Method, apparatus, acupuncture robot, and storage medium for identifying position of acupoints on side of head

Abstract

Provided are a method, an apparatus, an acupuncture robot, and a storage medium for identifying the positions of acupoints on the lateral side of the head, which improve the accuracy and efficiency of identifying the positions of acupoints on the lateral side of the head. **Solution**: The method includes training an acupoint identification model based on a set of acupoint samples on the lateral side of the head to obtain a target acupoint identification model. The set of acupoint samples on the lateral side of the head includes a plurality of acupoint sample points and a plurality of support sample points located on the curve of the Gallbladder Meridian of Foot-Shaoyang. Inputting an image of the lateral side of the head to be identified into the target acupoint identification model to obtain a plurality of acupoint points and a plurality of support points, fitting the plurality of acupoint points and the plurality of support points to obtain a fitting curve, determining whether it is necessary to correct the acupoint points based on the fitting curve, and if so, correcting the acupoint points based on the fitting curve to obtain the target acupoint points.

Description

Technical Field

[0001] The present invention belongs to the technical field of acupoint identification, and relates to a method for identifying the positions of acupoints on the side of the head, an apparatus, an acupuncture robot, and a storage medium.

Background Art

[0002] There are nine meridians in the human head, namely the Stomach Meridian of Foot-Yangming, the Gallbladder Meridian of Foot-Shaoyang, the Bladder Meridian of Foot-Taiyang, the Large Intestine Meridian of Hand-Yangming, the Triple Energizer Meridian of Hand-Shaoyang, the Small Intestine Meridian of Hand-Taiyang, the Conception Vessel, the Governor Vessel, and the Liver Meridian of Foot-Jueyin. There are also many acupoints on the head, including Yangbai Point, Xuannao Point, Qubin Point, Jiaosun Point, Luyi Point, Benshen Point, Toulinqi Point, Chenglin Point, Naokong Point, Hanyan Point, Shangguan Point, etc. The acupoints on the side of the head are widely applied in health care and medicine, and each acupoint has different effects. Among them, the acupoints on the side of the head have very important significance and effects in health care and medicine.

[0003] Currently, the identification and positioning of acupoints on the side of the head mainly rely on the long-term experience and techniques of traditional Chinese medicine doctors. Doctors manually analyze the characteristics of the patient's head and combine the theoretical knowledge of how the meridians run and distribute in the head to determine the approximate positions of the head acupoints. However, this traditional method is strongly influenced by personal subjectivity, and different doctors may make different judgments for the same patient, resulting in inconsistent acupoint positioning results. Furthermore, it takes a large amount of time and effort to identify acupoints, with low efficiency and high requirements for the experience of traditional Chinese medicine doctors.

[0004] In order to solve these problems, it is necessary to provide a method for identifying the positions of acupoints on the side of the head, an apparatus, an acupuncture robot, and a storage medium to improve the accuracy and efficiency of acupoint position identification on the side of the head, support doctors in diagnosing the patient's condition, and reduce the burden on doctors.

Summary of the Invention

Problems to be Solved by the Invention

[0005] In view of the above, in order to solve the technical problem in the prior art that the accuracy of identifying the position of the acupoints on the side of the head is low, the present invention needs to provide a method and apparatus for identifying the position of the acupoints on the side of the head, an acupuncture robot, and a storage medium.

[0006] In one aspect, in order to solve the above technical problem, the present invention provides a method for identifying the position of the acupoints on the side of the head, training an acupoint identification model based on a set of acupoint samples on the side of the head to obtain a target acupoint identification model, wherein the set of acupoint samples on the side of the head includes a plurality of acupoint sample points and a plurality of support sample points located on the curve of the Gallbladder Meridian of Foot-Shaoyang; inputting an image of the side of the head to be identified into the target acupoint identification model to obtain a plurality of acupoint points and a plurality of support points; fitting the plurality of acupoint points and the plurality of support points to obtain a fitting curve; identifying whether it is necessary to correct the acupoint points based on the fitting curve, and if so, correcting the acupoint points based on the fitting curve to obtain target acupoint points.

[0007] In a possible embodiment, before training the acupoint identification model based on the set of acupoint samples on the side of the head, acquiring a plurality of sample images of the side of the head, and identifying the curve of the Gallbladder Meridian of Foot-Shaoyang in each sample image of the side of the head; annotating acupoint points based on the curve of the Gallbladder Meridian of Foot-Shaoyang to obtain a plurality of acupoint sample points; dividing the curve of the Gallbladder Meridian of Foot-Shaoyang based on the inflection points of the curve of the Gallbladder Meridian of Foot-Shaoyang to obtain a plurality of segments of the sub-curve of the Gallbladder Meridian of Foot-Shaoyang; further including determining at least one support sample point inserted into the sub-curve of the Gallbladder Meridian of Foot-Shaoyang based on the number of acupoint sample points and the curve characteristics in each segment of the sub-curve of the Gallbladder Meridian of Foot-Shaoyang.

[0008] In a possible embodiment, the curve characteristics include the curvature change rate and the curve length. Based on the number of the acupoint sample points and the curve characteristics in the sub-curve of the Gallbladder Meridian of Foot-Shaoyang of each segment, determining at least one support sample point inserted into the sub-curve of the Gallbladder Meridian of Foot-Shaoyang includes: When the curvature change rate of the sub-curve of the Gallbladder Meridian of Foot-Shaoyang is less than or equal to the first threshold value, the total number of support sample points and acupoint sample points on the sub-curve of the Gallbladder Meridian of Foot-Shaoyang is the first threshold value number. Based on the first threshold value number and the number of the acupoint sample points, determining the target number of support sample points inserted into the sub-curve of the Gallbladder Meridian of Foot-Shaoyang; When the curvature change rate of the sub-curve of the Gallbladder Meridian of Foot-Shaoyang is greater than the first threshold value and the curve length of the sub-curve of the Gallbladder Meridian of Foot-Shaoyang is less than or equal to the length threshold value, the total number of support sample points and acupoint sample points on the sub-curve of the Gallbladder Meridian of Foot-Shaoyang is the second threshold value number. Based on the second threshold value number and the number of the acupoint sample points, determining the target number of support sample points inserted into the sub-curve of the Gallbladder Meridian of Foot-Shaoyang; When the curvature change rate of the sub-curve of the Gallbladder Meridian of Foot-Shaoyang is greater than the first threshold value and the curve length of the sub-curve of the Gallbladder Meridian of Foot-Shaoyang is greater than the length threshold value, the total number of support sample points and acupoint sample points on the sub-curve of the Gallbladder Meridian of Foot-Shaoyang is the third threshold value number. Based on the third threshold value number and the number of the acupoint sample points, determining the target number of support sample points inserted into the sub-curve of the Gallbladder Meridian of Foot-Shaoyang; Inserting the support sample points of the target number into the sub-curve of the Gallbladder Meridian of Foot-Shaoyang; including Here, the first threshold value number is smaller than the second threshold value number, and the second threshold value number is smaller than the third threshold value number.

[0009] In a possible implementation manner, for obtaining a fitting curve by fitting the plurality of acupoint points and the plurality of support points, identifying a plurality of curve segments based on the plurality of acupoint points and the plurality of support points, where the curve segments are relaxation curve segments or circular curve segments, and each of the curve segments includes at least one of the acupoint points and at least one of the support points.

[0010] In a possible implementation manner, identifying whether it is necessary to correct the acupoint based on the fitting curve includes judging whether the distance between each acupoint and the fitting curve is greater than zero and less than a distance threshold; when the distance between each acupoint and the fitting curve is zero, judging that the acupoint does not need to be corrected; when the distance between each acupoint and the fitting curve is greater than or equal to the distance threshold, removing the acupoint; and when the distance between each acupoint and the fitting curve is greater than zero and less than the distance threshold, judging that the acupoint needs to be corrected.

[0011] In a possible implementation manner, correcting the acupoint based on the fitting curve to obtain a target acupoint includes obtaining a tangent line of the fitting curve, and taking a direction perpendicular to the tangent line and in the direction of the fitting curve as the moving direction of the acupoint; using the distance between the acupoint and the fitting curve as a moving distance, and controlling the acupoint to move along the moving direction by the moving distance to obtain the target acupoint.

[0012] In a possible implementation manner, the head side acupoint sample set includes a plurality of head side sample images of different persons, a plurality of head side sample images of the same person at the same distance at different angles, and a plurality of head side sample images of the same person at the same angle at different distances.

[0013] In another aspect, the present invention further provides a device for identifying the position of the head side acupoint, a model training unit that trains an acupoint identification model based on a head side acupoint sample set to obtain a target acupoint identification model, where the head side acupoint sample set includes a plurality of acupoint sample points and a plurality of support sample points located on the Gallbladder Meridian of Foot-Shaoyang curve; An identification target image identification unit that inputs a head side image of an identification target into the target acupoint identification model to obtain a plurality of acupoint points and a plurality of support points, A curve fitting unit that fits the plurality of acupoint points and the plurality of support points to obtain a fitting curve, An acupoint position identification unit that specifies whether it is necessary to correct the acupoint points based on the fitting curve, and if correction is necessary, corrects the acupoint points based on the fitting curve to obtain target acupoint points.

[0014] On the other hand, the present invention further provides an acupuncture robot, including a memory and a processor, The memory stores a program, The processor is coupled to the memory and executes the program stored in the memory to implement the steps of the method for identifying the positions of the head side acupoints described in any of the above possible embodiments.

[0015] On the other hand, the present invention further provides a computer-readable storage medium, in which programs and instructions are stored, and when the programs and instructions are executed by a processor, the steps of the method for determining the positions of the head side acupoints in any of the above embodiments are realized.

Advantages of the Invention

[0016] The beneficial effect of the present invention is that the method for identifying the positions of the head side acupoints provided by the present invention uses a deep learning model, namely, a target acupoint identification model, to identify acupoint points, thereby reducing the burden on doctors and improving the identification efficiency of the head side acupoints. In addition, the head side acupoint sample set of the present invention includes a plurality of acupoint sample points and a plurality of support sample points located on the curve of the Gallbladder Meridian of Foot-Shaoyang. That is, by restricting the acupoint sample points to be located on the curve of the Gallbladder Meridian of Foot-Shaoyang, the accuracy of the acupoint points in the head side acupoint samples is ensured, thereby improving the accuracy of the acupoint points identified by the target acupoint identification model, that is, improving the accuracy of identifying the positions of the head side acupoints.

[0017] Furthermore, the present invention installs a plurality of identified acupoint points and a plurality of support points to fit and obtain a fitting curve, and corrects the acupoint points that need to be corrected based on the fitting curve, thereby further improving the accuracy of the determined target acupoint points, that is, further improving the positioning accuracy of the head acupoint positioning.

Brief Description of the Drawings

[0018] To more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required for the description of the embodiments are briefly introduced below. Obviously, the drawings in the following description are only part of the embodiments of the present invention, and those skilled in the art can obtain other drawings based on these drawings without creative efforts.

[0019]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Best Mode for Carrying Out the Invention

[0020] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative efforts belong to the protection scope of the present invention.

[0021] It should be understood that the schematic drawings are not drawn to actual scale. The flowcharts used in the present invention show operations based on some embodiments of the present invention. The operations in the flowchart can be implemented in any order, and steps without a logical hierarchical relationship can be reversed or implemented simultaneously. Also, those skilled in the art can add one or more other operations to the flowchart or delete one or more operations from the flowchart based on the content of the present invention. Some of the block diagrams shown in the drawings are functional entities and do not necessarily correspond to physically or logically independent entities. These functional entities can be realized in the form of software, or in one or more hardware modules or integrated circuits, or in different networks and / or processor systems and / or microcontroller systems.

[0022] As used herein, the term "embodiment" means at least one embodiment of the present invention incorporating the specific features, structures, or characteristics described in the description of the embodiment. When the term appears at each position in the specification, it does not necessarily refer to the same embodiment, nor does it mean an independent or alternative embodiment mutually exclusive with other embodiments. What those skilled in the art understand explicitly and implicitly is that the embodiments described in this specification can be combined with other embodiments.

[0023] The present invention provides a method and apparatus for identifying the positions of acupoints on the lateral side of the head, an acupuncture robot, and a storage medium, which will be described below respectively.

[0024] FIG. 1 is a schematic flowchart of an embodiment of a method for identifying the positions of acupoints on the lateral side of the head provided by the present invention. As shown in FIG. 1, the method for identifying the positions of acupoints on the lateral side of the head includes: S101: Training an acupoint identification model based on a set of acupoint samples on the lateral side of the head to obtain a target acupoint identification model. The set of acupoint samples on the lateral side of the head includes a plurality of acupoint sample points and a plurality of support sample points located on the curve of the Gallbladder Meridian of Foot-Shaoyang. S102: Inputting an image of the lateral side of the head to be identified into the target acupoint identification model to obtain a plurality of acupoint points and a plurality of support points. S103: Fitting a plurality of acupoint points and a plurality of support points to obtain a fitting curve. S104: Determining whether it is necessary to correct the acupoint points based on the fitting curve. If correction is necessary, correcting the acupoint points based on the fitting curve to obtain target acupoint points.

[0025] It should be understood that when there is no need to correct the acupoint points, the acupoint points directly become the target acupoint points.

[0026] Here, the identification model based on the set of acupoint samples on the lateral side of the head in step S101 can include processes of training, testing, and verification. Specifically, the set of acupoint samples on the lateral side of the head is divided into a training set, a test set, and a verification set at a preset ratio, for example, a ratio of 8:1:1. The acupoint identification model is trained based on the training set. After the training is completed, the model performance of the trained acupoint identification model is tested based on the test set. If the test is passed, the acupoint identification model is verified based on the verification set. After the verification is passed, the target acupoint identification model is obtained. If the test fails or the verification fails, the acupoint identification model is retrained based on the training set.

[0027] Specifically, the acupoint identification model is YOLOv8, which has been improved and innovated based on YOLOv5 and YOLOv7. It uses Anchor-Free instead of Anchor-Base, replaces the C3 module with the lightweight C2f module, and reduces the computational load while maintaining detection accuracy. Additionally, YOLOv8 uses the SPPF module and the PAN-FPN structure, and improves the loss function and matching strategy to further enhance detection performance and efficiency. The execution process of the YOLOv8 network structure is as follows. The input image is converted into an RGB image of 640×640×3 and preprocessed through the CBS (Conv BatchNorm SiLU) module, which includes mosaic augmentation, spatial perturbation, and color perturbation. The CBS module consists of convolution, normalization, and the SiLU activation function, and is configured to extract texture and color information to solve the gradient vanishing and explosion problems, while also improving non-linear transformation. Next is the C2f module, which contains a convolutional kernel with downsampling and two branches. One branch includes split and concat operations, and the other branch includes the serial and parallel outputs of three Bottleneck networks. Finally, the features extracted from the overlapping of the features of the two branches are obtained. The input and output channel sizes of the C2f layer are the same. Finally, a maximum pooling operation is performed on the feature image using a 5×5 convolutional kernel to extract features. In the Head layer, YOLOv8 borrows the idea of the decoupled head of YOLOX, separates the regression branch and the prediction branch, and extracts category features and position features respectively.

[0028] Compared with the existing technologies, the method for identifying the positions of the acupoints on the side of the head provided by the embodiments of the present invention uses a deep learning model called the target acupoint identification model to identify acupoint points, thereby reducing the burden on doctors and improving the identification efficiency of the acupoints on the side of the head. Further, the acupoint sample set on the side of the head in the embodiments of the present invention includes a plurality of acupoint sample points and a plurality of support sample points located on the curve of the Gallbladder Meridian of Foot-Shaoyang. That is, by restricting the acupoint sample points to be located on the curve of the Gallbladder Meridian of Foot-Shaoyang, the accuracy of the acupoint points in the acupoint sample on the side of the head is ensured, thereby improving the accuracy of the acupoint points identified by the target acupoint identification model, that is, improving the accuracy of identifying the positions of the acupoints on the side of the head.

[0029] Furthermore, the embodiments of the present invention fit a plurality of identified acupoint points and a plurality of support points to obtain a fitting curve, and correct the acupoint points that need to be corrected based on the fitting curve, thereby further improving the accuracy of the identified target acupoint points, that is, further improving the accuracy of identifying the positions of the acupoints on the head.

[0030] In addition, in order to improve the robustness of the target acupoint identification model, that is, to realize accurately identifying the side head images of the identification target obtained at any angle and any shooting distance, in some embodiments of the present invention, the acupoint sample set on the side of the head includes a plurality of side head sample images of different people, a plurality of side head sample images of the same person at the same distance at different angles, and a plurality of side head sample images of the same person at the same angle at different distances.

[0031] The embodiments of the present invention include a plurality of side head sample images of different people, different angles of the same person, and different distances at the same angle in the acupoint sample set on the side of the head, so that the acupoint identification model learns a plurality of samples, and the obtained target acupoint identification model can accurately identify the side head images of the identification target at any angle and any distance, thereby improving the identification accuracy of the target acupoint identification model.

[0032] Furthermore, the head side acupoint sample set can include head side sample images under different hairstyles, head shapes, skin colors, and lighting conditions. This further improves the robustness and discrimination accuracy of the target acupoint discrimination model.

[0033] The fitting curve needs to be fitted based on the acupoint points and support points. To improve the accuracy of the fitting curve and the positioning accuracy of the head side acupoints, it is necessary to ensure accurate annotation of the acupoint sample points and support sample points in the head side acupoint sample set. To achieve this purpose, in some embodiments of the present invention, as shown in FIG. 2, before step S101, S201, acquiring a plurality of head side sample images, and identifying the Gallbladder Meridian of Foot-Shaoyang curve in each of the head side sample images; S202, annotating acupoint points based on the Gallbladder Meridian of Foot-Shaoyang curve to obtain a plurality of acupoint sample points; S203, dividing the Gallbladder Meridian of Foot-Shaoyang curve based on the inflection points of the Gallbladder Meridian of Foot-Shaoyang curve to obtain a plurality of segments of the Gallbladder Meridian of Foot-Shaoyang sub-curves; S204, determining at least one support sample point inserted into the Gallbladder Meridian of Foot-Shaoyang sub-curve based on the number of acupoint sample points and curve features in each segment of the Gallbladder Meridian of Foot-Shaoyang sub-curve.

[0034] Here, the specific method of the acupoint point annotation method in step S202 is to annotate a plurality of acupoint sample points on the Gallbladder Meridian of Foot-Shaoyang curve using the Labelme annotation tool.

[0035] In a specific embodiment of the present invention, as shown in FIG. 3, the Gallbladder Meridian of Foot-Shaoyang curve includes 7 inflection points, each of which is circular and is the points numbered 1, 2, 4, 7, 12, 14, and 20. Correspondingly, two adjacent inflection points form one section of the Gallbladder Meridian of Foot-Shaoyang sub-curve, and there are 6 sections of the Gallbladder Meridian of Foot-Shaoyang sub-curve.

[0036] Embodiments of the present invention improve the accuracy of acupoints on the lateral side of the head identified by a trained target acupoint identification model by marking acupoint points based on the limitation of the curve of the Gallbladder Meridian of Foot-Shaoyang. Furthermore, embodiments of the present invention ensure the rationality of the inserted support sample points by determining the number of inserted support sample points based on the number of acupoint sample points and curve characteristics, prevent excessive or insufficient insertion of support sample points, and improve the efficiency of acupoint identification while ensuring the accuracy of acupoint identification.

[0037] In some embodiments of the present invention, the curve features include the curvature change rate and the curve length. As shown in FIG. 4, step S204 includes S401. When the curvature change rate of the sub-curve of the Gallbladder Meridian of Foot-Shaoyang is less than or equal to the first threshold, the total number of support sample points and acupoint sample points on the sub-curve of the Gallbladder Meridian of Foot-Shaoyang is the number of the first threshold. Based on the number of the first threshold and the number of acupoint sample points, determining the target number of support sample points to be inserted into the sub-curve of the Gallbladder Meridian of Foot-Shaoyang; S402. When the curvature change rate of the sub-curve of the Gallbladder Meridian of Foot-Shaoyang is greater than the first threshold and the curve length of the sub-curve of the Gallbladder Meridian of Foot-Shaoyang is less than or equal to the length threshold, the total number of support sample points and acupoint sample points on the sub-curve of the Gallbladder Meridian of Foot-Shaoyang is the number of the second threshold. Based on the number of the second threshold and the number of acupoint sample points, determining the target number of support sample points to be inserted into the sub-curve of the Gallbladder Meridian of Foot-Shaoyang; S403. When the curvature change rate of the sub-curve of the Gallbladder Meridian of Foot-Shaoyang is greater than the first threshold and the curve length of the sub-curve of the Gallbladder Meridian of Foot-Shaoyang is greater than the length threshold, the total number of support sample points and acupoint sample points on the sub-curve of the Gallbladder Meridian of Foot-Shaoyang is the number of the third threshold. Based on the number of the third threshold and the number of acupoint sample points, determining the target number of support sample points to be inserted into the sub-curve of the Gallbladder Meridian of Foot-Shaoyang; S404. Inserting the target number of support sample points into the sub-curve of the Gallbladder Meridian of Foot-Shaoyang. Here, the number of the first threshold is smaller than the number of the second threshold, and the number of the second threshold is smaller than the number of the third threshold.

[0038] In a specific embodiment of the present invention, the number of the first threshold is 4, the number of the second threshold is 7, and the number of the third threshold is 9.

[0039] It should be understood that the first threshold value, the length threshold value, the number of first threshold values, the number of second threshold values, and the number of third threshold values can be adjusted according to the actual application scenario and will not be elaborated here.

[0040] In a specific embodiment of the present invention, as shown in FIG. 3, the circular points represent the acupoint sample points, and the head side sample image includes a total of 20 acupoint sample points numbered 1-20 respectively.

[0041] Here, when the curvature change rate of the curve between the acupoint sample point A and the acupoint sample point B is small and smaller than the first threshold value, the total number of support sample points and acupoint sample points in the curve segment AB is 4, and it is necessary to insert 2 support sample points between the acupoint sample point A and the acupoint sample point B.

[0042] Here, when the curvature change rate of the curve between the acupoint sample point C and the acupoint sample point D is large and larger than the first threshold value, and the curve length between the acupoint sample point C and the acupoint sample point D is smaller than the length threshold value, the total number of support sample points and acupoint sample points in the curve segment CD is 7, and it is necessary to insert 4 support sample points between the acupoint sample point C and the acupoint sample point D.

[0043] Here, when the curvature change rate of the curve between the acupoint sample point D and the acupoint sample point E is large and larger than the first threshold value, and the curve length between the acupoint sample point D and the acupoint sample point E is larger than the length threshold value, the total number of support sample points and acupoint sample points in the curve segment DE is 9, and it is necessary to insert 2 support sample points between the acupoint sample point D and the acupoint sample point E.

[0044] In addition, when it is necessary to insert a plurality of support sample points, it is required that the plurality of support sample points and the plurality of acupoint sample points be arranged as evenly as possible.

[0045] As shown in FIG. 3, the rectangular points represent support sample points, and the head side sample image includes a total of 20 through-hole sample points (numbers 1-20) and 9 support sample points (numbers 1-9).

[0046] In some embodiments of the present invention, step S103 fits through-hole points and support points based on the six curve segments divided in step S102, that is, six fitting curves are obtained by fitting.

[0047] However, as shown in FIG. 3, the curvature change of the curve formed by the through-hole points numbered 7-12 is large as a whole and cannot be represented by a simple expression. If this curve segment is fitted as a whole, both the calculation amount and accuracy of fitting are low. To avoid these problems, in a specific embodiment of the present invention, step S103 determines a plurality of curve segments based on a plurality of through-hole points and a plurality of support points. The curve segment is a relaxation curve segment or a circular curve segment, and each curve segment includes at least one through-hole point and at least one support point.

[0048] The embodiment of the present invention divides and fits the curve segments in two forms: relaxation curve segments and circular curve segments, that is, by expressing each curve segment with one type of curve formula, the fitting efficiency is improved while ensuring the fitting accuracy.

[0049] Specifically, the fitting curve includes 11 curve segments, and the through-hole points and support points included in each curve segment are as shown in Table 1.

[0050] Curve segment and the through-hole points and support points included therein

Table 1

[0051] As shown in Table 1, in the embodiment of the present invention, the curve segment is divided into 11 segments, and fitting is performed in order for each segment.

[0052] The acupoint points identified based on the target acupoint identification model may not exactly match the fitting curve obtained by fitting. However, the fitting curve represents the meridian of the Foot-Shaoyang Gallbladder Meridian. In order to ensure the accuracy of the location of the acupoint points, in the specific embodiment of the present invention, as shown in FIG. 5, in the process of determining whether it is necessary to correct the acupoint points based on the fitting curve in step S104, S501, a step of determining whether the distance between each acupoint point and the fitting curve is greater than zero and less than the distance threshold; S502, a step of determining that there is no need to correct the acupoint point when the distance between each acupoint point and the fitting curve is zero; S503, a step of removing the acupoint point when the distance between each acupoint point and the fitting curve is greater than or equal to the distance threshold; S504, including a step of determining that it is necessary to correct the acupoint point when the distance between each acupoint point and the fitting curve is greater than zero and less than the distance threshold.

[0053] In the embodiment of the present invention, by setting to remove the acupoint point when the distance between the acupoint point and the fitting curve is greater than or equal to the distance threshold, it is possible to avoid misidentifying the acupoint point with identification error as the target acupoint point, and further ensure the accuracy of the target acupoint point.

[0054] In the specific embodiment of the present invention, as shown in FIG. 6, correcting the acupoint point based on the fitting curve of step S104 to obtain the target acupoint point includes S601, a step of obtaining the tangent line of the fitting curve and setting the direction perpendicular to the tangent line and facing the direction of the fitting curve as the moving direction of the acupoint point; S602, including a step of using the distance between the acupoint point and the fitting curve as the moving distance and controlling the acupoint point to move along the moving direction by the moving distance to obtain the target acupoint point.

[0055] In a specific embodiment of the present invention, as shown in FIG. 7, when the identified acupuncture point No. 8 does not coincide with the fitting curve segment and does not exceed the correction range, it is necessary to move this acupuncture point onto the curve segment after fitting.

[0056] Here, the unit vector u in the moving direction is as follows. JPEG0007702763000003.jpg16170

[0057] Here, the point (x0, f(x0)) is the point on the curve closest to the acupuncture point No. 8, f'(x0) is the first derivative of this point, and the equation of the tangent line at this point is as follows. JPEG0007702763000004.jpg16170

[0058] Then the moving distance d is as follows. JPEG0007702763000005.jpg31170

[0059] In the formula, (x8, y8) is the acupuncture point No. 8.

[0060] In summary, the method for identifying the positions of acupoints on the lateral side of the head provided by the embodiments of the present invention uses deep learning, that is, obtains acupoint points using a target acupoint identification model, which does not require manual labor, enables rapid response, and can obtain acupoint points in real time and dynamically to support the diagnosis of doctors. In addition, by setting the annotation and correction of acupoint points based on the curve of the Gallbladder Meridian of Foot-Shaoyang, the identification of acupoints on the head is restricted by the meridian curve features pre-annotated, and the accuracy of acupoint position identification is improved. Furthermore, the embodiments of the present invention do not use expensive imaging diagnostic devices and have low costs. Furthermore, the embodiments of the present invention are configured to include, in the sample set of acupoints on the lateral side of the head, a plurality of lateral head sample images of different persons, a plurality of lateral head sample images taken from the same distance at different angles of the same person, and a plurality of lateral head sample images taken from the same angle and the same distance of the same person, so as to be able to cope with variations such as different head shapes, postures, shooting distances, lighting conditions, etc., and further improve the accuracy of the recognized acupoints.

[0061] To better implement the method for identifying the positions of acupoints on the lateral side of the head according to the embodiments of the present invention, based on the method for identifying the positions of acupoints on the lateral side of the head, correspondingly, the embodiments of the present invention further provide a device for identifying the positions of acupoints on the lateral side of the head. As shown in FIG. 8, the device 800 for identifying the positions of acupoints on the lateral side of the head includes a model training unit 801 that trains an acupoint identification model based on a sample set of acupoints on the lateral side of the head to obtain a target acupoint identification model, and the sample set of acupoints on the lateral side of the head includes a plurality of acupoint sample points and a plurality of support sample points located on the curve of the Gallbladder Meridian of Foot-Shaoyang an identification target image identification unit 802 that inputs a lateral head image of an identification target into the target acupoint identification model to obtain a plurality of acupoint points and a plurality of support points a curve fitting unit 803 that fits a plurality of acupoint points and a plurality of support points to obtain a fitting curve and an acupoint position identification unit 804 that determines whether it is necessary to correct the acupoint points based on the fitting curve, and if correction is necessary, corrects the acupoint points based on the fitting curve to obtain target acupoint points.

[0062] The head side acupoint position identification device 800 provided by the above embodiment can implement the technical solutions described in the above embodiment of the head side acupoint position identification method. For the specific implementation principles of the above modules or units, refer to the corresponding content in the above embodiment of the head side acupoint position identification method, and will not be described in detail here.

[0063] As shown in FIG. 9, the present invention also provides an acupuncture robot 900 accordingly. The acupuncture robot 900 includes a processor 901, a memory 902, and a display 903. Although FIG. 9 shows only some components of the acupuncture robot 900, it should be understood that not all of the shown components need to be implemented, and more components or fewer components may be alternatively implemented.

[0064] In some embodiments, the processor 901 may be a central processing unit (CPU), a microprocessor, or other data processing chips, and is used for executing the program code stored in the memory 902 and processing data such as the method for identifying the positions of acupoints on the head side in the present invention.

[0065] In some embodiments of the present invention, the processor 901 may be a single server or a server group. The server group may be centralized or distributed. In some embodiments, the processor 901 may be local or remote. In some embodiments, the processor 901 may be implemented on a cloud platform. In some embodiments, the cloud platform may include a private cloud, a public cloud, a hybrid cloud, a community cloud, a distributed cloud, an on-premises cloud, a multi-cloud, etc., or any combination thereof.

[0066] In some embodiments, the memory 902 may be an internal storage unit of the acupuncture robot 900, such as a hard disk or memory of the acupuncture robot 900. In other embodiments, the memory 902 may also be an external storage device of the acupuncture robot 900, such as a plug-in hard drive, a Smart Media Card (SMC), a Secure Digital (SD) card, a Flash Card, etc.

[0067] Furthermore, the memory 902 can also include both the internal storage unit and the external memory device of the acupuncture robot 900. The memory 902 is used to store the application software installed in the acupuncture robot 900 and various data.

[0068] The display 903 may be, in some embodiments, an LED display, a liquid crystal display, a touch liquid crystal display, an OLED (Organic Light-Emitting Diode) touch display, etc. The display 903 is used to display the information of the acupuncture robot 900 and the visualized user interface. The components 901 - 903 of the acupuncture robot 900 communicate with each other through the system bus.

[0069] In some embodiments of the present invention, when the processor 901 executes the acupoint location identification program on the head side in the memory 902, the following steps can be realized: Train an acupoint identification model based on the head side acupoint sample set to obtain a target acupoint identification model. The head side acupoint sample set includes a plurality of acupoint sample points and a plurality of support sample points located on the curve of the Gallbladder Meridian of Foot-Shaoyang. Input the head side image to be identified into the target acupoint identification model to obtain a plurality of acupoint points and a plurality of support points. Fit the plurality of acupoint points and the plurality of support points to obtain a fitting curve. Based on the fitting curve, it is determined whether it is necessary to correct the acupuncture point. If correction is necessary, the acupuncture point is corrected based on the fitting curve to obtain the target acupuncture point.

[0070] It should be understood that when the processor 901 executes the head side acupuncture point position identification program in the memory 902, other functions can also be realized in addition to the above-mentioned functions. For details, please refer to the description of the corresponding method embodiments above.

[0071] Furthermore, in the embodiments of the present invention, the type of the mentioned acupuncture robot 900 is not particularly limited. The acupuncture robot 900 may be a portable acupuncture robot such as a mobile phone, a tablet, a personal digital assistant (PDA), a wearable device, or a laptop. Exemplary embodiments of the portable acupuncture robot include, but are not limited to, portable acupuncture robots equipped with IOS, Android, Microsoft, or other operating systems. The above-mentioned portable acupuncture robots may be other portable acupuncture robots. Also, in some other embodiments of the present invention, it should be understood that the acupuncture robot 900 may also be a desktop computer having a touch-sensitive surface (for example, a touch panel) instead of a portable acupuncture robot.

[0072] Accordingly, the embodiments of the present invention further provide a computer-readable storage medium, which is used to store a program or instruction readable by a computer. When the program or instruction is executed by a processor, the steps or functions of the head side acupuncture point position identification method in any of the above embodiments can be realized.

[0073] All or part of the process for implementing the method of the above embodiments can be completed by instructing related hardware (such as a processor, a controller, etc.) through a computer program, and the computer program can be stored in a computer-readable storage medium. Among them, those skilled in the art can understand that the computer-readable storage medium includes a disk, an optical disk, a read-only memory, a random access memory, and the like.

[0074] In the above, the method for identifying the position of the head side through-hole, the device, the acupuncture robot, and the storage medium provided by the present invention are introduced in detail. In this specification, specific examples are used to explain the principle and implementation method of the present invention. The description of the above embodiments is only used to understand the method and the central idea of the present invention. At the same time, for those skilled in the art, based on the idea of the present invention, changes will occur in the specific implementation and application scope. In summary, the content of this specification should not be construed as limiting the present invention.

Claims

1. A device for identifying the positions of acupoints on the lateral side of the head, comprising: a model training unit that trains an acupoint identification model based on a set of acupoint samples on the lateral side of the head to obtain a target acupoint identification model, wherein the set of acupoint samples on the lateral side of the head includes a plurality of acupoint sample points and a plurality of support sample points located on the curve of the Gallbladder Meridian of Foot-Shaoyang; an identification target image identification unit that inputs an image of the lateral side of the head to be identified into the target acupoint identification model to obtain a plurality of acupoint points and a plurality of support points; a curve fitting unit that fits the plurality of acupoint points and the plurality of support points to obtain a fitting curve; an acupoint position identification unit that determines whether it is necessary to correct the acupoint points based on the fitting curve, and if correction is necessary, corrects the acupoint points based on the fitting curve to obtain target acupoint points; Before training an acupoint identification model based on the set of acupoint samples on the lateral side of the head, acquiring a plurality of lateral head sample images, and identifying the curve of the Gallbladder Meridian of Foot-Shaoyang in each of the lateral head sample images; labeling acupoint points based on the curve of the Gallbladder Meridian of Foot-Shaoyang to obtain a plurality of acupoint sample points; dividing the curve of the Gallbladder Meridian of Foot-Shaoyang based on the inflection points of the curve of the Gallbladder Meridian of Foot-Shaoyang to obtain a plurality of segments of the sub-curve of the Gallbladder Meridian of Foot-Shaoyang; determining at least one support sample point to be inserted into the sub-curve of the Gallbladder Meridian of Foot-Shaoyang based on the number of acupoint sample points and the curve characteristics in the sub-curve of the Gallbladder Meridian of Foot-Shaoyang for each segment; further comprising: wherein the curve characteristics include the curvature change rate and the curve length, and determining at least one support sample point to be inserted into the sub-curve of the Gallbladder Meridian of Foot-Shaoyang based on the number of acupoint sample points in the sub-curve of the Gallbladder Meridian of Foot-Shaoyang for each segment and the curve characteristics in the sub-curve of the Gallbladder Meridian of Foot-Shaoyang for each segment means that when the curvature change rate of the sub-curve of the Gallbladder Meridian of Foot-Shaoyang is less than or equal to a first threshold value, the total number of support sample points and acupoint sample points on the sub-curve of the Gallbladder Meridian of Foot-Shaoyang is a first threshold number, and based on the first threshold number and the number of acupoint sample points, determining the target number of support sample points to be inserted into the sub-curve of the Gallbladder Meridian of Foot-Shaoyang; When the curvature change rate of the sub - curve of the Gallbladder Meridian of Foot - Shaoyang is greater than the first threshold value and the curve length of the sub - curve of the Gallbladder Meridian of Foot - Shaoyang is less than or equal to the length threshold value, the total number of support sample points and acupoint sample points on the sub - curve of the Gallbladder Meridian of Foot - Shaoyang is the second threshold number. Based on the second threshold number and the number of acupoint sample points, determining the target number of support sample points to be inserted into the sub - curve of the Gallbladder Meridian of Foot - Shaoyang; When the curvature change rate of the sub - curve of the Gallbladder Meridian of Foot - Shaoyang is greater than the first threshold value and the curve length of the sub - curve of the Gallbladder Meridian of Foot - Shaoyang is greater than the length threshold value, the total number of support sample points and acupoint sample points on the sub - curve of the Gallbladder Meridian of Foot - Shaoyang is the third threshold number. Based on the third threshold number and the number of acupoint sample points, determining the target number of support sample points to be inserted into the sub - curve of the Gallbladder Meridian of Foot - Shaoyang; Inserting the support sample points of the target number into the sub - curve of the Gallbladder Meridian of Foot - Shaoyang; including; A device for identifying the positions of acupoints on the lateral side of the head, characterized in that the first threshold number is smaller than the second threshold number, and the second threshold number is smaller than the third threshold number.

2. The curve fitting unit is identifying a plurality of curve segments based on the plurality of acupoint points and the plurality of support points, where the curve segments are easement curve segments or circular curve segments, and each of the curve segments includes at least one of the acupoint points and at least one of the support points. The device for identifying the positions of acupoints on the lateral side of the head according to claim 1, characterized by including this.

3. Identifying whether it is necessary to correct the acupoint points based on the fitting curve is judging whether the distance between each of the acupoint points and the fitting curve is greater than zero and less than the distance threshold; when the distance between each of the acupoint points and the fitting curve is zero, judging that there is no need to correct the acupoint point; when the distance between each of the acupoint points and the fitting curve is greater than or equal to the distance threshold, removing the acupoint point; when the distance between each of the acupoint points and the fitting curve is greater than zero and less than the distance threshold, judging that it is necessary to correct the acupoint point; The device for identifying the positions of acupoints on the lateral side of the head according to claim 1, characterized by including this.

4. Modifying the acupoint points based on the fitting curve to obtain the target acupoint points is Obtain the tangent line of the fitting curve, and set the direction perpendicular to the tangent line and in the direction of the fitting curve as the moving direction of the acupoint point. Set the distance between the acupoint point and the fitting curve as the moving distance, and control the acupoint point to move only the moving distance along the moving direction to obtain the target acupoint point. The head side acupoint position identification device according to claim 1, characterized by including the above.

5. The head side acupoint sample set includes a plurality of head side sample images of different persons, a plurality of head side sample images at the same distance at different angles of the same person, and a plurality of head side sample images at different distances at the same angle of the same person. The head side acupoint position identification device according to claim 1, characterized by including the above.

6. Including a memory and a processor. The memory stores a program. The processor is coupled to the memory and executes the program stored in the memory to realize the head side acupoint position identification device according to any one of claims 1 to 5. A acupuncture robot characterized by the above.

7. A computer-readable storage medium, characterized in that a program and instructions are stored, and the head side acupoint position identification device according to any one of claims 1 to 5 is realized by the program and instructions being executed by a processor.

Images