Method for stably recognizing acupuncture points on back of human body and intelligent terminal

Abstract

The invention discloses a method for stably recognizing human body back acupuncture points and an intelligent terminal.The method for stably recognizing the human body back acupuncture points comprises the steps that a back image of a current user is obtained, positioning coordinates of reference acupuncture points are obtained, and the human body proportion is calculated according to the reference acupuncture points; dividing the reference acupuncture points into a plurality of groups, determining a reference straight line according to the representative points in each group, and then obtaining a spine trend identification line; and calculating the specific positions of other acupuncture points according to the human body proportion and the spine trend identification line. According to the method, the reference straight line of a plurality of groups is determined by obtaining the user back image, marking the reference acupuncture point coordinates and calculating the representative points of the groups divided by the reference acupuncture points in groups, then the spine trend identification line is generated, and the positions of other acupuncture points are determined by combining the human body proportion and the spine trend identification line. According to the method, the spine trend and the human body proportion are comprehensively considered, the acupuncture points are accurately marked, the recognition efficiency is improved, and the accuracy of a positioning result and the physiology rationality are also ensured.

Description

Technical Field

[0001] The present invention relates to the technical field of robot vision recognition, and particularly relates to a method for stably identifying acupoints on the human back and an intelligent terminal. Background Art

[0002] With the continuous development of traditional Chinese medicine diagnosis and treatment technologies, the accuracy and efficiency of acupoint location have received increasing attention in the medical field. Especially in traditional treatment methods such as acupuncture, massage, and moxibustion, the accurate location of back acupoints is directly related to the treatment effect and patient experience. As an important treatment area of the human body, the back contains up to 95 key acupoint points, which are distributed on important meridians such as the Governor Vessel and the Bladder Meridian. These acupoints are not only closely related to the functions of internal organs but also play an important role in disease prevention and treatment. Accurate acupoint location can not only improve the treatment effect but also reduce medical risks such as incorrect acupuncture and moxibustion. At the same time, it is of great significance for traditional Chinese medicine teaching, clinical research, and standardization construction.

[0003] In the prior art, the location of back acupoints mainly relies on physicians to measure through the method of bone degrees and dimensions. This method has problems such as strong subjectivity, low operation efficiency, and lack of standardization. When multiple acupoints need to be located, there are differences in the location results among different physicians. Although there are also methods for identifying acupoints through mechanical vision in the prior art, such as the meridian acupoint massage method based on deep learning disclosed in CN115429683A, the existing mechanical recognition cannot target the body shape conditions of different users to locate acupoints specifically, resulting in inaccurate acupoint location.

[0004] Therefore, the prior art still needs to be improved and developed. Summary of the Invention

[0005] In order to solve the problems in the prior art that when locating multiple acupoints on the user's back, the prior art relies on subjective measurement by physicians, resulting in low efficiency, lack of standardization, and large differences in location results among different physicians, and that there are problems with inaccurate acupoint location when using mechanical vision to identify acupoints because it cannot target the body shape conditions of different users to locate acupoints specifically, the present invention proposes a method for stably identifying acupoints on the human back and an intelligent terminal.

[0006] The present invention is achieved through the following technical solutions:

[0007] A method for stably identifying acupoints on the human back, wherein the method for stably identifying acupoints on the human back includes:

[0008] Obtain the back image of the current user, obtain the positioning coordinates of the reference acupoints based on the back image, and calculate the body proportion of the current user based on the positioning coordinates of the reference acupoints;

[0009] Divide the reference acupoints into several groups that include the left and right symmetric reference acupoints, calculate the representative points on the left and right sides in each group according to the positioning coordinates; determine the reference line according to the representative points;

[0010] Obtain the spinal alignment identification line of the current user according to the reference line;

[0011] Calculate the specific positions of other acupoints according to the human body ratio and the spinal alignment identification line.

[0012] The method for stably identifying acupoints on the human back, wherein the calculating the representative points on the left and right sides in each group according to the positioning coordinates includes:

[0013] Calculate the representative points P l (x l ,y l ) and P r (x r ,y r ) in several of the groups respectively;

[0014] The representative point P l (x l ,y l ) is the mean point of several acupoints on the left side in the group;

[0015] The representative point P r (x r ,y r ) is the mean point of several acupoints on the right side in the group.

[0016] The method for stably identifying acupoints on the human back, wherein the determining the reference line according to the representative points includes:

[0017] Calculate the slope k and intercept b of the reference line L1 of the corresponding group according to the representative points P l (x l ,y l ) and P r (x r ,y r ) in several of the groups, and determine the reference line L1 according to the slope k and intercept b;

[0018] The calculation formula for the slope k is: k = (y r -y l ) / (x r -x l );

[0019] The calculation formula for the intercept b is: b = y l -k*x l 。

[0020] The method for stably identifying acupoints on the human back, wherein obtaining the spinal alignment identification line of the current user based on the reference line includes:

[0021] Calculating the slope k2 of the spinal alignment according to a plurality of the reference lines L1, and generating the spinal alignment identification line at the center point of the reference line L1 according to the slope k2;

[0022] The calculation formula for the slope k2 of the spinal alignment identification line is: k2 = -1 / k.

[0023] The method for stably identifying acupoints on the human back, wherein calculating the specific positions of other acupoints according to the human body ratio and the spinal alignment identification line includes:

[0024] Calculating the basic scale d, and determining the specific positions p of other acupoints according to the basic scale d new ;

[0025] The basic scale d is the distance between the highest point p max and the lowest point p min in the group;

[0026] The calculation formula for the basic scale d is:

[0027] The calculation formula for the specific position p new of other acupoints is:

[0028] x new = x min + t * d * cos(arctan(k2));

[0029] y new = y min + t * d * sin(arctan(k2)).

[0030] The method for stably identifying acupoints on the human back, wherein obtaining the back image of the current user and obtaining the positioning coordinates of the reference acupoints according to the back image includes:

[0031] Processing the back image of the current user using a training model;

[0032] Before processing the back image of the current user using the training model, it further includes:

[0033] Constructing a training model, the training model includes a human back acupoint data set, and the human back acupoint data set includes a plurality of human back images with coordinate data of the reference acupoints.

[0034] The method for stably identifying human back acupoints, wherein the human back acupoint dataset further includes enhanced data;

[0035] The enhanced data is an image obtained by randomly rotating, scaling, and performing brightness processing on a plurality of pairs of the human back images.

[0036] The method for stably identifying human back acupoints, wherein the dividing the reference acupoints into several groups including the left and right symmetric reference acupoints includes:

[0037] Grouping according to the region where the reference acupoints are located in the back image;

[0038] The groups include a shoulder region group, a middle back region group, and a waist region group;

[0039] The shoulder region group includes: 8 acupoints that are symmetric left and right, namely Jianzhongshu, Jianwaishu, Bingfeng, and Tianzong;

[0040] The middle back region group includes: 6 acupoints that are symmetric left and right, namely Geshu, Hunmen, and Sanjiaoshu;

[0041] The waist region group includes: 12 acupoints that are symmetric left and right, namely Guanyuanshu, Pangguangshu, Baihuanshu, Zhibian, Huiyang, and Jingmen.

[0042] The method for stably identifying human back acupoints, wherein the dividing the reference acupoints into several groups including the left and right symmetric reference acupoints further includes:

[0043] Or, grouping according to the meridian connection relationship of the reference acupoints in the back image;

[0044] Or, grouping according to the functional attributes of the reference acupoints in the back image;

[0045] Or, grouping according to the recognition confidence of the reference acupoints in the back image.

[0046] An intelligent terminal, wherein the intelligent terminal includes a memory, a processor, and a program for stably identifying human back acupoints stored in the memory and executable on the processor. When the processor executes the program for stably identifying human back acupoints, the method for stably identifying human back acupoints as described above is implemented;

[0047] The intelligent terminal further includes a display, and the display can simultaneously display at least two display pages. One display page is used to display the back image of the current user, and the other display page is used to display the back image of the current user with the reference acupoints and other acupoints marked.

[0048] The beneficial effects of the present invention are as follows: By obtaining the back image of the current user, accurately marking the positioning coordinates of the reference acupoints, dividing these reference acupoints into several groups, calculating the representative points of each group, determining the reference line based on the representative points, and then generating the identification line of the spinal column direction, and combining the spinal column direction identification line with the human body ratio, the present invention can accurately determine the specific positions of human acupoints. Compared with the prior art, by comprehensively considering the spinal column direction and the human body ratio, the present invention realizes the accurate marking of acupoints. This not only significantly increases the number of recognizable acupoints and improves the recognition efficiency, but also ensures the accuracy and physiological rationality of the positioning results. Brief Description of the Drawings

[0049] Figure 1 is the flowchart of the method for stably identifying human back acupoints of the present invention;

[0050] Figure 2 is the schematic diagram of the back image of the current user in the method for stably identifying human back acupoints of the present invention;

[0051] Figure 3 is the schematic diagram of the back image after the reference acupoints are marked in the method for stably identifying human back acupoints of the present invention;

[0052] Figure 4 is the schematic diagram of the back image after the qi acupoints are marked in the method for stably identifying human back acupoints of the present invention;

[0053] Figure 5 is the schematic diagram of the display page of the display in the intelligent terminal of the present invention;

[0054] Figure 6 is the internal structure principle block diagram of the intelligent terminal of the present invention. Detailed Embodiment

[0055] To make the objectives, technical solutions and effects of the present invention clearer and more definite, the following further describes the present invention in detail with reference to the accompanying drawings and by way of examples. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0056] It should be noted that if there are directional indications (such as up, down, left, right, front, back...) involved in the embodiments of the present invention, the directional indications are only used to explain the relative positional relationship and movement conditions between components in a specific posture (as shown in the drawings). If the specific posture changes, the directional indications will also change accordingly.

[0057] In addition, if there are descriptions such as "first", "second", etc. in the embodiments of the present invention, the descriptions of "first", "second", etc. are only for descriptive purposes and should not be construed as indicating or implying their relative importance or implicitly specifying the quantity of the indicated technical features. Thus, the features defined with "first" and "second" may explicitly or implicitly include at least one of the said features. In addition, the technical solutions between various embodiments can be combined with each other, but it must be based on the ability of those of ordinary skill in the art to implement. When the combination of technical solutions is contradictory or cannot be implemented, it should be considered that such a combination of technical solutions does not exist and is not within the scope of protection required by the present invention.

[0058] In the prior art, the positioning of back acupoints mainly relies on physicians to measure through the bone measurement method. This method has problems such as strong subjectivity, low operation efficiency, and lack of standardization. When multiple acupoints need to be located, there are differences in the positioning results among different physicians. Although there are also methods for identifying acupoints through machine vision in the prior art, such as the meridian acupoint massage method based on deep learning disclosed in CN115429683A, the existing mechanical recognition cannot target the body shape conditions of different users to accurately locate acupoints, resulting in inaccurate acupoint positioning.

[0059] Based on the above problems in the prior art, the present invention provides a method for stably identifying back acupoints of the human body, as Figure 1 shown, the method for stably identifying back acupoints of the human body includes:

[0060] S100. Obtain the back image of the current user, obtain the positioning coordinates of the reference acupoints according to the back image, and calculate the body proportion of the current user according to the positioning coordinates of the reference acupoints;

[0061] S200. Divide the reference acupoints into several groups including the left and right symmetric reference acupoints, calculate the representative points on the left and right sides of each group according to the positioning coordinates; determine the reference line according to the representative points;

[0062] S300. Obtain the spinal alignment identification line of the current user according to the reference line;

[0063] S400. Calculate the specific positions of other acupoints according to the body proportion and the spinal alignment identification line.

[0064] In this embodiment, the reference acupoints are human body acupoints located in easily locatable parts of the human back. After obtaining the back image of the user (as Figure 2 shown), the specific positions of the reference acupoints can be determined according to the characteristics of the user's back image, such as the characteristic distribution of the back bones, the characteristic distribution of the muscle trends, etc., as Figure 3As shown, in a specific embodiment, the reference acupoints include the left and right symmetrical points of Jianzhongshu, Jianwaishu, Bingfeng, Tianzong, Geshu, Hunmen, Sanjiaoshu, Jingmen, Guanyuanshu, Bladdershu, Baihuanshu, Zhibian, Huiyang and other acupoints. These acupoints are not only commonly used and safe massage acupoints in clinical practice, but also acupoints that are easy to distinguish in images. Therefore, the coordinates of the above-mentioned reference acupoints in the back image can be easily marked, that is, the positioning coordinates of the reference acupoints can be obtained.

[0065] At the same time, after the positioning coordinates of the benchmark acupoints are marked and determined, the body proportions of the current user can be calculated by processing several positioning coordinates. The body proportions include but are not limited to proportional relationships such as height, shoulder width, back width, and pelvic width. These proportional relationships provide key parameters for the subsequent calculation of the specific positions of other acupoints, thereby avoiding the subsequent marking of other acupoints being affected by the current user's body shape conditions such as fatness, thinness, height, etc.

[0066] Furthermore, if Figure 3 As shown, since the acupuncture points on the back of the human body are distributed symmetrically along the spine, the above-mentioned reference acupuncture points can be divided into several groups including bilaterally symmetrical reference acupuncture points, and the representative points on the left and right sides of each group are calculated according to the positioning coordinates marked above. The representative points are obtained by averaging multiple positioning coordinates in the group, and can represent the distribution of several reference acupuncture points in the group corresponding to the current user. At the same time, since two points determine a straight line, after calculating the representative points on the left and right sides of each group, the reference straight line of each group can be obtained by connecting the representative points on the left and right sides.

[0067] Furthermore, since the human body's acupuncture points are distributed symmetrically along the left and right sides of the spine, after obtaining the reference straight line, the reference straight line can be used to determine the spine direction identification line. Therefore, using the spine direction identification line as a basis, combined with the above-calculated body proportions of the current user and the above-calculated representative point coordinates, as well as traditional bone measurement theory, the calculation of other acupuncture points can be achieved to achieve the effect of calculating the specific positions of other acupuncture points.

[0068] The present invention obtains the back image of the current user, accurately marks the location coordinates of the reference acupuncture points, divides these reference acupuncture points into several groups, calculates the representative points of each group, determines the reference straight line based on the representative points, and then generates the identification line of the spine direction. By combining the identification line of the spine direction with the proportion of the human body, the present invention can accurately determine the specific location of the human body acupuncture points. Compared with the existing technology, the present invention realizes the accurate marking of acupuncture points (such as Figure 4 This not only significantly increases the number of identifiable acupoints and improves recognition efficiency, but also ensures the accuracy and physiological rationality of the positioning results.

[0069] In another feasible embodiment of the present invention, calculating the representative points on the left and right sides in each group according to the above-mentioned positioning coordinates includes:

[0070] S210. Calculate the representative points P l (x l ,y l ) and P r (x r ,y r ) in several of the above-mentioned groups respectively;

[0071] Among them, the representative point P l (x l ,y l ) is the mean point of several acupoints on the left side in the group; the representative point P r (x r ,y r ) is the mean point of several acupoints on the right side in the group.

[0072] In a specific embodiment, calculate the mean points of the acupoints on the left side (such as Zhongshu of the left shoulder, Waishu of the left shoulder, etc.) and the acupoints on the right side (such as Zhongshu of the right shoulder, Waishu of the right shoulder, etc.) in the above-mentioned group respectively, to obtain two representative points P l (x l ,y l ) and P r (x r ,y r ). By calculating the mean points of the acupoints on the left and right sides in each group, the distribution of the acupoints in this group can be reflected more accurately. When calculating the mean point, the positioning coordinates of each acupoint can be considered and weighted averaged to obtain more accurate representative point coordinates. This method not only improves the accuracy of acupoint positioning, but also helps to improve the accuracy of determining the reference line and the spinal alignment identification line based on the representative points in the subsequent process.

[0073] Based on the above-mentioned acquisition of the two representative points P l (x l ,y l ) and P r (x r ,y r ), the reference line corresponding to this group can be determined by the method of determining a straight line through two points. The above-mentioned determination of the reference line according to the representative points includes:

[0074] S220. According to the representative points P l (x l ,y l ) and P r (x r ,y r)Calculate the slope k and intercept b of the reference line L1 corresponding to the group, and determine the reference line L1 according to the slope k and intercept b;

[0075] Among them, the calculation formula for the slope k is: k = (y r - y l ) / (x r - x l ); The calculation formula for the intercept b is: b = y l - k * x l .

[0076] The reference line L1 can be drawn through the slope k and intercept b, that is, the reference lines L1 of all groups can be drawn. These reference lines not only reflect the acupoint distribution on the human back, but also provide an important basis for determining the spinal alignment marking line in the subsequent process.

[0077] Specifically, obtaining the spinal alignment marking line of the current user according to the reference line includes:

[0078] S310. Calculate the slope k2 of the spinal alignment according to several reference lines L1, and generate the spinal alignment marking line at the center point of the reference line L1 according to the slope k2;

[0079] Among them, according to the human anatomical characteristics, the spinal direction should be perpendicular to the reference line. The calculation formula for the slope k2 of the spinal direction is: k2 = -1 / k.

[0080] It should be noted that in the above embodiment, the reference lines L1 of each group are obtained. Therefore, the slope k2 of the spinal alignment marking line corresponding to a single group can be obtained through the above calculation formula. After obtaining the spinal alignment marking lines corresponding to all groups, the parts corresponding to each group of the spinal alignment marking lines can be processed to integrate several spinal alignment marking lines into one spinal alignment marking line. The integrated spinal alignment marking line can more accurately reflect the actual spinal alignment of the current user and provide a key reference for subsequent acupoint positioning.

[0081] Furthermore, the above-mentioned calculating the specific positions of other acupoints according to the human body ratio and the spinal alignment marking line includes:

[0082] [[ID=3!6]]S410. Calculate the basic scale d, and determine the specific position p of other acupoints according to the basic scale d new ;

[0083] Among them, the basic scale d is the distance between the highest point p max and the lowest point p min in the group;

[0084] The calculation formula for the basic scale d is: The basic scale d is a reference for determining the positions of other acupoints based on human proportions;

[0085] Based on the above steps, the specific position p of other acupoints is finally calculated new The calculation formula is:

[0086] x new = x min + t * d * cos(arctan(k2));

[0087] y new = y min + t * d * sin(arctan(k2)).

[0088] Among them, t is a proportionality coefficient determined according to the traditional bone measurement theory;

[0089] Among them, cos(arctan(k2)) is the projection distance of the x coordinate of other acupoints relative to the spinal alignment line; sin(arctan(k2)) is the projection distance of the y coordinate of other acupoints relative to the spinal alignment line;

[0090] t is the relative ratio between other acupoints and the representative point. This proportionality coefficient can be determined according to the bone measurement method in traditional Chinese medicine theory. By analyzing the proportional relationships of different human body parts, the relative position relationships between various acupoints are obtained. By combining human proportions, the spinal alignment line, and the traditional bone measurement theory, the present invention can achieve precise positioning of other acupoints on the human back. This method not only improves the accuracy and efficiency of acupoint positioning but also provides more reliable technical support for traditional Chinese medicine massage, acupuncture, and other therapies. In addition, the implementation manner of the present invention may further include preprocessing steps for the user's back image, such as denoising, enhancing contrast, etc., to improve the accuracy and stability of acupoint recognition. At the same time, the present invention can also continuously optimize the algorithm and model to adapt to the body shapes and back characteristics of different users, further improving the accuracy and applicability of acupoint recognition.

[0091] In another feasible implementation manner of the present invention, to achieve the effect of more accurate marking of the positioning coordinates of the reference acupoint when obtaining the positioning coordinates of the reference acupoint according to the back image in the above S100, in the present invention, obtaining the back image of the current user and obtaining the positioning coordinates of the reference acupoint according to the back image includes:

[0092] S110. Processing the back image of the current user using a trained model;

[0093] Before processing the back image of the current user using a trained model, it further includes:

[0094] S10. Construct a training model, where the training model includes a human back acupoint dataset, and the human back acupoint dataset includes a number of human back images with coordinate data of the reference acupoints.

[0095] The present invention first constructs a professional back acupoint dataset. By collecting more than 5000 human back images under different body types and different lighting conditions, and inviting professional Chinese medicine doctors to annotate these images. As Figure 3 shown, the annotated content includes 26 reference acupoint points, and these acupoints can be divided into a shoulder area (Jianzhongshu, Jianwaishu, Bingfeng, Tianzong, etc., 8 points in total for left and right symmetry), a middle back area (Geshu, Hunmen, Sanjiaoshu, etc., 6 points in total for left and right symmetry), and a waist area (Guanyuanshu, Pangguangshu, Baihuanshu, Zhibian, Huiyang, Jingmen, 12 points in total for left and right symmetry) according to the anatomical position. To ensure the annotation quality, multiple quality control mechanisms such as cross-validation by multiple experts, data cleaning, and outlier detection are also adopted.

[0096] In the training stage of the model, the present invention uses the YOLOv8 model to identify 26 reference acupoints. The training process first performs pre-training on the COCO dataset to obtain basic feature extraction capabilities, and then performs fine-tuning on the self-built back acupoint dataset.

[0097] Furthermore, the above-mentioned human back acupoint dataset also includes enhanced data, which are images obtained by randomly rotating, scaling, and adjusting the brightness of several pairs of the above-mentioned human back images.

[0098] The purpose of setting the above-mentioned enhanced data is to provide the generalization ability of the above-mentioned model training. In addition to using the above-mentioned enhanced data, by using Focal Loss as the loss function, the problem of sample imbalance is effectively solved, and through continuous optimization and verification, the recognition accuracy of the model on the test set is further improved.

[0099] Based on the above, using the training model to obtain the positioning coordinates of the reference acupoints of the user's back image, and calculating the specific positions of other acupoints by dividing the reference acupoints into groups to calculate the representative points, reference lines, and spinal alignment identification lines, the present invention can effectively improve the accuracy of identifying the reference acupoints and other acupoints.

[0100] There are various specific ways to divide the above-mentioned reference acupoints into several groups containing the left and right symmetric reference acupoints. In a specific embodiment, it can be grouped according to the regions where the reference acupoints are located in the back image, specifically as a shoulder area group, a middle back area group, and a waist area group; as Figure 3As shown in the figure, the shoulder area group includes 8 symmetric acupoints on the left and right, namely Jianzhongshu, Jianwaishu, Bingfeng, and Tianzong; the middle back area group includes 6 symmetric acupoints on the left and right, namely Geshu, Hunmen, and Sanjiaoshu; the waist area group includes 12 symmetric acupoints on the left and right, namely Guanyuanshu, Pangguangshu, Baihuanshu, Zhibian, Huiyang, and Jingmen.

[0101] In other feasible embodiments, the above division of the reference acupoints into several groups containing the symmetric reference acupoints on the left and right can also be grouped by other methods. For example:

[0102] It can be grouped according to the meridian connection relationship of the reference acupoints in the back image; according to traditional Chinese medicine meridian theory, there are specific meridian connection relationships between acupoints in the human body. Therefore, acupoints with the same or similar meridian connection relationships can be divided into the same group. Through this grouping method, the traditional Chinese medicine meridian theory can be explored and utilized more deeply, and the accuracy and scientific nature of acupoint positioning can be further improved.

[0103] It can also be grouped according to the functional attributes of the reference acupoints in the back image; acupoints in the human body not only have specific anatomical positions but also carry different functional attributes. For example, certain acupoints have significant effects on relieving shoulder and neck pain and improving blood circulation. Therefore, acupoints with the same or similar functional attributes can be divided into the same group. This grouping method can more comprehensively reflect the characteristics and functions of acupoints and provide more accurate guidance for subsequent acupoint positioning and treatment.

[0104] It can also be grouped according to the recognition confidence of the reference acupoints in the back image; that is, acupoints with higher recognition confidence are divided into one group, and acupoints with lower recognition confidence are divided into another group. Through this grouping method, targeted processing and analysis can be carried out on acupoints with different confidence levels, and the stability and reliability of acupoint recognition can be further improved.

[0105] When implementing the present invention, an appropriate grouping method can be selected according to specific requirements and scenarios to achieve the best acupoint recognition effect. The present invention makes full use of the position information of known acupoints and determines the key reference line through the overall distribution characteristics of acupoints within the group; secondly, by establishing the vertical relationship, the spinal column trend is accurately grasped, which is crucial for positioning other acupoints; finally, using the projection distance as the basic scale not only ensures the accuracy of calculation but also can adapt to the individual differences of different body types.

[0106] Based on the above embodiments, the present invention also provides an intelligent terminal, and its principle block diagram can be as Figure 6As shown in the figure. The smart terminal includes a processor, a memory, a network interface, and a display screen connected through a system bus. Among them, the processor of the smart terminal is used to provide computing and control capabilities. The memory of the smart terminal includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The network interface of the smart terminal is used to communicate with external terminals through a network connection. When the computer program is executed by the processor, it realizes a method for stably identifying acupoints on the human back.

[0107] Those skilled in the art can understand that Figure 6 the block diagram of the principle shown in the figure is only a block diagram of some structures related to the solution of the present invention, and does not constitute a limitation on the smart terminal to which the solution of the present invention is applied. The specific smart terminal may include more or fewer components than those shown in the figure, or combine some components, or have different component arrangements.

[0108] In one embodiment, a smart device is provided, including a memory, and one or more programs, where one or more programs are stored in the memory and are configured to be executed by one or more processors. The one or more programs include instructions for performing the following operations:

[0109] Obtain the back image of the current user, obtain the positioning coordinates of the reference acupoints according to the back image, and calculate the human body ratio of the current user according to the positioning coordinates of the reference acupoints;

[0110] Divide the reference acupoints into several groups including the left and right symmetric reference acupoints, calculate the representative points on the left and right sides of each group according to the positioning coordinates; determine the reference line according to the representative points;

[0111] Obtain the spinal cord orientation identification line of the current user according to the reference line;

[0112] Calculate the specific positions of other acupoints according to the human body ratio and the spinal cord orientation identification line.

[0113] In addition, as Figure 5 shown, the smart terminal further includes a display, and at least two display pages can be simultaneously displayed on the display. Among them, one display page is used to display the back image of the current user, and the other display page is used to display the back image of the current user with the reference acupoints and other acupoints marked. Through the dual-window side-by-side display design of the present invention, the original image is displayed on the left, and the recognition result is displayed in real time on the right. The operation is simple and intuitive, fully meeting the clinical use requirements, and providing strong technical support for traditional Chinese medicine clinical practice.

[0114] Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be completed by instructing relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above methods. Among them, any reference to a memory, storage, database, or other medium used in the embodiments provided by the present invention can include non-volatile and / or volatile memories. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), memory bus (Rambus), direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0115] In summary, the present invention provides a method for stably identifying acupoints on the human back. The method for stably identifying acupoints on the human back includes: obtaining a back image of the current user, obtaining the positioning coordinates of reference acupoints based on the back image, and calculating the human body ratio of the current user according to the positioning coordinates of the reference acupoints; dividing the reference acupoints into several groups including the left and right symmetric reference acupoints, and calculating the representative points on the left and right sides of each group according to the positioning coordinates; determining a reference line according to the representative points; obtaining an identification line of the spinal column trend of the current user according to the reference line; and calculating the specific positions of other acupoints according to the human body ratio and the identification line of the spinal column trend. By obtaining the back image of the current user, accurately marking the positioning coordinates of the reference acupoints, dividing these reference acupoints into several groups, calculating the representative points of each group, determining the reference line based on the representative points, and then generating an identification line of the spinal column trend, and combining the identification line of the spinal column trend with the human body ratio, the present invention can accurately determine the specific positions of human acupoints. Compared with the prior art, by comprehensively considering the spinal column trend and the human body ratio, the present invention realizes the accurate marking of acupoints. This not only significantly increases the number of identifiable acupoints, improves the identification efficiency, but also ensures the accuracy and physiological rationality of the positioning results.

[0116] It should be understood that the application of the present invention is not limited to the above examples. For those of ordinary skill in the art, improvements or modifications can be made according to the above description, and all such improvements and modifications shall fall within the protection scope of the appended claims of the present invention.

Claims

1. A method for stably identifying acupoints on the human back, characterized in that, The method for stably identifying human back acupoints includes: Obtaining a back image of the current user, obtaining the positioning coordinates of the reference acupoints based on the back image, and calculating the human body ratio of the current user according to the positioning coordinates of the reference acupoints; Dividing the reference acupoints into several groups including the left and right symmetric reference acupoints, calculating the representative points on the left and right sides in each group according to the positioning coordinates; determining the reference line according to the representative points; Obtaining the spinal alignment identification line of the current user according to the reference line; Calculating the specific positions of other acupoints according to the human body ratio and the spinal alignment identification line.

2. The method for stably identifying human back acupoints according to claim 1, wherein The calculating the representative points on the left and right sides in each group according to the positioning coordinates includes: Calculate the representative points P in several of the groups respectively l (x l ,y l ) and P r (x r ,y r ); The representative point P l (x l , y l ) is the mean point of several acupoints on the left side in the group; The representative point P r (x r , y r ) is the mean point of several acupoints on the right side in the group.

3. The method for stably identifying human back acupoints according to claim 2, wherein The determining the reference line according to the representative points includes: Based on the representative points P in several of the groups l (x l , y l ) and P r (x r , y r ), calculate the slope k and intercept b of the reference line L1 for the corresponding group, and determine the reference line L1 according to the slope k and intercept b; The calculation formula for the slope k is: k = (y r - y l ) / (x r - x l ); The calculation formula for the intercept b is: b = y l - k * x l .

4. The method for stably identifying human back acupoints according to claim 3, characterized in that The obtaining the spinal alignment identification line of the current user according to the reference line includes: Calculating the slope k2 of the spinal alignment according to several reference lines L1, and generating the spinal alignment identification line at the center point of the reference line L1 according to the slope k2; The calculation formula for the slope k2 of the spinal alignment identification line is: k2 = -1 / k.

5. The method for stably identifying human back acupoints according to claim 4, characterized in that, The calculating the specific positions of other acupoints according to the human body ratio and the spinal alignment identification line includes: Calculate the basic scale d, and determine the specific positions p of other acupoints according to the basic scale d new ; The basic dimension d is the distance between the highest point p max and the lowest point p min in the group; The calculation formula for the basic dimension d is as follows: The specific locations p of the other acupoints new The calculation formula is: x new = x min + t * d * cos(arctan(k2)); y new = y min + t * d * sin(arctan(k2)).

6. The method for stably identifying human back acupoints according to claim 1, wherein The obtaining a back image of the current user and obtaining the positioning coordinates of the reference acupoints based on the back image includes: Processing the back image of the current user using a trained model; Before processing the back image of the current user using the trained model, it further includes: Constructing a trained model, where the trained model includes a human back acupoint data set, and the human back acupoint data set includes several human back images with coordinate data of the reference acupoints.

7. The method for stably identifying human back acupoints according to claim 6, wherein The human back acupoint data set further includes enhanced data; The enhanced data is the image obtained by randomly rotating, scaling, and performing brightness processing on several pairs of the human back images.

8. The method for stably identifying human back acupoints according to claim 1, characterized in that, The dividing the reference acupoints into several groups including the left and right symmetric reference acupoints includes: Grouping according to the area where the reference acupoints are located in the back image; The groups include a shoulder area group, a middle back area group, and a waist area group; The shoulder area group includes 8 acupoints symmetrically distributed on the left and right, namely Jianzhongshu, Jianwaishu, Bingfeng, and Tianzong; The middle back area group includes 6 acupoints symmetrically distributed on the left and right, namely Geshu, Hunmen, and Sanjiaoshu; The waist area group includes 12 acupoints symmetrically distributed on the left and right, namely Guanyuanshu, Pangguangshu, Baihuanshu, Zhibian, Huiyang, and Jingmen.

9. The method for stably identifying human back acupoints according to claim 8, characterized in that, The dividing the reference acupoints into several groups including the left and right symmetric reference acupoints further includes: Or, grouping according to the meridian connection relationship of the reference acupoints in the back image; Or, grouping according to the functional attributes of the reference acupoints in the back image; Or, grouping according to the recognition confidence of the reference acupoints in the back image.

10. An intelligent terminal, characterized in that, The intelligent terminal includes a memory, a processor, and a program for stably identifying acupoints on the human back stored in the memory and executable on the processor. When the processor executes the program for stably identifying acupoints on the human back, the method for stably identifying acupoints on the human back as described in any one of the above claims 1-9 is implemented; The intelligent terminal further includes a display, and the display can simultaneously display at least two display pages. One of the display pages is used to display the back image of the current user, and the other display page is used to display the back image of the current user with the reference acupoints and other acupoints marked.

Images