile hard disk, and can be downloaded or copied and saved in the memory of the local device, or the system of the local device is updated, and when the program in the memory is executed by the processor, all or part of the functions in the above-mentioned embodiments can be implemented.

[0116] The above specific examples are used to illustrate the present application, which is only used to help understand the present application and is not intended to limit the present application. For technicians in the technical field to which the present application belongs, they can also make some simple deductions, deformations or substitutions based on the ideas of the present application.

Claims

1. A registration method for a target of a surgical navigation and positioning system, characterized in that: The registration method comprises: Identify the target to obtain point cloud data of the target; the target includes at least one marking point, the point cloud data includes at least one virtual point, and at least one virtual point corresponds to at least one marking point of the target; Acquire a first feature description of at least one virtual point in the point cloud data; the first feature description of the virtual point includes at least one first element, and the at least one first element is used to characterize the relative position information between the virtual point and at least one other virtual point around it; Acquire a second feature description of at least one marking point of the target; the second feature description of the marking point includes at least one second element, and the at least one second element is used to characterize the relative position information between the marking point and at least one other marking point around it; Match at least one virtual point with at least one corresponding marked point to obtain at least one group of mutually matching virtual points and marked points; wherein, when a first condition is satisfied between a first feature description of a virtual point and a second feature description of a corresponding marked point, the virtual point and the marked point corresponding to the virtual point are mutually matched; and the first condition is set based on the similarity between the first feature description and the second feature description.

2. The target registration method for a surgical navigation and positioning system according to claim 1, characterized in that: The step of identifying the target to obtain point cloud data of the target includes: Identify the target through the identification device to obtain the point cloud data of the target; in: The recognition device includes a three-dimensional scanner, and / or a visual sensor, and / or an RGB-D camera, and / or a projector and a camera; When the identification device is a three-dimensional scanner, the target is scanned by the three-dimensional scanner to obtain point cloud data about the marking points on the target; When the recognition device is a binocular vision sensor, the target is recognized by the vision sensor and image processing is performed to obtain point cloud data about the marking points on the target; When the recognition device is an RGB-D camera, the target is photographed by the RGB-D camera to obtain a depth image of the target, and point cloud data about the marking points on the target are obtained according to the depth image; When the recognition device is a projector and a camera, the projector transmits a grating image to the target, and the camera captures a reflected image of the target to obtain point cloud data about the marking points on the target.

3. The target registration method for a surgical navigation and positioning system according to claim 1, characterized in that: The obtaining of a first feature description of at least one virtual point in the point cloud data comprises: For any virtual point P in the point cloud data i , then: Get the virtual point P i at least one other virtual point within a preset distance; According to the virtual point P i The virtual point P is obtained by combining the relative position information of at least one other virtual point with the virtual point P. i The first feature description P id ; Where i = 1, ..., n; n is the number of virtual points in the point cloud data; the first neighborhood point set P id At least one of the first elements corresponds to at least one other virtual point within a preset distance.

4. The target registration method for a surgical navigation and positioning system according to claim 3, characterized in that: At the virtual point P i The neighborhood point set P id At least one other virtual point is in accordance with the virtual point P i The distances of the virtual points are sorted from near to far, or at least one other virtual point is sorted according to the distance from the virtual point P i The distances are sorted from far to near.

5. The target registration method for a surgical navigation and positioning system according to claim 1, characterized in that: The step of obtaining a second feature description of at least one marking point of the target includes: Acquire the real position information of at least one marking point on the target; the real position information includes the three-dimensional coordinates of the marking point; For any target point P Ij , then: According to the real position information of at least one marking point, calculate and obtain the position information of the marking point P Ij at least one other marked point within a preset distance; According to the marking point P Ij The set of relative position information between the marker point P and at least one other marker point is obtained Ij The second feature description P Ijd ; Where I represents the number of the target, P Ij Refers to the jth marked point of the Ith target; the second neighborhood point set P Ijd At least one of the second elements corresponds to at least one other marking point within a preset distance.

6. The target registration method for a surgical navigation and positioning system according to claim 5, characterized in that: At the marked point P Ij The neighborhood point set P Ijd At least one other marked point is in accordance with the marked point P Ij The distances of the points are sorted from near to far, or at least one other point is sorted according to the distance from the point P. Ij The distances are sorted from far to near.

7. The target registration method for a surgical navigation and positioning system according to claim 1, characterized in that: The matching of at least one virtual point with at least one corresponding marked point comprises: Compare at least one of the first elements of the first feature description of any virtual point with at least one of the second elements of the second feature description of the corresponding marking point at least once; At each comparison, determine whether the difference between the first element and the second element is less than the error threshold: if so, increase the similarity between the first feature description and the second feature description; if not, do not increase the similarity between the first feature description and the second feature description; After the first element of the first feature description and the second element of the second feature description are compared, determine whether the similarity between the first feature description and the second feature description exceeds a preset threshold: if so, the first condition is satisfied between the first feature description and the second feature description, and the virtual points and marked points respectively corresponding to the first feature description and the second feature description are matched with each other; if not, the first condition is not satisfied between the first feature description and the second feature description, and the virtual points and marked points respectively corresponding to the first feature description and the second feature description are not matched.

8. The target registration method for a surgical navigation and positioning system according to claim 7, characterized in that: The first feature description P id is {l1,l2,…,l k }, the second feature description P Ijd is {d1,d2,…,d k }; in: l1,l2,…,l k are all first elements, and the relative position information of the first elements includes the relative distance between two virtual points; d1,d2,…,d k are both second elements, and the relative position information of the second elements includes the relative distance between the two marking points.

9. A registration device for a target of a surgical navigation and positioning system, characterized in that: The registration device comprises: A point cloud data acquisition module, wherein the point cloud data acquisition module is used to identify the target to acquire point cloud data of the target; the target includes at least one marking point, and the point cloud data includes at least one virtual point, and at least one virtual point corresponds to at least one marking point of the target; A first feature description acquisition module, the first feature description acquisition module is used to acquire a first feature description of at least one virtual point in the point cloud data; the first feature description of the virtual point includes at least one first element, and the at least one first element is used to characterize the relative position information between the virtual point and at least one other virtual point around it; A second feature description acquisition module, the second feature description acquisition module is used to acquire a second feature description of at least one marking point of the target; the second feature description of the marking point includes at least one second element, and the at least one second element is used to characterize the relative position information between the marking point and at least one other marking point around it; A matching module, wherein the matching module is used to match at least one virtual point with at least one corresponding marking point to obtain at least one group of virtual points and marking points that match each other; wherein, when a first condition is satisfied between a first feature description of a virtual point and a second feature description of a corresponding marking point, the virtual point and the marking point corresponding to the virtual point are matched with each other; and the first condition is set based on the similarity between the first feature description and the second feature description.

10. A registration device, comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the registration method according to any one of claims 1 to 8 when executing the computer program.

Images