Ear acupuncture point reflex zone identification method, device, acupuncture robot and storage medium

Abstract

The present invention provides an ear acupuncture point reflex zone recognition method, device, acupuncture robot and storage medium, belonging to the field of acupuncture point recognition technology, and the method includes: based on a coarse-grained segmentation module, coarse-grained segmentation is performed on an ear image to obtain an ear structure segmentation map and an ear function zone segmentation map; based on the ear structure segmentation map, whether the ear function zone segmentation map is accurate is determined; when the ear function zone segmentation map is accurate, fine-grained segmentation is performed on the ear function zone segmentation map based on a fine-grained segmentation module to obtain multiple ear acupuncture point reflex zones. The present invention utilizes the relationship between the ear's own structure and the ear function zone to verify the accuracy of the ear function zone segmentation map, and when it is determined that the ear function zone segmentation map is accurate, fine-grained segmentation is performed on the ear function zone segmentation map based on the fine-grained segmentation module to obtain the ear acupuncture point reflex zone, thereby ensuring the accuracy of the determined ear acupuncture point reflex zone.

Description

Technical Field

[0001] The present invention relates to the technical field of acupoint recognition, and in particular to an ear acupoint reflex zone recognition method, a device, an acupuncture robot and a storage medium. Background Art

[0002] Ear acupuncture points are specific areas distributed on the ear, corresponding to various organs and functional systems of the human body. By stimulating these acupuncture points, the physiological and pathological states of the body can be regulated. Auricular acupuncture therapy is a traditional Chinese medical treatment method that combines the theory of meridians and internal organs with modern medical methods to prevent and treat diseases by stimulating the corresponding reaction areas on the ear. Accurate identification of ear acupuncture points is crucial to the use of auricular acupuncture therapy and is also an important part of intelligent acupuncture robots and massage robots.

[0003] The existing method for ear acupoint recognition is to input the ear image into a mature neural network model to obtain the recognition result. Since there are many ear acupoints with different densities, for example, a small area includes many ear acupoints, while another larger area includes only a few ear acupoints. The existing ear acupoint recognition method does not consider the distribution characteristics of ear acupoints, resulting in a low recognition accuracy of ear acupoints.

[0004] Therefore, there is an urgent need to provide an ear acupuncture point reflex zone identification method, device, acupuncture robot and storage medium to achieve accurate identification of ear acupuncture points. Summary of the invention

[0005] In view of this, it is necessary to provide a method, device, acupuncture robot and storage medium for identifying ear acupoint reflex zones to solve the technical problem in the prior art that the distribution characteristics of ear acupoints are not taken into consideration, resulting in low recognition accuracy of ear acupoints.

[0006] On the one hand, in order to solve the above technical problems, the present invention provides an ear acupuncture point reflexology area recognition method, which is used to recognize ear images based on a multi-level deep convolutional network model to obtain multiple ear acupuncture point reflexology areas, wherein the multi-level deep convolutional network model includes a coarse-grained segmentation module and a fine-grained segmentation module; the method includes:

[0007] Performing coarse-grained segmentation on the ear image based on the coarse-grained segmentation module to obtain an ear structure segmentation map and an ear function area segmentation map;

[0008] Determining whether the ear function area segmentation map is accurate based on the ear structure segmentation map;

[0009] When the ear function area segmentation map is accurate, fine-grained segmentation is performed on the ear function area segmentation map based on the fine-grained segmentation module to obtain the multiple ear acupuncture point reflexology areas.

[0010] In a possible implementation, the ear structure segmentation map includes multiple ear structure regions, and the ear function region segmentation map includes multiple ear function regions; then determining whether the ear function region segmentation map is accurate based on the ear structure segmentation map includes:

[0011] Determining the inclusion relationship between the ear structure area and the ear function area;

[0012] Determining whether the inclusion relationship conforms to the theoretical inclusion relationship;

[0013] When the inclusion relationship conforms to the theoretical inclusion relationship, the ear functional area segmentation map is accurate;

[0014] When the inclusion relationship does not conform to the theoretical inclusion relationship, the ear function area segmentation map is inaccurate.

[0015] In a possible implementation, the multiple ear structure areas include an earlobe area, a concha area, a scaphoid area, an antihelix area, a triangular fossa area, and a helix area; the multiple ear function areas include a head and neck function area, a circulation function area, a respiratory function area, an endocrine function area, a digestive function area, a trunk function area, an upper limb function area, a spinal function area, a lower limb function area, a urinary function area, a reproductive function area, and a neural function area;

[0016] The theory contains the relationship:

[0017] The head and neck functional area is located in the earlobe area and the helix area;

[0018] The circulatory function area, the respiratory function area and the endocrine function area are located in the concha area;

[0019] The trunk functional area, the spinal column functional area and the lower limb functional area are located in the antihelix area;

[0020] The upper limb functional area is located in the scaphoid region;

[0021] The reproductive function area is located in the triangular fossa area;

[0022] The digestive function area is located in the antihelix area and the concha area;

[0023] The urinary function area is located in the helix area and the concha area;

[0024] The nerve function area is located in the antihelix area and the triangular fossa area.

[0025] In a possible implementation manner, before determining whether the ear function area segmentation map is accurate based on the ear structure segmentation map, the method further includes:

[0026] Acquiring the positional relationship of the multiple ear structure regions;

[0027] Determining whether the orientation relationship conforms to the theoretical orientation relationship;

[0028] When the orientation relationship conforms to the theoretical orientation relationship, determining that the ear structure segmentation map is accurate;

[0029] When the orientation relationship does not conform to the theoretical orientation relationship, it is determined that the ear structure segmentation map is inaccurate.

[0030] In a possible implementation, the theoretical orientation relationship is:

[0031] The earlobe region is adjacent to the concha region, the scaphoid region, the antihelix region and the helix region respectively;

[0032] The concha region is adjacent to the earlobe region, the antihelix region and the helix region respectively;

[0033] The scaphoid region is adjacent to the helix region, the antihelix region and the earlobe region respectively;

[0034] The antihelix region is adjacent to the scaphoid region, the helix region and the earlobe region respectively;

[0035] The triangular fossa region is adjacent to the antihelix region and the scaphoid region respectively;

[0036] The helix region is adjacent to the earlobe region, the concha region, the scaphoid region, the antihelix region, and the triangular fossa region respectively.

[0037] In a possible implementation, the multiple ear acupuncture point reflex areas include tonsil reflex area, inner ear reflex area, eye reflex area, frontal reflex area, jaw reflex area, tongue reflex area, tooth reflex area, occipital reflex area, apex reflex area, forehead reflex area, wheel one reflex area, wheel two reflex area, wheel three reflex area, wheel four reflex area, nodule reflex area and ear tip reflex area located in the head and neck functional area, heart reflex area located in the circulation functional area, lung reflex area and trachea reflex area located in the respiratory functional area, endocrine reflex area located in the endocrine functional area, mouth reflex area, esophagus reflex area, cardia reflex area, stomach reflex area, spleen reflex area, liver reflex area, pancreas and gallbladder reflex area, duodenum reflex area, small intestine reflex area located in the digestive functional area The reflex areas include the radicular reflex areas, large intestine reflex areas, rectal reflex areas, middle ear reflex areas and triple burner reflex areas; the cervical reflex areas, chest reflex areas, abdominal reflex areas and arm reflex areas located in the trunk functional area; the clavicle reflex areas, shoulder reflex areas, elbow reflex areas, wrist reflex areas and finger reflex areas located in the upper limb functional area; the cervical vertebra reflex areas, spinal reflex areas and lumbar sacral vertebra reflex areas located in the spinal column functional area; the hip reflex areas, knee reflex areas, ankle reflex areas, toe reflex areas and heel reflex areas located in the lower limb functional area; the kidney reflex areas, bladder reflex areas, urethra reflex areas and anus reflex areas located in the urinary function area; the pelvic reflex areas, angular fossa reflex areas and internal genital reflex areas located in the reproductive function area; the Shenmen reflex area and sciatic nerve reflex area located in the nerve function area.

[0038] In a possible implementation, the coarse-grained segmentation module includes a shared feature extraction unit and a parallel structural coarse-grained segmentation unit and a functional coarse-grained segmentation unit; then the coarse-grained segmentation of the ear image based on the coarse-grained segmentation module to obtain an ear structure segmentation map and an ear function area segmentation map includes:

[0039] Performing feature extraction on the ear image based on the shared feature extraction unit to obtain an initial feature map;

[0040] Based on the structural coarse-grained segmentation unit, the initial feature map is coarsely divided to obtain the ear structure segmentation map;

[0041] The initial feature map is coarsely divided based on the functional coarse-grained segmentation unit to obtain the ear functional area segmentation map.

[0042] On the other hand, the present invention also provides an ear acupuncture point reflex zone recognition device, which is used to recognize ear images based on a multi-level deep convolutional network model to obtain multiple ear acupuncture point reflex zones, wherein the multi-level deep convolutional network model includes a coarse-grained segmentation module and a fine-grained segmentation module; the device includes:

[0043] A coarse-grained segmentation unit, configured to perform coarse-grained segmentation on the ear image based on the coarse-grained segmentation module to obtain an ear structure segmentation map and an ear function area segmentation map;

[0044] an accuracy judging unit, configured to determine whether the ear function area segmentation map is accurate based on the ear structure segmentation map;

[0045] A fine-grained segmentation unit is used to perform fine-grained segmentation on the ear function area segmentation map based on the fine-grained segmentation module when the ear function area segmentation map is accurate, so as to obtain the multiple ear acupuncture point reflex areas.

[0046] On the other hand, the present invention also provides an acupuncture robot, comprising an acupuncture point search subsystem for determining an ear acupuncture point reflex zone and a moxibustion subsystem for performing moxibustion based on the ear acupuncture point reflex zone, wherein the acupuncture point search subsystem comprises a memory and a processor, wherein:

[0047] The memory is used to store programs;

[0048] The processor is coupled to the memory and is used to execute the program stored in the memory to implement the steps in the ear acupuncture point reflex zone identification method described in any one of the possible implementations above.

[0049] On the other hand, the present invention also provides a computer-readable storage medium for storing computer-readable programs or instructions, which, when executed by a processor, can implement the steps in the ear acupoint reflex zone identification method described in any of the above possible implementation methods.

[0050] The beneficial effects of the present invention are as follows: the ear acupuncture point reflex zone identification method provided by the present invention performs coarse-grained segmentation on the ear image by setting a coarse-grained segmentation module to obtain an ear structure segmentation map and an ear function area segmentation map, and judges the accuracy of the ear function area segmentation map based on the ear structure segmentation map, and verifies the accuracy of the ear function area segmentation map by utilizing the relationship between the ear's own structure and the ear function area. When it is determined that the ear function area segmentation map is accurate, the ear function area segmentation map is fine-grainedly segmented based on the fine-grained segmentation module to obtain the ear acupuncture point reflex zone, thereby ensuring the accuracy of the determined ear acupuncture point reflex zone.

[0051] Furthermore, the multi-level deep convolutional network model sets a coarse-grained segmentation module and a fine-grained segmentation module, which can retain the multi-level information of the ear image from the shallow layer to the deep layer, and perform feature extraction of different coarseness and fineness. When performing coarse-grained segmentation, coarse-grained features can be used, which improves the segmentation efficiency while ensuring the accuracy of coarse-grained segmentation. When performing fine-grained segmentation, fine-grained features are used to improve the refinement of the features, thereby ensuring the accuracy of the identified ear acupuncture point reflex areas. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative work.

[0053] Figure 1 A schematic flow chart of an embodiment of the ear acupuncture point reflexology zone identification method provided by the present invention;

[0054] Figure 2 A schematic diagram of the structure of an embodiment of a multi-level deep convolutional network model provided by the present invention;

[0055] Figure 3 For the present invention Figure 1 A schematic flow chart of an embodiment of S102;

[0056] Figure 4 A schematic diagram of the structure of an embodiment of multiple ear structure areas provided by the present invention;

[0057] Figure 5 A schematic diagram of the structure of an embodiment of multiple ear functional areas provided by the present invention;

[0058] Figure 6 A schematic diagram of a flow chart of an embodiment of verifying the accuracy of an ear structure segmentation map provided by the present invention;

[0059] Figure 7 A schematic diagram of the structure of an embodiment of multiple ear acupuncture point reflexology areas provided by the present invention;

[0060] Figure 8 A schematic diagram of the structure of an embodiment of the ear acupuncture point reflex zone identification device provided by the present invention;

[0061] Fig. 9 This is a schematic structural diagram of an embodiment of the acupuncture robot provided by the present invention. DETAILED DESCRIPTION

[0062] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0063] It should be understood that the schematic drawings are not drawn to scale. The flowchart used in the present invention shows the operations implemented according to some embodiments of the present invention. It should be understood that the operations of the flowchart can be implemented out of order, and the steps without logical context can be reversed in order or implemented simultaneously. In addition, those skilled in the art, under the guidance of the content of the present invention, can add one or more other operations to the flowchart, and can also remove one or more operations from the flowchart. Some of the block diagrams shown in the accompanying drawings are functional entities, which do not necessarily correspond to physically or logically independent entities. These functional entities can be implemented in software form, or in one or more hardware modules or integrated circuits, or in different networks and / or processor systems and / or microcontroller systems.

[0064] Reference to an "embodiment" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiment may be included in at least one embodiment of the present invention. The appearance of the phrase in various places in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0065] The present invention provides an ear acupuncture point reflex zone recognition method, device, acupuncture robot and storage medium, which are described below respectively.

[0066] Before describing the specific embodiments, the relationship between acupoints and acupoint reflexology areas is described. In some scenarios in the industry, an acupoint refers to an acupoint, and an acupoint reflexology area refers to an area including an acupoint. In other scenarios in the industry, an acupoint refers to the area where the acupoint is located, and the acupoint reflexology area has the same meaning as the acupoint. In other words, an acupoint reflexology area refers to an area that can characterize an acupoint.

[0067] Figure 1 A schematic diagram of an embodiment of the ear acupuncture point reflexology zone identification method provided by the present invention. Figure 2 A schematic diagram of the structure of an embodiment of a multi-level deep convolutional network (Hierarchical Deep CNNs, HD-CNNs) model provided by the present invention is shown in FIG. Figure 1 and Figure 2 As shown, the multi-level deep convolutional network model includes a coarse-grained segmentation module and a fine-grained segmentation module, and the ear acupuncture point reflexology area recognition method includes:

[0068] S101, performing coarse-grained segmentation on the ear image based on a coarse-grained segmentation module to obtain an ear structure segmentation map and an ear function area segmentation map;

[0069] S102, determining whether the ear function area segmentation map is accurate based on the ear structure segmentation map;

[0070] S103. When the ear function area segmentation map is accurate, fine-grained segmentation is performed on the ear function area segmentation map based on a fine-grained segmentation module to obtain a plurality of ear acupuncture point reflex areas.

[0071] It should be noted that both the coarse-grained segmentation module and the fine-grained segmentation module are composed of multiple convolutional layers.

[0072] It should be understood that before recognizing ear images based on a multi-level deep convolutional network model, the multi-level deep convolutional network model needs to be trained, tested and verified based on a sample set to ensure the model performance of the multi-level deep convolutional network model.

[0073] In some embodiments of the present invention, in order to further ensure the recognition accuracy of the ear acupuncture point reflex zones, before step S101, the ear image may be subjected to image processing such as denoising and enhancement to prevent other factors such as noise from affecting the recognition result.

[0074] Among them, denoising and enhancement processing are mature image processing methods and will not be described in detail here.

[0075] Compared with the prior art, the ear acupoint reflex zone identification method provided by the embodiment of the present invention performs coarse-grained segmentation on the ear image by setting a coarse-grained segmentation module to obtain an ear structure segmentation map and an ear function area segmentation map, and judges the accuracy of the ear function area segmentation map based on the ear structure segmentation map, and verifies the accuracy of the ear function area segmentation map by utilizing the relationship between the ear's own structure and the ear function area. When it is determined that the ear function area segmentation map is accurate, the ear function area segmentation map is fine-grainedly segmented based on the fine-grained segmentation module to obtain the ear acupoint reflex zone, thereby ensuring the accuracy of the determined ear acupoint reflex zone.

[0076] Furthermore, the multi-level deep convolutional network model sets a coarse-grained segmentation module and a fine-grained segmentation module, which can retain the multi-level information of the ear image from the shallow layer to the deep layer, and perform feature extraction of different coarseness and fineness. When performing coarse-grained segmentation, coarse-grained features can be used, which improves the segmentation efficiency while ensuring the accuracy of coarse-grained segmentation. When performing fine-grained segmentation, fine-grained features are used to improve the refinement of the features, thereby ensuring the accuracy of the identified ear acupuncture point reflex areas.

[0077] In some embodiments of the present invention, the ear structure segmentation map includes multiple ear structure regions, and the ear function region segmentation map includes multiple ear function regions.

[0078] Since the ear structure area is an inherent feature of the ear, there is an unchanging theoretical correlation between the ear function area and the ear structure area. Therefore, the accuracy of the segmented ear function area can be verified based on this theoretical correlation.

[0079] Among them, the theoretical correlation relationship can be a theoretical inclusion relationship, a theoretical position relationship, etc., which will not be elaborated here one by one.

[0080] The theoretical inclusion relationship is the area inclusion relationship between the ear structure area and the ear function area, that is, a certain ear structure area theoretically contains several ear function areas. The theoretical position relationship is the position relationship between the center point of the ear function area and the center point of the ear structure area determined based on the established coordinate system.

[0081] In order to simplify the amount of calculation and improve the efficiency of judging whether the ear function area segmentation map is accurate, in some embodiments of the present invention, the accuracy of the ear function area segmentation map is judged by using a theoretical inclusion relationship. Specifically, Figure 3 As shown, step S102 includes:

[0082] S301, determining the inclusion relationship between the ear structure area and the ear function area;

[0083] S302, judging whether the inclusion relationship conforms to the theoretical inclusion relationship;

[0084] S303, when the inclusion relationship conforms to the theoretical inclusion relationship, the ear functional area segmentation map is accurate;

[0085] S304. When the inclusion relationship does not conform to the theoretical inclusion relationship, the ear functional area segmentation map is inaccurate.

[0086] In a specific embodiment of the present invention, Figure 4 As shown, the multiple ear structure regions include the earlobe region (the skin color region in the figure), the concha region (the green region in the figure), the scaphoid region (the gray region in the figure), the antihelix region (the yellow region in the figure), the triangular fossa region (the orange region in the figure), and the helix region (the pink region in the figure). Figure 5 As shown, multiple ear functional areas include head and neck functional area 1, circulatory functional area 2, respiratory functional area 3, endocrine functional area 4, digestive functional area 5, trunk functional area 6, upper limb functional area 7, spinal functional area 8, lower limb functional area 9, urinary functional area 10, reproductive functional area 11 and nervous functional area 12.

[0087] The theoretical inclusion relationship is specifically:

[0088] The functional areas of the head and neck are located in the earlobe and helix regions;

[0089] The circulatory, respiratory and endocrine functional areas are located in the concha area;

[0090] The trunk functional area, spine functional area and lower limb functional area are located in the antihelix area;

[0091] The functional area of ​​the upper limb is located in the scaphoid region;

[0092] The reproductive function area is located in the triangular fossa area;

[0093] The digestive function area is located in the antihelix and concha areas;

[0094] The urinary function area is located in the helix and concha regions;

[0095] The functional areas of the nerve are located in the antihelix area and the triangular fossa area.

[0096] The embodiment of the present invention ensures the accuracy of each ear functional area by setting a theoretical inclusion relationship between each ear functional area and multiple ear structural areas, thereby further improving the accuracy of the determined ear acupuncture point reflex areas.

[0097] From the above description, it can be seen that the accuracy of the ear function area segmentation map needs to be verified based on the ear structure segmentation map, and the accuracy of the ear structure segmentation map is crucial. In order to avoid the technical problem of low accuracy of ear acupuncture point reflexology areas when the ear structure segmentation map is inaccurate, in some embodiments of the present invention, before step S102, the ear acupuncture point reflexology area recognition method also includes verifying the accuracy of the ear structure segmentation map.

[0098] There is a theoretical inherent distribution in the distribution of each ear structure region in the ear structure segmentation map. For example, when the earlobe region is located above the triangular fossa region, the ear structure segmentation map is inaccurate. Therefore, the accuracy of the ear structure segmentation map can be verified based on the theoretical inherent distribution of each ear structure region in the ear structure segmentation map.

[0099] It should be noted that the theoretical inherent distribution of the ear structure area can be characterized by position parameters, adjacency parameters, etc. Since the position parameters require accurate position coordinates, in a preferred embodiment of the present invention, verification is performed based on the adjacency relationship, that is, the theoretical orientation relationship of multiple ear structure areas.

[0100] Specifically, Figure 6 As shown, before step S102, the ear acupuncture point reflexology area recognition method further includes:

[0101] S601, obtaining the orientation relationship of multiple ear structure regions;

[0102] S602, determining whether the orientation relationship conforms to the theoretical orientation relationship;

[0103] S603, when the orientation relationship conforms to the theoretical orientation relationship, determining that the ear structure segmentation map is accurate;

[0104] S604: When the orientation relationship does not conform to the theoretical orientation relationship, determine that the ear structure segmentation map is inaccurate.

[0105] Before executing step S102, the embodiment of the present invention verifies the accuracy of the ear structure segmentation map based on the theoretical orientation relationship, thereby ensuring the accuracy of the ear structure segmentation map and further improving the accuracy of the ear acupuncture point reflexology area finally determined.

[0106] It should be noted that, in some embodiments of the present invention, the multi-level deep convolutional network model also includes a feature fusion module arranged between the coarse-grained segmentation module and the fine-grained segmentation module. The feature fusion module is used to execute the above-mentioned process of correcting the ear structure segmentation map and the ear function area segmentation map, that is: based on the theoretical inclusion relationship and orientation relationship, the ear structure area and the ear function area are corrected. After the correction, the ear function area segmentation map is input into the fine-grained segmentation module for fine-grained segmentation to obtain multiple ear acupoint reflex areas.

[0107] In a specific embodiment of the present invention, the theoretical orientation relationship is:

[0108] The earlobe region is adjacent to the concha region, scaphoid region, antihelix region, and helix region;

[0109] The concha region is adjacent to the earlobe region, the antihelix region, and the helix region;

[0110] The scaphoid region is adjacent to the helix region, antihelix region, and earlobe region;

[0111] The antihelix region is adjacent to the scaphoid region, the helix region, and the earlobe region;

[0112] The triangular fossa region is adjacent to the antihelix region and the scaphoid region;

[0113] The helix region is adjacent to the earlobe region, concha region, scaphoid region, antihelix region, and triangular fossa region respectively.

[0114] In a specific embodiment of the present invention, the ear acupuncture point reflex areas obtained after fine-grained segmentation based on the ear structure area and the ear function area are as follows: Figure 7As shown, the ear acupuncture point reflex areas include tonsil reflex area 1.1, inner ear reflex area 1.2, eye reflex area 1.3, anterior vertebral reflex area 1.4, jaw reflex area 1.5, tongue reflex area 1.6, tooth reflex area 1.7, occipital reflex area 1.8, apex reflex area 1.9, forehead reflex area 1.10, chakra 1 reflex area 1.14, chakra 2 reflex area 1.13, chakra 3 reflex area 1.12, chakra 4 reflex area 1.11, tubercle reflex area 1.15 and ear tip reflex area located in the functional area of ​​the head and neck. 1.16, heart reflex zone 2.1 located in the circulatory function area, lung reflex zone 3.1 and trachea reflex zone 3.2 located in the respiratory function area, endocrine reflex zone 4.1 located in the endocrine function area, mouth reflex zone 5.1, esophagus reflex zone 5.2, cardia reflex zone 5.3, stomach reflex zone 5.4, spleen reflex zone 5.5, liver reflex zone 5.6, pancreas and gallbladder reflex zone 5.7, duodenum reflex zone 5.8, small intestine reflex zone 5.9, large intestine reflex zone 5.10, The rectal reflex zone is 5.11, the middle ear reflex zone is 5.12 and the triple energizer reflex zone is 5.13; the cervical reflex zone is 6.1, the chest reflex zone is 6.2, the abdomen reflex zone is 6.3 and the arm reflex zone is 6.4 in the trunk functional area; the clavicle reflex zone is 7.1, the shoulder reflex zone is 7.2, the elbow reflex zone is 7.3, the wrist reflex zone is 7.4 and the finger reflex zone is 7.5 in the upper limb functional area; the cervical reflex zone is 8.1, the spinal reflex zone is 8.2 and the lumbar reflex zone is 8.3 in the spine functional area; the The hip reflex zone 9.1, knee reflex zone 9.2, ankle reflex zone 9.3, toe reflex zone 9.4 and heel reflex zone 9.5 in the area of ​​​​the urinary function; the kidney reflex zone 10.1, bladder reflex zone 10.2, urethra reflex zone 10.3 and anus reflex zone 10.4 in the urinary function area; the pelvic reflex zone 11.1, the fossa reflex zone 11.2 and the internal genital reflex zone 11.3 in the reproductive function area; the Shenmen reflex zone 12.1 and the sciatic nerve reflex zone 12.2 in the nerve function area.

[0115] To further improve the recognition efficiency of ear acupuncture point reflexology areas, in some embodiments of the present invention, Figure 2 As shown, the coarse-grained segmentation module includes a shared feature extraction unit and a parallel structural coarse-grained segmentation unit and a functional coarse-grained segmentation unit; then step S101 includes:

[0116] Extract features from the ear image based on the shared feature extraction unit to obtain an initial feature map;

[0117] The initial feature map is coarsely divided based on the structural coarse-grained segmentation unit to obtain an ear structure segmentation map;

[0118] The initial feature map is coarsely divided based on the functional coarse-grained segmentation unit to obtain the ear functional area segmentation map.

[0119] The embodiment of the present invention extracts the initial feature map of the ear image through a shared feature extraction unit. The initial feature map can be simultaneously provided to the parallel structural coarse-grained segmentation unit and the functional coarse-grained segmentation unit for identification, thereby realizing the simultaneous identification of the ear structure segmentation map and the ear functional area segmentation map, shortening the recognition time of the ear structure segmentation map and the ear functional area segmentation map, and thereby improving the recognition efficiency of the ear acupuncture point reflex area.

[0120] It should be understood that the shared feature extraction unit, the structural coarse-grained segmentation unit and the functional coarse-grained segmentation unit are all composed of multiple layers of convolutional layers.

[0121] In order to better implement the ear acupuncture point reflex zone recognition method in the embodiment of the present invention, based on the ear acupuncture point reflex zone recognition method, the embodiment of the present invention also provides an ear acupuncture point reflex zone recognition device, which is used to recognize ear images based on a multi-level deep convolutional network model to obtain multiple ear acupuncture point reflex zones, and the multi-level deep convolutional network model includes a coarse-grained segmentation module and a fine-grained segmentation module; Figure 8 As shown, the ear acupuncture point reflection zone identification device 800 includes:

[0122] A coarse-grained segmentation unit 801 is used to perform coarse-grained segmentation on the ear image based on the coarse-grained segmentation module to obtain an ear structure segmentation map and an ear function area segmentation map;

[0123] An accuracy judging unit 802 is used to determine whether the ear function area segmentation map is accurate based on the ear structure segmentation map;

[0124] The fine-grained segmentation unit 803 is used to perform fine-grained segmentation on the ear function area segmentation map based on the fine-grained segmentation module when the ear function area segmentation map is accurate, so as to obtain multiple ear acupuncture point reflexology areas.

[0125] The ear acupoint reflex zone identification device 800 provided in the above embodiment can implement the technical solution described in the above ear acupoint reflex zone identification method embodiment. The specific implementation principles of the above modules or units can refer to the corresponding contents in the above ear acupoint reflex zone identification method embodiment, which will not be repeated here.

[0126] like Fig. 9 As shown, the present invention also provides an acupuncture robot 900. The acupuncture robot 900 includes an acupuncture point search subsystem 910 for determining ear acupuncture point reflex zones and a moxibustion subsystem 920 for performing acupuncture based on the ear acupuncture point reflex zones. The acupuncture point search subsystem 910 includes a processor 911, a memory 912 and a display 913. Fig. 9 Only some of the components of the acupuncture robot 900 are shown, but it should be understood that it is not required to implement all of the shown components, and more or fewer components may be implemented instead.

[0127] In some embodiments, the memory 912 may be an internal storage unit of the cave-finding subsystem 910, such as a hard disk or memory of the cave-finding subsystem 910. In other embodiments, the memory 912 may also be an external storage device of the cave-finding subsystem 910, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc., equipped on the cave-finding subsystem 910.

[0128] Furthermore, the memory 912 may include both an internal storage unit and an external storage device of the cave-finding subsystem 910. The memory 912 is used to store application software installed in the cave-finding subsystem 910 and various data.

[0129] In some embodiments, the processor 911 may be a central processing unit (CPU), a microprocessor or other data processing chip, which is used to run the program code or process data stored in the memory 912, such as the ear acupuncture point reflex zone identification method of the present invention.

[0130] In some embodiments, the display 913 may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, an OLED (Organic Light-Emitting Diode) touch device, etc. The display 913 is used to display information of the acupoint-finding subsystem 910 and to display a visual user interface. The components 911-913 of the acupoint-finding subsystem 910 communicate with each other via a system bus.

[0131] In some embodiments of the present invention, when the processor 911 executes the ear acupuncture point reflexology zone recognition program in the memory 912, the following steps may be implemented:

[0132] Based on the coarse-grained segmentation module, the ear image is coarsely segmented to obtain an ear structure segmentation map and an ear function area segmentation map;

[0133] Determining whether the ear functional area segmentation map is accurate based on the ear structure segmentation map;

[0134] When the ear function area segmentation map is accurate, fine-grained segmentation is performed on the ear function area segmentation map based on the fine-grained segmentation module to obtain multiple ear acupuncture point reflex areas.

[0135] It should be understood that: when the processor 911 executes the ear acupuncture point reflex zone identification program in the memory 912, in addition to the above functions, other functions can also be implemented. For details, please refer to the description of the previous related method embodiments.

[0136] Among them, Fig. 9 As shown, the moxibustion subsystem 920 includes a path planning unit 921 and a moxibustion unit 922. The path planning unit 921 is used to perform path planning according to the moxibustion needle positions and acupoint coordinates in the moxibustion unit 922 to generate a planned path. The moxibustion unit 922 moves the moxibustion needles to the locations of the acupoint reflex areas according to the planned path to perform moxibustion operations.

[0137] The embodiment of the present invention can replace doctors to complete long-term acupuncture treatment work by providing an acupuncture robot 900, thereby alleviating the burden on doctors and reducing medical expenses.

[0138] Correspondingly, an embodiment of the present invention also provides a computer-readable storage medium, which is used to store computer-readable programs or instructions. When the program or instructions are executed by a processor, the steps or functions in the ear acupoint reflex zone identification method provided in the above-mentioned method embodiments can be implemented.

[0139] Those skilled in the art will appreciate that all or part of the processes of the above-mentioned embodiments can be implemented 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, wherein the computer-readable storage medium is a disk, an optical disk, a read-only storage memory, or a random access memory, etc.

[0140] The above is a detailed introduction to the ear acupuncture point reflex zone identification method, device, acupuncture robot and storage medium provided by the present invention. Specific examples are used in this article to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea; at the same time, for technical personnel in this field, according to the idea of ​​the present invention, there will be changes in the specific implementation method and application scope. In summary, the content of this specification should not be understood as a limitation on the present invention.

Claims

1. A method for identifying ear acupuncture point reflex zones, characterized in that: For identifying ear images based on a multi-level deep convolutional network model to obtain multiple ear acupuncture point reflexology areas, the multi-level deep convolutional network model includes a coarse-grained segmentation module and a fine-grained segmentation module; the method includes: Performing coarse-grained segmentation on the ear image based on the coarse-grained segmentation module to obtain an ear structure segmentation map and an ear function area segmentation map; Determining whether the ear function area segmentation map is accurate based on the ear structure segmentation map; When the ear function area segmentation map is accurate, fine-grained segmentation is performed on the ear function area segmentation map based on the fine-grained segmentation module to obtain the multiple ear acupuncture point reflexology areas; The ear structure segmentation map includes a plurality of ear structure regions, and the ear function region segmentation map includes a plurality of ear function regions; then determining whether the ear function region segmentation map is accurate based on the ear structure segmentation map includes: Determining the inclusion relationship between the ear structure area and the ear function area; Determining whether the inclusion relationship conforms to the theoretical inclusion relationship; When the inclusion relationship conforms to the theoretical inclusion relationship, the ear functional area segmentation map is accurate; When the inclusion relationship does not conform to the theoretical inclusion relationship, the ear function area segmentation map is inaccurate; The multiple ear structure areas include an earlobe area, a concha area, a scaphoid area, an antihelix area, a triangular fossa area and a helix area; the multiple ear function areas include a head and neck function area, a circulation function area, a respiratory function area, an endocrine function area, a digestive function area, a trunk function area, an upper limb function area, a spinal function area, a lower limb function area, a urinary function area, a reproductive function area and a nerve function area; The theory contains the relationship: The head and neck functional area is located in the earlobe area and the helix area; The circulatory function area, the respiratory function area and the endocrine function area are located in the concha area; The trunk functional area, the spinal column functional area and the lower limb functional area are located in the antihelix area; The upper limb functional area is located in the scaphoid region; The reproductive function area is located in the triangular fossa area; The digestive function area is located in the antihelix area and the concha area; The urinary function area is located in the helix area and the concha area; The nerve function area is located in the antihelix area and the triangular fossa area.

2. The ear acupuncture point reflexology zone identification method according to claim 1, characterized in that: Before determining whether the ear function area segmentation map is accurate based on the ear structure segmentation map, the method further includes: Acquiring the positional relationship of the multiple ear structure regions; Determining whether the orientation relationship conforms to the theoretical orientation relationship; When the orientation relationship conforms to the theoretical orientation relationship, determining that the ear structure segmentation map is accurate; When the orientation relationship does not conform to the theoretical orientation relationship, it is determined that the ear structure segmentation map is inaccurate.

3. The ear acupuncture point reflexology zone identification method according to claim 2, characterized in that: The theoretical orientation relationship is: The earlobe region is adjacent to the concha region, the scaphoid region, the antihelix region and the helix region respectively; The concha region is adjacent to the earlobe region, the antihelix region and the helix region respectively; The scaphoid region is adjacent to the helix region, the antihelix region and the earlobe region respectively; The antihelix region is adjacent to the scaphoid region, the helix region and the earlobe region respectively; The triangular fossa region is adjacent to the antihelix region and the scaphoid region respectively; The helix region is adjacent to the earlobe region, the concha region, the scaphoid region, the antihelix region, and the triangular fossa region respectively.

4. The ear acupuncture point reflexology zone identification method according to claim 1, characterized in that: The multiple ear acupuncture point reflex areas include tonsil reflex area, inner ear reflex area, eye reflex area, anterior vertebral reflex area, jaw reflex area, tongue reflex area, tooth reflex area, occipital reflex area, apex reflex area, forehead reflex area, chakra one reflex area, chakra two reflex area, chakra three reflex area, chakra four reflex area, nodule reflex area and ear tip reflex area located in the head and neck functional area, heart reflex area located in the circulation functional area, lung reflex area and trachea reflex area located in the respiratory functional area, endocrine reflex area located in the endocrine functional area, mouth reflex area, esophagus reflex area, cardia reflex area, stomach reflex area, spleen reflex area, liver reflex area, pancreas and gallbladder reflex area, duodenum reflex area, small intestine reflex area, large intestine reflex area located in the digestive functional area The reflex zone, rectal reflex zone, middle ear reflex zone and triple burner reflex zone are located in the trunk functional area; the cervical reflex zone, chest reflex zone, abdominal reflex zone and arm reflex zone are located in the upper limb functional area; the clavicle reflex zone, shoulder reflex zone, elbow reflex zone, wrist reflex zone and finger reflex zone are located in the upper limb functional area; the cervical vertebra reflex zone, spinal reflex zone and lumbar sacral reflex zone are located in the spinal column functional area; the hip reflex zone, knee reflex zone, ankle reflex zone, toe reflex zone and heel reflex zone are located in the lower limb functional area; the kidney reflex zone, bladder reflex zone, urethra reflex zone and anus reflex zone are located in the urinary function area; the pelvic reflex zone, angular fossa reflex zone and internal genital reflex zone are located in the reproductive function area; the Shenmen reflex zone and sciatic nerve reflex zone are located in the nerve function area.

5. The ear acupuncture point reflexology zone identification method according to claim 1, characterized in that: The coarse-grained segmentation module includes a shared feature extraction unit and a parallel structural coarse-grained segmentation unit and a functional coarse-grained segmentation unit; then the coarse-grained segmentation of the ear image based on the coarse-grained segmentation module to obtain an ear structure segmentation map and an ear function area segmentation map includes: Performing feature extraction on the ear image based on the shared feature extraction unit to obtain an initial feature map; Based on the structural coarse-grained segmentation unit, the initial feature map is coarsely divided to obtain the ear structure segmentation map; The initial feature map is coarsely divided based on the functional coarse-grained segmentation unit to obtain the ear functional area segmentation map.

6. An ear acupuncture point reflex zone identification device, characterized in that: Used to identify ear images based on a multi-level deep convolutional network model to obtain multiple ear acupuncture point reflex areas, the multi-level deep convolutional network model includes a coarse-grained segmentation module and a fine-grained segmentation module; the device includes: A coarse-grained segmentation unit, configured to perform coarse-grained segmentation on the ear image based on the coarse-grained segmentation module to obtain an ear structure segmentation map and an ear function area segmentation map; an accuracy judging unit, configured to determine whether the ear function area segmentation map is accurate based on the ear structure segmentation map; A fine-grained segmentation unit, configured to perform fine-grained segmentation on the ear function area segmentation map based on the fine-grained segmentation module when the ear function area segmentation map is accurate, so as to obtain the plurality of ear acupuncture point reflexology areas; The ear structure segmentation map includes a plurality of ear structure regions, and the ear function region segmentation map includes a plurality of ear function regions; then determining whether the ear function region segmentation map is accurate based on the ear structure segmentation map includes: Determining the inclusion relationship between the ear structure area and the ear function area; Determining whether the inclusion relationship conforms to the theoretical inclusion relationship; When the inclusion relationship conforms to the theoretical inclusion relationship, the ear functional area segmentation map is accurate; When the inclusion relationship does not conform to the theoretical inclusion relationship, the ear function area segmentation map is inaccurate; The multiple ear structure areas include an earlobe area, a concha area, a scaphoid area, an antihelix area, a triangular fossa area and a helix area; the multiple ear function areas include a head and neck function area, a circulation function area, a respiratory function area, an endocrine function area, a digestive function area, a trunk function area, an upper limb function area, a spinal function area, a lower limb function area, a urinary function area, a reproductive function area and a nerve function area; The theory contains the relationship: The head and neck functional area is located in the earlobe area and the helix area; The circulatory function area, the respiratory function area and the endocrine function area are located in the concha area; The trunk functional area, the spinal column functional area and the lower limb functional area are located in the antihelix area; The upper limb functional area is located in the scaphoid region; The reproductive function area is located in the triangular fossa area; The digestive function area is located in the antihelix area and the concha area; The urinary function area is located in the helix area and the concha area; The nerve function area is located in the antihelix area and the triangular fossa area.

7. An acupuncture robot, characterized in that: It includes an acupuncture point search subsystem for determining ear acupuncture point reflex zones and a moxibustion subsystem for performing moxibustion based on the ear acupuncture point reflex zones, wherein the acupuncture point search subsystem includes a memory and a processor, wherein: The memory is used to store programs; The processor is coupled to the memory and is used to execute the program stored in the memory to implement the steps in the ear acupuncture point reflex zone identification method described in any one of claims 1 to 5 above.

8. A computer-readable storage medium, characterized in that: The readable storage medium stores a program or instruction, and when the program or instruction is executed by the processor, the steps in the ear acupuncture point reflex zone identification method described in any one of claims 1 to 5 are implemented.

Images