Foot reflexology zone identification method, device, rehabilitation robot and storage medium

Abstract

The present invention provides a foot reflexology zone recognition method, device, rehabilitation robot and storage medium, belonging to the field of acupoint recognition technology, and the method includes: inputting a foot image to be recognized into a multi-task deep learning network model to obtain a foot part reflexology zone and a foot organ reflexology zone; the foot part reflexology zone includes a head reflexology zone, a neck reflexology zone, a shoulder and back reflexology zone, a chest reflexology zone, an abdomen reflexology zone and a reproductive reflexology zone; judging whether the multiple foot organ reflexology zones and the multiple foot part reflexology zones meet a theoretical position relationship, and when they meet, determining the foot organ reflexology zone as a target foot organ reflexology zone; wherein the theoretical position relationship is: the foot organ reflexology zone is a sub-area of ​​each of the foot part reflexology zones. The present invention uses the foot part reflexology zone as a verification condition for the foot organ reflexology zone, thereby improving the accuracy of the identified target foot organ reflexology zone.

Description

Technical Field

[0001] The present invention relates to the technical field of acupoint identification, and specifically to a foot reflexology zone identification method, device, rehabilitation robot and storage medium. Background Art

[0002] The reflex areas of the soles of the feet include the frontal sinus, nose, shoulder, thyroid, lung, stomach, kidney, liver, small intestine, thigh, gonad, buttocks and other reflex areas. Massaging the frontal sinus reflex area can relieve concussion, sinusitis, and fever. Massaging the nasal reflex area can relieve multiple sclerosis, poliomyelitis or spinal cord tumors. Massaging the thyroid reflex area can calm the mind and relieve the nerves, and massaging the lung reflex area can relieve and treat respiratory inflammation, tuberculosis, emphysema, and chest tightness. In clinical application, according to different conditions, certain reflex areas are selected for combination. Through the stimulation of the foot reflex area with massage techniques and scraping therapy, the functions of the corresponding organs, meridians, qi and blood can be adjusted, thereby achieving the purpose of treating the disease.

[0003] The traditional method of identifying foot reflex zones is to manually analyze and determine the foot reflex zones based on the experience and skills accumulated by TCM masters over the years. This method is highly subjective and requires high experience from TCM masters. The efficiency and accuracy of determining the foot reflex zones are low. In order to solve this technical problem, modern technology has proposed image recognition algorithms based on computer vision, such as YOLOv8 and other algorithms to identify foot reflex zones, which has improved the efficiency and accuracy of foot reflex zone identification to a certain extent. However, the recognition results of the image recognition algorithm cannot be verified. When the foot reflex zone identification is inaccurate, it will affect the treatment effect.

[0004] Therefore, there is an urgent need to provide a foot reflex zone identification method, device, rehabilitation robot and storage medium to verify the results of the identification algorithm, thereby ensuring the accuracy of the foot reflex zone identification results. Summary of the invention

[0005] In view of this, it is necessary to provide a foot reflex zone identification method, device, rehabilitation robot and storage medium to solve the technical problem in the prior art that the identification results cannot be verified, resulting in low accuracy of foot reflex zone identification results.

[0006] In one aspect, the present invention provides a method for identifying a foot reflexology zone, comprising:

[0007] Input the foot image to be identified into the multi-task deep learning network model to obtain the reflex areas of the foot parts and the reflex areas of the foot organs; the reflex areas of the foot parts include the reflex areas of the head, neck, shoulders and back, chest, abdomen and genitals;

[0008] Determine whether the plurality of foot organ reflexology zones and the plurality of foot part reflexology zones satisfy a theoretical position relationship, and if so, determine the foot organ reflexology zone as a target foot organ reflexology zone;

[0009] Wherein, the theoretical position relationship is: the foot organ reflex zone is a sub-area of ​​the reflex zone of each foot part.

[0010] In some possible implementations, determining whether the plurality of foot organ reflexology zones and the plurality of foot part reflexology zones satisfy a theoretical position relationship includes:

[0011] Determine a first area coordinate range of each of the foot part reflex zones in a preset coordinate system and a second area coordinate range of each of the foot organ reflex zones in the preset coordinate system;

[0012] Determining the actual positional relationship between the plurality of foot organ reflexology zones and the plurality of foot part reflexology zones based on the first area coordinate range and the second area coordinate range;

[0013] When the actual position relationship is the same as the theoretical position relationship, the multiple foot organ reflexology zones and the multiple foot part reflexology zones satisfy the theoretical position relationship.

[0014] In some possible implementations, the foot organ reflex zones include a first group of organ reflex sub-zones corresponding to the head reflex zones, a second group of organ reflex sub-zones corresponding to the neck reflex zones, a third group of organ reflex sub-zones corresponding to the shoulder and back reflex zones, a fourth group of organ reflex sub-zones corresponding to the chest reflex zones, a fifth group of organ reflex sub-zones corresponding to the abdomen reflex zones, and a sixth group of organ reflex sub-zones corresponding to the reproductive reflex zones;

[0015] Then, judging whether the plurality of foot organ reflexology zones and the plurality of foot part reflexology zones satisfy the theoretical position relationship includes:

[0016] Determine in turn whether the first group of organ reflex sub-areas and the head reflex area, the second group of organ reflex sub-areas and the neck reflex area, the third group of organ reflex sub-areas and the shoulder and back reflex area, the fourth group of organ reflex sub-areas and the chest reflex area, the fifth group of organ reflex sub-areas and the abdomen reflex area, and the sixth group of organ reflex sub-areas and the reproductive reflex area satisfy the theoretical position relationship.

[0017] In some possible implementations, before determining whether the plurality of foot organ reflexology zones and the plurality of foot part reflexology zones satisfy the theoretical position relationship, the method further includes:

[0018] Obtaining the theoretical position relationship and the actual position relationship between the reflex zones of the foot parts;

[0019] Determining whether the actual position relationship satisfies the theoretical position relationship;

[0020] Then, the step of judging whether the plurality of foot organ reflexology zones and the plurality of foot part reflexology zones satisfy the theoretical position relationship is as follows:

[0021] When the actual part orientation relationship satisfies the theoretical part orientation relationship, it is determined whether the plurality of foot organ reflexology areas and the plurality of foot part reflexology areas satisfy the theoretical position relationship.

[0022] In some possible implementations, before determining whether the plurality of foot organ reflexology zones and the plurality of foot part reflexology zones satisfy the theoretical position relationship, the method further includes:

[0023] Determine the first sub-area orientation relationship of the first group of organ reflection sub-areas, the second sub-area orientation relationship of the second group of organ reflection sub-areas, the third sub-area orientation relationship of the third group of organ reflection sub-areas, the fourth sub-area orientation relationship of the fourth group of organ reflection sub-areas, the fifth sub-area orientation relationship of the fifth group of organ reflection sub-areas, and the sixth sub-area orientation relationship of the sixth group of organ reflection sub-areas;

[0024] Determine whether the first sub-area orientation relationship, the second sub-area orientation relationship, the third sub-area orientation relationship, the fourth sub-area orientation relationship, the fifth sub-area orientation relationship, and the sixth sub-area orientation relationship satisfy a theoretical device orientation relationship;

[0025] Then, the step of judging whether the plurality of foot organ reflexology zones and the plurality of foot part reflexology zones satisfy the theoretical position relationship is as follows:

[0026] When the first sub-area orientation relationship, the second sub-area orientation relationship, the third sub-area orientation relationship, the fourth sub-area orientation relationship, the fifth sub-area orientation relationship and the sixth sub-area orientation relationship satisfy the theoretical organ position relationship, determine whether the multiple foot organ reflex zones and the multiple foot part reflex zones satisfy the theoretical position relationship.

[0027] In some possible implementations, the first group of organ reflex sub-areas includes a brain reflex sub-area, a nasal reflex sub-area, an oral reflex sub-area, an eye reflex sub-area, an ear reflex sub-area and a frontal sinus reflex sub-area; the second group of organ reflex sub-areas includes an esophageal tracheal reflex sub-area, a neck reflex sub-area, a thyroid reflex sub-area and a parathyroid reflex sub-area; the third group of organ reflex sub-areas includes a trapezius reflex sub-area and a shoulder reflex sub-area; the fourth group of organ reflex sub-areas includes a bronchial reflex sub-area and a lung reflex sub-area; the fifth group of organ reflex sub-areas includes a liver reflex sub-area, a kidney reflex sub-area, a stomach reflex sub-area, a duodenum reflex sub-area, a transverse colon reflex sub-area, an ascending colon reflex sub-area, a small intestine reflex sub-area and a bladder reflex sub-area; the sixth group of organ reflex sub-areas includes a pelvic reflex sub-area, a buttocks reflex sub-area and a gonad reflex sub-area.

[0028] In some possible implementations, the multi-task deep learning network model includes a shared backbone sub-model and a parallel organ reflexology detection head and a part reflexology detection head; then the step of inputting the foot image to be identified into the multi-task deep learning network model to obtain the foot part reflexology and foot organ reflexology includes:

[0029] Extracting features of the to-be-recognized foot image based on the shared backbone sub-model to obtain shared image features;

[0030] Performing key point detection on the shared image features based on the organ reflex zone detection head to obtain the foot organ reflex zone;

[0031] The shared image features are semantically segmented based on the part reflexology detection head to obtain the part reflexology of the foot.

[0032] On the other hand, the present invention also provides a foot reflexology zone identification device, comprising:

[0033] A foot reflexometer initial recognition unit is used to input the foot image to be recognized into a multi-task deep learning network model to obtain the foot part reflex area and the foot organ reflex area; the foot part reflex area includes the head reflex area, the neck reflex area, the shoulder and back reflex area, the chest reflex area, the abdomen reflex area and the reproductive reflex area;

[0034] a foot organ reflex zone determination unit, configured to determine whether the plurality of foot organ reflex zones and the plurality of foot part reflex zones satisfy a theoretical position relationship, and if so, determine the foot organ reflex zone as a target foot organ reflex zone;

[0035] Wherein, the theoretical position relationship is: the foot organ reflex zone is a sub-area of ​​the reflex zone of each foot part.

[0036] On the other hand, the present invention also provides a rehabilitation robot, comprising a memory and a processor, wherein:

[0037] The memory is used to store programs;

[0038] The processor is coupled to the memory and is used to execute the program stored in the memory to implement the steps in the foot reflexology zone identification method described in any one of the possible implementations above.

[0039] On the other hand, the present invention also provides a computer-readable storage medium, on which a program or instruction is stored. When the program or instruction is executed by a processor, the steps in the foot reflex zone identification method described in any possible implementation method mentioned above are implemented.

[0040] The beneficial effect of adopting the above-mentioned embodiment is: the foot reflex zone recognition method provided by the present invention, by setting the foot image to be identified to be input into the multi-task deep learning network model, obtains the foot part reflex zone and the foot organ reflex zone, and can verify the accuracy of the foot organ reflector based on the foot part reflex zone, that is: the foot part reflex zone is used as the verification condition of the foot organ reflex zone, when multiple foot organ reflex zones and multiple foot part reflex zones meet the theoretical position relationship, the foot organ reflex zone is determined to be the target foot organ reflex zone, thereby improving the accuracy of the identified target foot organ reflex zone.

[0041] Furthermore, since the organ reflex zone is a sub-area of ​​each foot part reflex zone, that is, the area of ​​the foot part reflex zone is larger than the area of ​​the foot organ reflex zone, the recognition accuracy of the foot part reflex zone is higher. Verifying the foot organ reflex zone based on the foot part reflex zone with higher recognition accuracy can further improve the accuracy of the foot organ reflex zone.

[0042] Furthermore, the multi-task deep learning network model in the present invention can obtain the reflex areas of the foot parts through semantic segmentation, and obtain the reflex areas of the foot organs through key point detection. One network model is used to realize multiple different tasks, which simplifies the model structure, saves computing resources, and thus improves the recognition efficiency of the foot reflex areas. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] 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.

[0044] Figure 1A schematic diagram of a flow chart of an embodiment of a foot reflexology zone identification method provided by the present invention;

[0045] Figure 2 A schematic diagram of an embodiment of the foot part reflexology area and foot organ reflexology area provided by the present invention;

[0046] Figure 3 For the present invention Figure 1 A schematic flow chart of an embodiment of S102;

[0047] Figure 4 A schematic diagram of a flow chart of an embodiment of verifying the reflex zones of the foot provided by the present invention;

[0048] Figure 5 A schematic diagram of a flow chart of an embodiment of verifying the reflex zones of the foot organs provided by the present invention;

[0049] Figure 6 A schematic diagram of the structure of an embodiment of the multi-task deep learning network model provided by the present invention;

[0050] Figure 7 For the present invention Figure 1 A schematic flow chart of an embodiment of S101;

[0051] Figure 8 A schematic structural diagram of an embodiment of a foot reflexology zone identification device provided by the present invention;

[0052] Fig. 9 This is a schematic structural diagram of an embodiment of the rehabilitation robot provided by the present invention. DETAILED DESCRIPTION

[0053] 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.

[0054] 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.

[0055] The descriptions of "first", "second", etc. involved in the embodiments of the present invention are only used for descriptive purposes and cannot be understood as indicating or implying their relative importance or implicitly indicating the number of the indicated technical features. Therefore, the technical features defined as "first" or "second" may explicitly or implicitly include at least one of the features.

[0056] 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.

[0057] The present invention provides a foot reflexology zone identification method, device, rehabilitation robot and storage medium, which are described below respectively.

[0058] Figure 1 This is a schematic diagram of an embodiment of the foot reflexology zone identification method proposed by the present invention. Figure 2 The figure is a schematic diagram of an embodiment of the foot part reflexology area and foot organ reflexology area proposed by the present invention, as shown in FIG. Figure 1 and Figure 2 As shown, the foot reflexology zone identification method includes:

[0059] S101, inputting the foot image to be identified into the multi-task deep learning network model to obtain the reflex areas of the foot parts and the reflex areas of the foot organs; the reflex areas of the foot parts include the reflex areas of the head, neck, shoulders and back, chest, abdomen and genitals;

[0060] S102, determining whether the multiple foot organ reflex zones and the multiple foot part reflex zones satisfy the theoretical position relationship, and if so, determining the foot organ reflex zone as the target foot organ reflex zone;

[0061] Among them, the theoretical position relationship is: the foot organ reflex zone is a sub-area of ​​the reflex zone of each foot part.

[0062] It should be understood that before step S101, the multi-task deep learning network model needs to be trained, tested and verified based on the sample data to obtain a fully trained multi-task deep learning network model that can recognize the foot image to be recognized.

[0063] Specifically, the sample data production process is as follows: using a camera to collect a foot data set, during which images from different angles are collected to ensure that the multi-task deep learning network model can accurately detect and identify foot reflexology areas and foot organ reflexology areas for foot images from different angles. In addition, to improve accuracy, the sole of the foot is required to be free of foreign matter and obstruction, so as to effectively reduce the obstruction of foreign matter on the foot reflexology area and improve the accuracy and stability of recognition.

[0064] It should also be understood that when the multiple foot organ reflex zones and the multiple foot part reflex zones do not satisfy the theoretical positional relationship, step S101 is re-executed, or the multi-task deep learning network model is re-trained.

[0065] Compared with the prior art, the foot reflex zone recognition method provided in an embodiment of the present invention obtains the foot part reflex zone and the foot organ reflex zone by setting the foot image to be identified to be input into a multi-task deep learning network model. The accuracy of the foot organ reflector can be verified based on the foot part reflex zone, that is: when multiple foot organ reflex zones and multiple foot part reflex zones meet the theoretical position relationship, the foot organ reflex zone is determined to be the target foot organ reflex zone, thereby improving the accuracy of the identified target foot organ reflex zone.

[0066] Furthermore, since the organ reflex zone is a sub-area of ​​each foot part reflex zone, that is, the area of ​​the foot part reflex zone is larger than the area of ​​the foot organ reflex zone, the recognition accuracy of the foot part reflex zone is higher. Verifying the foot organ reflex zone based on the foot part reflex zone with higher recognition accuracy can further improve the accuracy of the foot organ reflex zone.

[0067] Furthermore, the multi-task deep learning network model in the embodiment of the present invention can obtain the reflex zones of the foot parts through semantic segmentation, and obtain the reflex zones of the foot organs through key point detection. One network model is used to realize multiple different tasks, which simplifies the model structure, saves computing resources, and thus improves the recognition efficiency of the foot reflex zones.

[0068] In some embodiments of the present invention, Figure 3 As shown, step S102 includes:

[0069] S301, determining a first area coordinate range of each foot part reflex zone in a preset coordinate system and a second area coordinate range of each foot organ reflex zone in a preset coordinate system;

[0070] S302, determining the actual position relationship between a plurality of foot organ reflexology zones and a plurality of foot part reflexology zones based on the first area coordinate range and the second area coordinate range;

[0071] S303: When the actual position relationship is the same as the theoretical position relationship, the multiple foot organ reflexology zones and the multiple foot part reflexology zones satisfy the theoretical position relationship.

[0072] The process of determining the coordinate range of the first area is as follows: determining the boundary line of the first area and multiple pixel points on the boundary line, obtaining the pixel point coordinates of each pixel point in a preset coordinate system, and determining the coordinate range of the first area based on the pixel point coordinates.

[0073] The process of determining the coordinate range of the second region is the same as the process of determining the coordinate range of the first region, and will not be described in detail here.

[0074] The foot organ reflex zones include a first group of organ reflex sub-zones corresponding to the head reflex zones, a second group of organ reflex sub-zones corresponding to the neck reflex zones, a third group of organ reflex sub-zones corresponding to the shoulder and back reflex zones, a fourth group of organ reflex sub-zones corresponding to the chest reflex zones, a fifth group of organ reflex sub-zones corresponding to the abdomen reflex zones, and a sixth group of organ reflex sub-zones corresponding to the genital reflex zones.

[0075] If only the foot organ reflexology area is considered as a sub-area of ​​the foot part reflexology area, when the first group of organ reflexology sub-areas are located in the shoulder and back reflexology area, it is considered that the theoretical position relationship is satisfied. Obviously, in this case, the recognition result of the first group of organ reflexology sub-areas is wrong. In order to avoid erroneous recognition in similar situations, in some embodiments of the present invention, step S102 is specifically as follows:

[0076] Determine in turn whether the first group of organ reflex sub-areas and the head reflex area, the second group of organ reflex sub-areas and the neck reflex area, the third group of organ reflex sub-areas and the shoulder and back reflex area, the fourth group of organ reflex sub-areas and the chest reflex area, the fifth group of organ reflex sub-areas and the abdomen reflex area, and the sixth group of organ reflex sub-areas and the reproductive reflex area meet the theoretical position relationship.

[0077] Specifically, the theoretical position relationship is: the first group of organ reflex sub-areas is located in the head reflex zone, the second group of organ reflex sub-areas is located in the neck reflex zone, the third group of organ reflex sub-areas is located in the shoulder and back reflex zone, the fourth group of organ reflex sub-areas is located in the chest reflex zone, the fifth group of organ reflex sub-areas is located in the abdomen reflex zone, and the sixth group of organ reflex sub-areas is located in the reproductive reflex zone.

[0078] The embodiment of the present invention ensures that the organ reflex sub-areas of different groups are located within the corresponding foot part reflex areas by setting a further theoretical position relationship, thereby further ensuring the recognition accuracy of the foot organ reflex areas.

[0079] In a specific embodiment of the present invention, Figure 2 As shown, the first group of organ reflex sub-areas include brain reflex sub-area 1, nasal reflex sub-area 2, oral reflex sub-area 3, eye reflex sub-area 4, ear reflex sub-area 5 and frontal sinus reflex sub-area 6; the second group of organ reflex sub-areas include esophageal tracheal reflex sub-area 11, neck reflex sub-area 12, thyroid reflex sub-area 13 and parathyroid reflex sub-area 14; the third group of organ reflex sub-areas include trapezius reflex sub-area 7 and shoulder reflex sub-area 8; the fourth group of organ reflex sub-area include bronchial reflex sub-area 9 and lung reflex sub-area 10; the fifth group of organ reflex sub-area include liver reflex sub-area 15, kidney reflex sub-area 16, stomach reflex sub-area 17, duodenum reflex sub-area 18, transverse colon reflex sub-area 19, ascending colon reflex sub-area 20, small intestine reflex sub-area 21 and bladder reflex sub-area 22; the sixth group of organ reflex sub-areas include pelvic reflex sub-area 23, buttocks reflex sub-area 24 and gonad reflex sub-area 25.

[0080] Depend on Figure 2 It can be seen that the reflex zones of the foot parts have different orientation relationships, for example, the head reflex zone is located above the shoulder and back reflex zone, the shoulder and back reflex zone and the chest reflex zone are located on the left side of the neck reflex zone, etc. In order to further ensure the accuracy of the identified reflex zones of the foot parts, in some embodiments of the present invention, such as Figure 4 As shown, before step S102, it also includes:

[0081] S401, obtaining theoretical position relationship and actual position relationship between reflex zones of foot parts;

[0082] S402: Determine whether the actual position-orientation relationship satisfies the theoretical position-orientation relationship.

[0083] The embodiment of the present invention can ensure the recognition accuracy of the reflex zones of the foot parts by verifying the reflex zones of the foot parts based on the theoretical part orientation relationship.

[0084] Specifically, the theoretical position relationship is: the head reflex zone is located above the shoulder and back reflex zone and the neck reflex zone, the shoulder and back reflex zone and the chest reflex zone are located on the left side of the neck reflex zone, the abdomen reflex zone is located below the chest reflex zone, and the reproductive reflex zone is located below the abdomen reflex zone.

[0085] Here, the terms "up" and "down" refer to the direction of the toes of the foot as up and the direction of the heel as down.

[0086] Then step S102 is specifically as follows:

[0087] When the actual part orientation relationship satisfies the theoretical part orientation relationship, it is determined whether the plurality of foot organ reflexology areas and the plurality of foot part reflexology areas satisfy the theoretical position relationship.

[0088] Further, by Figure 2 It can be seen that different groups of organ reflex sub-areas also have different orientation relationships, for example, the transverse colon reflex sub-area is located above the small intestine reflex sub-area, etc. In order to further improve the recognition accuracy of the foot organ reflex area, in some embodiments of the present invention, such as Figure 5 As shown, before step S102, it also includes:

[0089] S501, determining the first sub-area orientation relationship of the first group of organ reflex sub-areas, the second sub-area orientation relationship of the second group of organ reflex sub-areas, the third sub-area orientation relationship of the third group of organ reflex sub-areas, the fourth sub-area orientation relationship of the fourth group of organ reflex sub-areas, the fifth sub-area orientation relationship of the fifth group of organ reflex sub-areas, and the sixth sub-area orientation relationship of the sixth group of organ reflex sub-areas;

[0090] S502, determining whether the first sub-area orientation relationship, the second sub-area orientation relationship, the third sub-area orientation relationship, the fourth sub-area orientation relationship, the fifth sub-area orientation relationship, and the sixth sub-area orientation relationship satisfy the theoretical device official orientation relationship;

[0091] Then step S102 is specifically as follows:

[0092] When the first sub-area orientation relationship, the second sub-area orientation relationship, the third sub-area orientation relationship, the fourth sub-area orientation relationship, the fifth sub-area orientation relationship and the sixth sub-area orientation relationship meet the theoretical organ position relationship, determine whether multiple foot organ reflex zones and multiple foot part reflex zones meet the theoretical position relationship.

[0093] The embodiment of the present invention verifies the recognition accuracy of each group of organ reflex sub-areas identified based on the theoretical organ-official position relationship, thereby ensuring the accuracy of the foot organ reflex area when executing step S102, thereby improving the accuracy of the target foot reflex area finally obtained.

[0094] Specifically, the theoretical position relationship is: the position relationship of the first group of organ reflex sub-areas from left to right is: ear reflex sub-area, eye reflex sub-area, brain reflex sub-area, left side of nasal reflex sub-area, oral reflex sub-area is below the nasal reflex sub-area, frontal sinus reflex sub-area is above the ear reflex sub-area, eye reflex sub-area, brain reflex sub-area, left side of nasal reflex sub-area; the position relationship of the second group of organ reflex sub-areas is: from top to bottom, including cervical reflex sub-area, esophageal tracheal reflex sub-area, thyroid reflex sub-area and parathyroid reflex sub-area; the position relationship of the third group of organ reflex sub-areas is: the trapezius reflex sub-area is located on the right side of the shoulder reflex sub-area; the position relationship of the fourth group of organ reflex sub-areas The bronchial reflex sub-area is located on the right side of the lung reflex sub-area; the orientation relationship of the fifth group of organ reflex sub-areas is as follows: the kidney reflex sub-area is located below the midpoint of the line connecting the liver reflex sub-area and the stomach reflex sub-area, the duodenum reflex sub-area is located below the stomach reflex sub-area, the transverse colon reflex sub-area is located below the kidney reflex sub-area, the small intestine reflex sub-area is located below the transverse colon reflex sub-area, the ascending colon reflex sub-area is located on the left side of the small intestine reflex sub-area, the bladder reflex sub-area is located below the duodenum reflex sub-area and on the right side of the small intestine reflex sub-area; the orientation relationship of the sixth group of organ reflex sub-areas is as follows: the gonad reflex sub-area is located below the pelvic reflex sub-area, and the buttocks reflex sub-area is located on the left side of the gonad reflex sub-area.

[0095] In some embodiments of the present invention, Figure 6 As shown in , the multi-task deep learning network model includes a shared backbone sub-model and a parallel organ reflexology detection head and a part reflexology detection head; Figure 7 As shown, step S101 includes:

[0096] S701, extracting features of the foot image to be identified based on the shared backbone sub-model to obtain shared image features;

[0097] S702, performing key point detection on the shared image features based on the organ reflex zone detection head to obtain the foot organ reflex zone;

[0098] S703 , performing semantic segmentation on the shared image features based on the part reflex zone detection head to obtain the part reflex zones of the foot.

[0099] The embodiment of the present invention sets a shared backbone sub-model for the organ reflex zone detection head and the part reflex zone detection head of the multi-task deep learning network model, so that the shared image features determined by the shared backbone sub-model can be used as the input of the organ reflex zone detection head and the part reflex zone detection head at the same time, further simplifying the model structure and improving the recognition efficiency of the part reflex zones of the foot and the organ reflex zones of the foot.

[0100] It should be noted that: when the foot image to be identified contains other objects besides the foot, in order to ensure the accuracy of identifying the foot part reflexology area and foot organ reflexology area, Figure 6 As shown, the multi-task deep learning network model also includes a target detection head, which performs target detection on the shared image features to obtain the foot contour. The part reflex zone detection head determines the foot part reflex zone based on the shared image features and the foot contour, and the organ reflex zone detection head determines the foot organ reflex zone based on the shared image features and the foot contour.

[0101] In a specific embodiment of the present invention, the multi-task deep learning network model is CenterNet, which can directly use the graph generated by the feature map without using the default bounding box to discretize the bounding box detection.

[0102] Among them, the part reflex area detection head is implemented by full convolution, and the organ reflex area detection head is implemented by regression offset.

[0103] In order to better implement the foot reflexology zone recognition method in the embodiment of the present invention, based on the foot reflexology zone recognition method, the embodiment of the present invention also provides a foot reflexology zone recognition device, such as Figure 8 As shown, the foot reflexology zone identification device 800 includes:

[0104] The foot reflexometer initial recognition unit 801 is used to input the foot image to be recognized into the multi-task deep learning network model to obtain the foot part reflex zone and the foot organ reflex zone; the foot part reflex zone includes the head reflex zone, the neck reflex zone, the shoulder and back reflex zone, the chest reflex zone, the abdomen reflex zone and the reproductive reflex zone;

[0105] The foot organ reflex zone determining unit 802 is used to determine whether the multiple foot organ reflex zones and the multiple foot part reflex zones satisfy the theoretical position relationship, and if so, determine the foot organ reflex zone as the target foot organ reflex zone;

[0106] Among them, the theoretical position relationship is: the foot organ reflex zone is a sub-area of ​​the reflex zone of each foot part.

[0107] The foot reflexology zone recognition device 800 provided in the above embodiment can implement the technical solution described in the above foot reflexology zone recognition method embodiment. The specific implementation principles of the above modules or units can refer to the corresponding contents in the above foot reflexology zone recognition method embodiment, which will not be repeated here.

[0108] like Fig. 9 As shown, the present invention also provides a rehabilitation robot 900 . The rehabilitation robot 900 includes a processor 901 , a memory 902 , and a display 903 . Fig. 9 Only some of the components of the rehabilitation 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.

[0109] In some embodiments, the memory 902 may be an internal storage unit of the rehabilitation robot 900, such as a hard disk or memory of the rehabilitation robot 900. In other embodiments, the memory 902 may also be an external storage device of the rehabilitation robot 900, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc., equipped on the rehabilitation robot 900.

[0110] Furthermore, the memory 902 may include both an internal storage unit of the rehabilitation robot 900 and an external storage device. The memory 902 is used to store application software installed in the rehabilitation robot 900 and various data.

[0111] In some embodiments, the processor 901 may be a central processing unit (CPU), a microprocessor or other data processing chip, and is used to run program codes or process data stored in the memory 902, such as the foot reflexology zone identification method of the present invention.

[0112] In some embodiments, the display 903 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 903 is used to display information of the rehabilitation robot 900 and to display a visual user interface. The components 901-903 of the rehabilitation robot 900 communicate with each other through a system bus.

[0113] In some embodiments of the present invention, when the processor 901 executes the foot reflexology zone recognition program in the memory 902, the following steps may be implemented:

[0114] Input the foot image to be identified into the multi-task deep learning network model to obtain the reflex areas of the foot parts and the reflex areas of the foot organs; the reflex areas of the foot parts include the reflex areas of the head, neck, shoulders and back, chest, abdomen and genitals;

[0115] Determine whether the multiple foot organ reflexology zones and the multiple foot part reflexology zones satisfy the theoretical position relationship, and if so, determine the foot organ reflexology zone as the target foot organ reflexology zone;

[0116] Among them, the theoretical position relationship is: the foot organ reflex zone is a sub-area of ​​the reflex zone of each foot part.

[0117] It should be understood that: when the processor 901 executes the foot reflexology zone recognition program in the memory 902, in addition to the above functions, other functions may also be implemented. For details, please refer to the description of the above related method embodiments.

[0118] 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 foot reflex zone identification method steps or functions provided in the above-mentioned method embodiments can be implemented.

[0119] 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.

[0120] The foot reflex zone identification method, device, rehabilitation robot and storage medium provided by the present invention are introduced in detail above. 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 foot reflexology areas, It is characterized in that include: Inputting the foot image to be identified into the multi-task deep learning network model to obtain multiple foot part reflexology zones and multiple foot organ reflexology zones; the foot part reflexology zones include the head reflexology zone, the neck reflexology zone, the shoulder and back reflexology zone, the chest reflexology zone, the abdomen reflexology zone and the reproductive reflexology zone; Determine whether the plurality of foot organ reflexology zones and the plurality of foot part reflexology zones satisfy a theoretical position relationship, and if so, determine the foot organ reflexology zone as a target foot organ reflexology zone; Wherein, the theoretical position relationship is: the foot organ reflexology area is a sub-area of ​​each foot part reflexology area; The foot organ reflexology zone includes a first group of organ reflexology sub-zones corresponding to the head reflexology zone, a second group of organ reflexology sub-zones corresponding to the neck reflexology zone, a third group of organ reflexology sub-zones corresponding to the shoulder and back reflexology zone, a fourth group of organ reflexology sub-zones corresponding to the chest reflexology zone, a fifth group of organ reflexology sub-zones corresponding to the abdomen reflexology zone, and a sixth group of organ reflexology sub-zones corresponding to the genital reflexology zone; Before determining whether the plurality of foot organ reflexology zones and the plurality of foot part reflexology zones satisfy the theoretical position relationship, the method further includes: Determine the first sub-area orientation relationship of the first group of organ reflection sub-areas, the second sub-area orientation relationship of the second group of organ reflection sub-areas, the third sub-area orientation relationship of the third group of organ reflection sub-areas, the fourth sub-area orientation relationship of the fourth group of organ reflection sub-areas, the fifth sub-area orientation relationship of the fifth group of organ reflection sub-areas, and the sixth sub-area orientation relationship of the sixth group of organ reflection sub-areas; Determine whether the first sub-area orientation relationship, the second sub-area orientation relationship, the third sub-area orientation relationship, the fourth sub-area orientation relationship, the fifth sub-area orientation relationship, and the sixth sub-area orientation relationship satisfy a theoretical device orientation relationship; Then, the step of judging whether the plurality of foot organ reflexology zones and the plurality of foot part reflexology zones satisfy the theoretical position relationship is as follows: When the first sub-area orientation relationship, the second sub-area orientation relationship, the third sub-area orientation relationship, the fourth sub-area orientation relationship, the fifth sub-area orientation relationship and the sixth sub-area orientation relationship satisfy the theoretical organ position relationship, determine whether the multiple foot organ reflex zones and the multiple foot part reflex zones satisfy the theoretical position relationship.

2. The foot reflexology zone identification method according to claim 1, It is characterized in that The determining whether the plurality of foot organ reflexology zones and the plurality of foot part reflexology zones satisfy a theoretical position relationship comprises: Determine a first area coordinate range of each of the foot part reflex zones in a preset coordinate system and a second area coordinate range of each of the foot organ reflex zones in the preset coordinate system; Determining the actual positional relationship between the plurality of foot organ reflexology zones and the plurality of foot part reflexology zones based on the first area coordinate range and the second area coordinate range; When the actual position relationship is the same as the theoretical position relationship, the multiple foot organ reflexology zones and the multiple foot part reflexology zones satisfy the theoretical position relationship.

3. The foot reflexology zone identification method according to claim 1, It is characterized in that The determining whether the plurality of foot organ reflexology zones and the plurality of foot part reflexology zones satisfy a theoretical position relationship comprises: Determine in turn whether the first group of organ reflex sub-areas and the head reflex area, the second group of organ reflex sub-areas and the neck reflex area, the third group of organ reflex sub-areas and the shoulder and back reflex area, the fourth group of organ reflex sub-areas and the chest reflex area, the fifth group of organ reflex sub-areas and the abdomen reflex area, and the sixth group of organ reflex sub-areas and the reproductive reflex area satisfy the theoretical position relationship.

4. The foot reflexology zone identification method according to claim 1, It is characterized in that Before determining whether the plurality of foot organ reflexology zones and the plurality of foot part reflexology zones satisfy the theoretical position relationship, the method further includes: Obtaining the theoretical position relationship and the actual position relationship between the reflex zones of the foot parts; Determining whether the actual position relationship satisfies the theoretical position relationship; Then, the step of judging whether the plurality of foot organ reflexology zones and the plurality of foot part reflexology zones satisfy the theoretical position relationship is as follows: When the actual part orientation relationship satisfies the theoretical part orientation relationship, it is determined whether the plurality of foot organ reflexology areas and the plurality of foot part reflexology areas satisfy the theoretical position relationship.

5. The foot reflexology zone identification method according to claim 3, It is characterized in that The first group of organ reflex sub-areas includes brain reflex sub-areas, nasal reflex sub-areas, oral reflex sub-areas, eye reflex sub-areas, ear reflex sub-areas and frontal sinus reflex sub-areas; the second group of organ reflex sub-areas includes esophageal tracheal reflex sub-areas, neck reflex sub-areas, thyroid reflex sub-areas and parathyroid reflex sub-areas; the third group of organ reflex sub-areas includes trapezius reflex sub-areas and shoulder reflex sub-areas; the fourth group of organ reflex sub-areas includes bronchial reflex sub-areas and lung reflex sub-areas; the fifth group of organ reflex sub-areas includes liver reflex sub-areas, kidney reflex sub-areas, stomach reflex sub-areas, duodenum reflex sub-areas, transverse colon reflex sub-areas, ascending colon reflex sub-areas, small intestine reflex sub-areas and bladder reflex sub-areas; the sixth group of organ reflex sub-areas includes pelvic reflex sub-areas, buttocks reflex sub-areas and gonad reflex sub-areas.

6. The foot reflexology zone identification method according to claim 1, It is characterized in that The multi-task deep learning network model includes a shared backbone sub-model and a parallel organ reflexology detection head and a part reflexology detection head; the foot image to be identified is input into the multi-task deep learning network model to obtain the foot part reflexology and foot organ reflexology, including: Extracting features of the to-be-recognized foot image based on the shared backbone sub-model to obtain shared image features; Performing key point detection on the shared image features based on the organ reflex zone detection head to obtain the foot organ reflex zone; The shared image features are semantically segmented based on the part reflexology detection head to obtain the part reflexology of the foot.

7. A foot reflexology zone identification device, It is characterized in that include: A foot reflexometer initial recognition unit is used to input the foot image to be recognized into a multi-task deep learning network model to obtain multiple foot part reflex zones and multiple foot organ reflex zones; the foot part reflex zones include the head reflex zone, the neck reflex zone, the shoulder and back reflex zone, the chest reflex zone, the abdomen reflex zone and the reproductive reflex zone; a foot organ reflex zone determination unit, configured to determine whether the plurality of foot organ reflex zones and the plurality of foot part reflex zones satisfy a theoretical position relationship, and if so, determine the foot organ reflex zone as a target foot organ reflex zone; Wherein, the theoretical position relationship is: the foot organ reflexology area is a sub-area of ​​each foot part reflexology area; The foot organ reflexology zone includes a first group of organ reflexology sub-zones corresponding to the head reflexology zone, a second group of organ reflexology sub-zones corresponding to the neck reflexology zone, a third group of organ reflexology sub-zones corresponding to the shoulder and back reflexology zone, a fourth group of organ reflexology sub-zones corresponding to the chest reflexology zone, a fifth group of organ reflexology sub-zones corresponding to the abdomen reflexology zone, and a sixth group of organ reflexology sub-zones corresponding to the genital reflexology zone; Before determining whether the plurality of foot organ reflexology zones and the plurality of foot part reflexology zones satisfy the theoretical position relationship, the method further includes: Determine the first sub-area orientation relationship of the first group of organ reflection sub-areas, the second sub-area orientation relationship of the second group of organ reflection sub-areas, the third sub-area orientation relationship of the third group of organ reflection sub-areas, the fourth sub-area orientation relationship of the fourth group of organ reflection sub-areas, the fifth sub-area orientation relationship of the fifth group of organ reflection sub-areas, and the sixth sub-area orientation relationship of the sixth group of organ reflection sub-areas; Determine whether the first sub-area orientation relationship, the second sub-area orientation relationship, the third sub-area orientation relationship, the fourth sub-area orientation relationship, the fifth sub-area orientation relationship, and the sixth sub-area orientation relationship satisfy a theoretical device orientation relationship; Then, the step of judging whether the plurality of foot organ reflexology zones and the plurality of foot part reflexology zones satisfy the theoretical position relationship is as follows: When the first sub-area orientation relationship, the second sub-area orientation relationship, the third sub-area orientation relationship, the fourth sub-area orientation relationship, the fifth sub-area orientation relationship and the sixth sub-area orientation relationship satisfy the theoretical organ position relationship, determine whether the multiple foot organ reflex zones and the multiple foot part reflex zones satisfy the theoretical position relationship.

8. A rehabilitation robot, It is characterized in that comprising 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 foot reflexology zone identification method as described in any one of claims 1 to 6.

9. A computer-readable storage medium, It is 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 foot reflexology zone identification method according to any one of claims 1 to 6 are implemented.

Images