An acupoint registration method based on the fusion of the midline and hip edge features

Abstract

The present invention discloses an acupoint registration method based on the fusion of the midline and hip edge features, which relates to the field of graphic image processing and includes: S1, constructing a standard coordinate system adapted to human acupoints; S2, obtaining a depth image based on a structured light camera to generate three-dimensional point cloud data of the human back; S3, identifying the midline of the human body from the three-dimensional point cloud data based on symmetry analysis; S4, identifying the hip edge from the three-dimensional point cloud data based on the detection of gradient extrema; S5, establishing a local human coordinate system based on the midline and hip edge of the human body; S6, using a non-linear registration algorithm to register the local human coordinate system with the standard coordinate system, and mapping the registered coordinates into the user's world coordinate system to complete the acupoint registration for people with different body types. The present invention provides an acupoint registration method based on the fusion of the midline and hip edge features to reduce the error caused by body type differences and improve the accuracy of acupoint identification.

Description

Technical Field

[0001] The present invention relates to the field of graphic image processing. More specifically, the present invention relates to an acupoint registration method based on the fusion of the midline and hip edge features. Background Art

[0002] The identification of traditional Chinese medicine acupoints is an important part of traditional Chinese medicine treatment. Currently, it mainly relies on manual experience and manual positioning, which has problems such as large errors, low efficiency, and poor repeatability. With the development of computer vision and deep learning technologies, the research and development of acupoint recognition systems have become an important direction in the field of intelligent traditional Chinese medicine. However, due to individual differences in the human body, posture changes, and complex lighting conditions, it is difficult for existing technologies to achieve high-precision and automated acupoint positioning.

[0003] Currently, acupoint positioning technologies are mainly divided into two categories:

[0004] One is the recognition method based on two-dimensional images. By using a camera or ordinary RGB images to collect the surface information of the human body, and combining image processing algorithms (such as edge detection, skin color segmentation) for acupoint positioning. The problems with this method are: it depends on lighting conditions and cannot process three-dimensional curved surfaces, is easily affected by posture changes, and has a large positioning error. For example, as proposed by Wang Hao et al. in "Computational Intelligence and Neuroscience", the positioning accuracy of acupoint recognition is low in complex scenarios (such as uneven lighting, occlusion, skin color interference) (DOI: PMC10659787).

[0005] The other is the fixed template matching method. A standard human body three-dimensional point cloud model is established, and the body surface data of the patient is matched through rigid registration. The problems with this method are: it matches the human body point cloud through a predefined three-dimensional template, but cannot adapt to individual body shape differences (such as spinal curvature, uneven muscle distribution), resulting in poor clinical adaptability. For example, when the displacement of the Zhongwan acupoint point cloud caused by respiratory movement > 4 mm, the system needs to be recalibrated;

[0006] In addition, there are also patents that propose to solve the problems existing in the existing acupoint recognition algorithms through the acupoint algorithm of the human body's transverse finger same-body inch. However, it can only be applied to the recognition of acupoints on the back of the human body. For example, the acupoint recognition algorithm of the human body's transverse finger same-body inch proposed in patent number CN116864079A is prone to measurement errors due to light changes, finger occlusion, or posture tilt when extracting hand data, and cannot handle three-dimensional deformation problems.

[0007] Therefore, it can be seen from the above that the main defects of the existing technologies are:

[0008] 1. Insufficient recognition accuracy: The two-dimensional method ignores depth information, and the three-dimensional template method cannot dynamically correct individual differences.

[0009] 2. Low processing efficiency: Traditional symmetry analysis needs to traverse the entire graph, with a high computational complexity (O(n2)), making it difficult to process in real time.

[0010] 3. Poor robustness: Gradient detection is prone to failure under noise interference, resulting in the deviation of the reference point. Summary of the Invention

[0011] An object of the present invention is to solve at least the above problems and / or defects and provide at least the advantages described hereinafter.

[0012] To achieve these objects and other advantages of the present invention, there is provided an acupoint registration method based on the fusion of the midline and hip edge features, including:

[0013] S1. Based on the analysis of the traditional Chinese medicine anatomy database and clinical measurement data, construct a standard coordinate system adapted to human acupoints;

[0014] S2. Obtain a depth image based on a structured light camera to generate three-dimensional point cloud data of the human back;

[0015] S3. Identify the midline of the human body from the three-dimensional point cloud data obtained in S2 based on symmetry analysis;

[0016] S4. Identify the hip edge from the three-dimensional point cloud data obtained in S2 based on dual-threshold second-order gradient detection;

[0017] S5. Establish a local human coordinate system based on the midline of the human body and the hip edge obtained in S3 and S4;

[0018] S6. Use a non-linear registration algorithm to register the local human coordinate system obtained in S5 with the standard coordinate system obtained in S1, and map the registered coordinates into the user's world coordinate system to complete acupoint registration for people with different body types.

[0019] Preferably, in S1, the standard coordinate system is obtained by using principal component analysis (PCA) and kernel density estimation (KDE), and is characterized by the following formula:

[0020]

[0021] In the above formula, P i represents the standard coordinates of the i th acupoint, n represents the total number of acupoint position samples in the data, K (⋅) is a kernel function used to represent the Euclidean distance between the point to be evaluated P and the sample point P j , h represents the bandwidth parameter.

[0022] Preferably, in S2, the three-dimensional point cloud data is characterized by the following formula:

[0023] P w = K -1 R [t] P c

[0024] In the above formula, P w is the world coordinate, K is the camera intrinsic matrix, R [t] is the viewpoint change matrix, P c is the three-dimensional point in the camera coordinate system.

[0025] Preferably, in S3, the process of identifying the midline of the human body through symmetry analysis is as follows:

[0026] S30. Based on the global highest point, roughly locate and directly search for the point set with the minimum depth value in the three-dimensional point cloud data P global , and use the points included in P global as reference points in turn to obtain the initial position of the candidate midline from the occipital region of the skull x c ;

[0027] S31. Based on a predetermined sliding step s , divide the depth image with a width of W into k overlapping windows, and s should be less than the width Δ of each window w ;

[0028] S32. Based on x c in S30, calculate the local symmetry error E ( y , x c ) of each window;

[0029] S33. Find the local minimum error position k from each window , and obtain the global midline position x center by weighted average to complete the global error integration and screening of the midline of the human body.

[0030] Preferably, in S30, the point setP global Characterized by the following formula:

[0031]

[0032] In the above formula, I ( x , y ) represents the depth value of the pixel point ( x , y ) in the depth map, and Ω is the predefined range of the occipital region of the skull;

[0033] In S32, the local symmetry error E ( y , x c ) is characterized by the following formula:

[0034]

[0035] In the above formula, δ is the boundary buffer parameter in the sliding window calculation;

[0036] In S33, the global midline position x center is characterized by the following formula:

[0037]

[0038] In the above formula, the weight W k is the reciprocal of the minimum error of the window k .

[0039] Preferably, in S4, the process of obtaining the hip edge through double-threshold second-order gradient detection is as follows:

[0040] S40. Calculate the depth gradient longitudinally along the depth map to identify the hip edge position through a second-order difference operator G y ( y );

[0041] S41. Based on the double-threshold mechanism, determine whether G y ( y ) is the hip edge.

[0042] Preferably, in S40, the G y ( y ) is obtained through the following formula:

[0043]

[0044] In the above formula,I line ( y ) is the depth value of the hip edge detection line, I line ( y - 1) is the depth value of the line above the hip edge detection line, I line ( y + 1) is the depth value of the line below the hip edge detection line;

[0045] In S41, the double - threshold mechanism is judged by the following formula G y ( y ) whether it is the hip edge;

[0046]

[0047] In the above formula, is the gradient threshold, is the depth mutation threshold, y edge is the hip edge, is the depth value of the line below the hip edge detection line, is the depth value of the hip edge detection line.

[0048] Preferably, in S5, the local human body coordinate system is based on the connection line of the global mid - line position x center as the x axis, and the connection line of the hip edge y edge as the y axis, and is constructed by the following formula:

[0049]

[0050] In the above formula, x、 y is the coordinate of the pixel point.

[0051] Preferably, in S6, the process of acupoint registration includes:

[0052] S61. Normalize the standard coordinate system by the following formula to obtain the standard acupoint position :

[0053]

[0054] In the above formula, W is the width of the depth image, H is the height of the depth image, is the abscissa of the local coordinate system, is the ordinate of the local coordinate system, is the abscissa after normalization, is the ordinate after normalization;

[0055] S62. Perform coordinate transformation and normalization processing on the depth image of the patient's back collected by the camera through the following formula to obtain the acupoint positions of the patient :

[0056]

[0057] In the above formula, x center is the position of the global midline, y edge is the hip edge, is the original coordinate of the depth image of the patient's back collected by the camera;

[0058] S63. Define the deformation field using the thin plate spline TPS model T ( P ), complete the mapping from the standard acupoint position to the acupoint position of the patient through the coordinate transformation of the following formula:

[0059]

[0060] In the above formula, γ is the smoothing term weight, T (.) is the thin plate spline TPS transformation function, n is the total number of samples of the acupoint positions in the data, is the second-order gradient operator, P is the point in the coordinate system;

[0061] S64. Map the registered acupoint coordinates back to the original coordinate system of the patient through the following formula to obtain the absolute position in the patient's actual coordinate system by restoring the proportional coordinates :

[0062] .

[0063] The present invention has at least the following beneficial effects:

[0064] First, by integrating the double-feature coordinate system of the midline and the hip edge, the present invention solves the problem that the acupoint positioning on the human back in the prior art is inaccurate and cannot effectively cope with body shape differences.

[0065] Second, through image processing technologies such as horizontal symmetry error analysis and second-order gradient detection, the present invention solves the problem of low data collection and processing efficiency in the prior art, which affects clinical applications.

[0066] Thirdly, through the gradient-depth dual-threshold detection mechanism, the present invention solves the problem in the prior art that gradient detection is prone to failure under noise interference, resulting in the deviation of the reference point.

[0067] Other advantages, objectives, and features of the present invention will be partially reflected by the following description and partially understood by those skilled in the art through the research and practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0068] Figure 1 It is a schematic diagram of the overall process when the method of the present invention is adopted;

[0069] Figure 2 It is a schematic diagram of three-dimensional image acquisition and coordinate conversion when the method of the present invention is adopted;

[0070] Figure 3 It is a schematic diagram of the connecting line of the global midline position obtained after the depth image of the human back is recognized by the method of the present invention;

[0071] Figure 4 It is a schematic diagram of the connecting line of the hip edge obtained after the depth image of the human back is recognized by the method of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0072] The following further describes the present invention in detail with reference to the drawings, so that those skilled in the art can implement it according to the description in the specification.

[0073] The present invention relates to an acupoint positioning system based on structured light three-dimensional depth imaging and reference point detection, including: a three-dimensional image acquisition module, and a terminal machine communicatively connected thereto, on which an acupoint registration software for acupoint recognition is installed. The acupoint registration software is provided with a midline and hip edge recognition module and an acupoint registration and mapping module to provide an acupoint registration method based on the fusion of midline and hip edge features to achieve high-precision acupoint recognition. Further, the acupoint registration method is to achieve high-precision acupoint recognition through symmetry analysis, gradient extreme value detection, and non-linear registration algorithms, and is applicable to intelligent auxiliary devices for traditional Chinese medicine treatments such as acupuncture and moxibustion.

[0074] Further, the present invention proposes an acupoint registration method based on the fusion of midline and hip edge features to achieve high-precision acupoint recognition, providing technical support for the standardization and intelligence of traditional Chinese medicine treatments. Specifically, the acupoint registration method of the present invention is an acupoint recognition method based on the image recognition technology of "highest point" selection. By establishing a standard acupoint database, symmetry analysis and gradient extreme value detection are carried out, and then the midline of the human back and the hip edge are determined, and then the acupoints are located by using a non-linear registration algorithm. As Figure 1 shown, its processing flow includes the following steps:

[0075] Step 1. Construct a human body standard acupoint database and a standard coordinate system: Based on the existing available traditional Chinese medicine knowledge, construct a depth coordinate system for human acupoints and normalize it as the standard coordinate system.

[0076] Step 2. Obtain a human back image through a structured light camera or a depth camera in the image acquisition module to generate three-dimensional point cloud data of the human body.

[0077] Step 3. Conduct symmetry analysis and gradient extreme value detection through the midline and hip edge recognition module to determine the midline of the human back and the hip edge, and then establish a local human coordinate system.

[0078] Step 4. Register the local human coordinate system with the standard coordinate system through the non-linear registration algorithm in the acupoint registration and mapping module, and then map the registered coordinates into the user's world coordinates.

[0079] Furthermore, in Step 1, the construction of the standard acupoint template is based on the traditional Chinese medicine anatomy database and clinical measured data to establish the standard acupoint template representing the coordinates of the i th acupoint. The standard acupoint template models the acupoint distribution through the principal component analysis (PCA) and kernel density estimation (KDE) methods in the following formula:

[0080]

[0081] where P i represents the standard coordinates of the i th acupoint, n represents the total number of acupoint position samples in the data, K(⋅) represents the kernel function, the Euclidean distance between the point P to be evaluated and the sample point P j , and h represents the bandwidth parameter.

[0082] Furthermore, in Step 2, as Figure 2 shown, the three-dimensional image acquisition module uses a structured light camera to obtain depth information Z(x, y) and generate three-dimensional point cloud data of the back. Its three-dimensional coordinate calculation formula is:

[0083] P w =K -1 R [t] P c

[0084] where P w is the world coordinate, K is the camera internal parameter matrix, R [t] is the viewpoint change matrix, P cis a three-dimensional point in the camera coordinate system.

[0085] Furthermore, in this step, statistical outlier rejection (SOR) is used to remove outliers, and Gaussian filtering is used to smooth the point cloud surface to reduce the impact of noise on subsequent symmetry analysis.

[0086] Furthermore, in step three, the midline and hip edge recognition module can include, according to the working mode: a midline recognition sub-module and a hip edge recognition and feature extraction sub-module;

[0087] Among them, the working process of the midline recognition sub-module includes:

[0088] 1. By the method of rough positioning of the global highest point, directly search for the point set with the minimum depth value in the three-dimensional point cloud data P global (skull occipital region):

[0089]

[0090] In the above formula, I ( x , y ) represents the depth value of the pixel point ( x , y ) in the depth map, Ω is the predefined range of the skull occipital region (here 20 is taken), and the points included in P global are used as reference points in turn, and the column where the reference point is located x c is used as the initial position of the candidate midline.

[0091] 2. Determine the midline position by calculating the horizontal symmetry error in small batches with a sliding window

[0092] ① Window division:

[0093] Divide the width W of the depth map into k overlapping windows, each window with a width of Δ w , and the sliding step size is s (s < Δ w to avoid gaps and cover the whole map), then the column range of the k -th window is:

[0094]

[0095] ② Calculation of symmetry error within the window:

[0096] For each window k , calculate the candidate midline position through the following formula x cLocal symmetry error:

[0097]

[0098] Wherein, x c is the initial position of the candidate midline, I ( x , y ) represents the depth value of the pixel point ( x , y ) in the depth map, δ is the boundary buffer parameter in the sliding window calculation. When calculating the symmetry error within the sliding window, if x c is too close to the window edge, it may cause some symmetric points ( x c - x , y ) or ( x c + x , y ) to exceed the current window range.

[0099] ③ Global error integration and midline screening:

[0100] For each window k , find the local minimum error position as the optimal solution for this window:

[0101]

[0102] Take the weighted average of all window optimal solutions as the global midline position x center :

[0103]

[0104] Wherein, the weight W k is the reciprocal of the minimum error of window k , ensuring that high-confidence windows dominate the result.

[0105] The working process of the hip edge recognition and feature extraction sub-module includes:

[0106] (1) Calculate the depth gradient longitudinally along the depth map, and identify the hip edge position through the second-order difference operator.

[0107] Define the longitudinal gradient of the hip edge position G y ( y ) as:

[0108] (2) The hip edge is positioned as follows:

[0109]

[0110] In the above formula, is the gradient threshold for screening significant mutations (here = 10 mm), is the depth mutation threshold (here = 15 mm), excluding minor fluctuations.

[0111] After the hip edge and the midline are recognized by the midline and hip edge recognition module, a local human coordinate system is constructed. The local human coordinate system uses the connection line of the global midline position x center as the x-axis and the connection line of the hip edge y edge as the y-axis for construction, that is:

[0112]

[0113] In this step, the second-order gradient is mainly used to be more sensitive to depth mutations to reduce noise interference, and then the double-threshold mechanism is used to ensure the robustness of the hip edge positioning. Specifically, the local human coordinate system constructed in this step can effectively fuse the midline and the hip edge, reduce the error caused by body shape differences, improve the accuracy of acupoint recognition, and at the same time, in the application of the midline and hip edge recognition module, the gradient-depth double-threshold detection is used to enhance the robustness of hip edge recognition and resist noise interference.

[0114] Furthermore, in step four, the workflow of the acupoint registration and mapping module includes:

[0115] (1) Normalize the standard acupoint template in the coordinate system , that is:

[0116]

[0117] Similarly, the actual original coordinates of the patient also undergo the above changes to obtain the acupoint positions of the patient , then there is:

[0118]

[0119] (2) Adopt the thin plate spline TPS model to define the deformation field T ( P ), that is:

[0120]

[0121] ① Linear term: , describes the global affine transformation (translation, rotation, scaling).

[0122] ② Nonlinear term: , describes the local deformation, P j is the control point (taking the standard acupoint P i position, also known as the sample point), P is the point in the coordinate system, φ (.) is the radial basis function, W j is the weight coefficient, m is the number of control points, r is the point P and the control point P j the Euclidean distance between.

[0123] ③ Radial basis function: , minimizes the bending energy and ensures smooth deformation.

[0124] Through coordinate system transformation, map the standard acupoint position to the patient's acupoint position , that is, substitute T ( P ) into the following formula to optimize the objective:

[0125]

[0126] In the above formula, represents the control point position (taking the key acupoints of the standard template), γ the smooth term weight (here take γ = 0.1), suppresses excessive deformation. Considering the local deformation of the human body, the TPS model can simultaneously describe the global linear transformation and local nonlinear deformation, adapt to spinal curvature and muscle distribution differences, and achieve precise alignment between the standard acupoint template and the patient's body. Normalized coordinates eliminate size differences and improve the generalization of registration.

[0127] (3) Map the registered acupoint coordinates back to the patient's original coordinate system, that is:

[0128]

[0129] is to restore the proportional coordinates to the absolute position in the patient's actual coordinate system.

[0130] The acupoint registration and mapping module in this solution eliminates size differences through normalized TPS registration and improves the cross-individual generalization ability.

[0131] Example 1: Recognition of adult back acupoints

[0132] The depth image is adopted by the structured light camera to input a depth image with W = 512 and H = 1024 into the algorithm processing software of the terminal.

[0133] The midline and hip edge recognition module of the present invention is adopted to detect the midline and hip edge of the human back in the input image, and obtain as Figures 3 - 4 shown x center = 256, y edge = 320;

[0134] The acupoint registration and mapping module of the present invention is adopted to perform normalization first, so as to convert the standard acupoint of the Shenshu acupoint P i =(256, 200) into (0.5, 0.25).

[0135] The acupoint registration and mapping module of the present invention is adopted to perform TPS registration and output the registration result: TPS outputs =(0.51, 0.24), and the actual coordinates after inverse transformation are (256.5, 324), completing the acupoint registration operation of the coordinates.

[0136] The traditional method (such as the existing acupoint recognition method based on deep learning) is compared with the method of the present invention from different indexes (such as processing time, algorithm complexity, average error, maximum error) to prove the beneficial effect of the method of the present invention. The comparison results are shown in Table 1. It can be seen that the present invention has a significant improvement in each index compared with the existing traditional method.

[0137] Table 1

[0138]

[0139] The above scheme is only an illustration of a preferred example, but is not limited thereto. When implementing the present invention, appropriate substitution and / or modification can be made according to the needs of users.

[0140] Although the embodiments of the present invention have been disclosed as above, it is not limited to the applications listed in the specification and embodiments. It can be fully applied to various fields suitable for the present invention. For those familiar with the field, additional modifications can be easily achieved. Therefore, without departing from the general concept defined by the claims and the equivalent scope, the present invention is not limited to the specific details and the illustrated examples here.

Claims

1. An acupoint registration method based on the fusion of midline and buttock edge features, characterized in that: include: S1. Based on the analysis of the TCM anatomical database and clinical measured data, a standard coordinate system adapted to human acupuncture points is constructed; S2, acquiring a depth image based on a structured light camera to generate three-dimensional point cloud data of the back of the human body; S3, identifying the human body midline from the three-dimensional point cloud data obtained in S2 based on symmetry analysis; S4, identifying the edge of the buttocks from the three-dimensional point cloud data obtained in S2 based on double threshold second-order gradient detection; S5, establishing a local human body coordinate system based on the human body midline and hip edge obtained in S3 and S4; S6, using a nonlinear registration algorithm to register the local human body coordinate system obtained in S5 with the standard coordinate system obtained in S1, and mapping the registered coordinates to the user's world coordinate system to complete the acupoint registration for people of different body shapes; In S4, the process of obtaining the buttocks edge through double-threshold second-order gradient detection is: S40, calculating the depth gradient along the depth map longitudinally to identify the hip edge position by a second-order difference operator G y ( y ); S41, based on the dual threshold mechanism G y ( y ) is the hip edge y edge Make a determination; In S40, the G y ( y ) is obtained by the following formula: In the above formula, I line ( y ) is the depth value of the hip edge detection line, I line ( y -1) is the depth value of the previous line of the hip edge detection line, I line ( y +1) is the depth value of the next line of the hip edge detection line; In S41, the dual threshold mechanism is determined by the following formula G y ( y ) is the buttock edge; In the above formula, is the gradient threshold, is the deep mutation threshold, y edge For the hip edge, is the depth value of the next line of the hip edge detection line, The depth value of the hip edge detection line.

2. The acupoint registration method based on the fusion of midline and buttock edge features as claimed in claim 1, characterized in that: In S1, the standard coordinate system is obtained by principal component analysis PCA and kernel density estimation KDE, and is characterized by the following formula: In the above formula, P i Indicates i The standard coordinates of the acupuncture points, n Represents the total number of acupoint location samples in the data, K (⋅) is the kernel function, which is used to represent the point to be evaluated. P With sample points P j The Euclidean distance between h Indicates the bandwidth parameter.

3. The acupoint registration method based on the fusion of midline and buttock edge features as claimed in claim 1, characterized in that: In S2, the three-dimensional point cloud data is represented by the following formula: P w = K -1 R [t] P c In the above formula, P w is the world coordinate, K is the camera intrinsic parameter matrix, R [t] is the viewpoint change matrix, P c is a 3D point in the camera coordinate system.

4. The acupoint registration method based on the fusion of midline and buttock edge features as claimed in claim 1, characterized in that: In S3, the process of identifying the human body midline through symmetry analysis is as follows: S30, based on the global highest point rough positioning, directly search for the point set with the smallest depth value in the three-dimensional point cloud data P global ,by P global The included points are used as reference points in turn to obtain the initial position of the candidate midline from the occipital region of the skull x c ; S31, based on a predetermined sliding step length s Set the width to W The depth image is divided into k overlapping windows, and s should be smaller than the width of each window Δ w ; S32, based on S30 x c The local symmetry error for each window E ( y , x c ) to perform calculations; S33. From every window k Find the local minimum error position , to obtain the global midline position by weighted average x center , complete the global error integration and the screening of the human body midline.

5. The acupoint registration method based on the fusion of midline and buttock edge features as claimed in claim 4, characterized in that: In S30, the point set P global Characterized by the following formula: In the above formula, I ( x , y ) represents pixel points ( x , y ), A predefined range for the occipital region of the skull; In S32, the local symmetry error E ( y , x c ) is characterized by the following formula: In the above formula, δ is the boundary buffer parameter in the sliding window calculation; In S33, the global midline position x center Characterized by the following formula: In the above formula, the weight W k For Window k The inverse of the minimum error.

6. The acupoint registration method based on the fusion of midline and buttock edge features as claimed in claim 1, characterized in that: In S5, the local human body coordinate system The global midline position x center The connection line is x Shaft, hip edge y edge The connection line is y The axis is constructed by: In the above formula, x , y is the coordinate of the pixel point.

7. The acupoint registration method based on the fusion of midline and buttock edge features as claimed in claim 1, characterized in that: In S6, the acupoint registration process includes: S61, normalize the standard coordinate system by the following formula to obtain the standard acupuncture point position : In the above formula, W is the width of the depth image, H is the height of the depth image, is the abscissa of the local coordinate system, is the ordinate of the local coordinate system, is the normalized horizontal axis, is the normalized vertical coordinate; S62, coordinate transformation and normalization processing are performed on the patient's back depth image acquired by the camera to obtain the patient's acupuncture point position by the following formula: : In the above formula, x center is the global midline position, y edge For the hip edge, The original coordinates of the patient's back depth image collected by the camera; S63, using thin plate spline TPS model to define the deformation field T ( P ), the standard acupoint position is completed by the following coordinate system transformation To the patient's acupuncture point location The mapping: In the above formula, γ is the smoothing term weight, T (.) is the thin plate spline TPS transformation function, n is the total number of samples of acupoint locations in the data, is the second-order gradient operator, P is a point in the coordinate system; S64, mapping the registered acupoint coordinates back to the patient's original coordinate system through the following formula: : 。

Images