Face acupoint space positioning method and system applied to intelligent robot

Abstract

The invention relates to the technical field of machine recognition, and discloses a face acupuncture point space positioning method and system applied to an intelligent robot, and the method comprises the steps: obtaining a face image collected by a depth camera; recognizing all key point locations of the face based on the face image, and generating a key point location map; wherein each key point location on the key point location map comprises a corresponding three-dimensional point location coordinate; mapping the target acupuncture point to the key point position map to obtain a face acupuncture point map; updating the face acupoint atlas; based on the updated face acupuncture point map, the acupuncture point normal of the target acupuncture point is determined, and the acupuncture point normal is used for guiding the intelligent robot to conduct directional operation. According to the method, the target acupuncture points of the face can be accurately recognized and positioned, the accurate operation direction can be provided through normal calculation, and the method has high positioning accuracy and operability.

Description

Technical Field

[0001] The present invention relates to the field of machine recognition technology, and in particular to a facial acupuncture point spatial positioning method and system applied to an intelligent robot. Background Art

[0002] Acupoints are specific locations along the body's meridians. Currently, there are over 300 identified acupoints on the human body. Traditional Chinese Medicine practitioners can treat these ailments through acupuncture, massage, acupressure, and moxibustion. The selection of acupoints and the effectiveness of treatments depend heavily on the practitioner's experience and understanding of TCM theory.

[0003] For Traditional Chinese Medicine practitioners, due to individual differences, the accuracy of acupoint selection and the appropriateness of acupuncture techniques are crucial indicators of their proficiency. Manual acupoint selection is subject to subjectivity and uncertainty, and the lack of standardization of TCM techniques results in poor reproducibility in acupuncture experiments. Against this backdrop, automated and standardized acupoint localization has become a key research direction within the medical community both domestically and internationally in recent years.

[0004] Currently, mainstream automated acupoint location methods fall into two main categories: human acupoint recognition based on feature point matching algorithms and computer vision-based acupoint recognition. Feature point matching can be considered an image matching problem, which is highly dependent on image quality. Significant discrepancies between two matching images can lead to feature point mismatches, resulting in acupoint location failure. Computer vision-based acupoint recognition, on the other hand, requires capturing facial images from various angles and manually annotating facial acupoints to provide a large amount of accurate data for establishing and training deep learning-based acupoint recognition models. Furthermore, binocular stereo imaging technology is required to provide three-dimensional spatial information from the images for the deep learning-based acupoint recognition models. While this method offers high acupoint recognition accuracy and is simple to use, it is highly dependent on model accuracy, requiring a large amount of training data and resulting in high training costs. Furthermore, the current lack of data related to facial acupoints significantly reduces the recognition performance of deep learning-based acupoint recognition models. Summary of the Invention

[0005] In view of this, the present invention provides a facial acupoint spatial positioning method and system applied to an intelligent robot to solve the problem in the prior art that the recognition accuracy of facial acupoints is difficult to ensure.

[0006] In a first aspect, the present invention provides a method for spatially locating facial acupuncture points for an intelligent robot, the method comprising:

[0007] Get the facial image captured by the depth camera;

[0008] Identify all key points of the face based on the facial image and generate a key point map; wherein each key point on the key point map includes corresponding three-dimensional point coordinates;

[0009] Map the target acupuncture points to the key point map to obtain the facial acupuncture point map;

[0010] Updated facial acupuncture point map;

[0011] Based on the updated facial acupoint map, the acupoint normal of the target acupoint is determined, and the acupoint normal is used to guide the intelligent robot to perform directional operations.

[0012] The present invention specifically relates to a spatial positioning method for facial acupoint recognition. A depth vision camera can obtain the physical distance between the face in the scene and the camera, providing three-dimensional data for spatial positioning. Then, a facial recognition algorithm is used to identify key facial features, and the key points are mapped one-to-one with the desired target acupoints. Finally, the robot's operating direction is determined by the acupoint normal of the target acupoint. In this embodiment, the facial acupoint spatial positioning method based on depth vision and facial recognition algorithms can not only accurately identify and locate the target facial acupoints, but also provide accurate operating directions through normal calculation, with high positioning accuracy and operability.

[0013] In an optional embodiment, determining the acupoint normal of the target acupoint based on the facial acupoint atlas includes:

[0014] Determine a preset number of key points surrounding the target acupuncture point;

[0015] Multiple triangles are constructed with the target acupuncture points as common vertices and key points as vertices;

[0016] Determine the normal vector of each triangle, wherein each normal vector takes the target acupuncture point as a starting point;

[0017] Based on the normal vectors of all triangles, the acupoint normal of the target acupoint is determined.

[0018] In this embodiment, by analyzing the acupoint normal of the target acupoint, the operation direction of the robot can be accurately calculated to ensure the accuracy of the robot operation and thus ensure the treatment effect.

[0019] In an optional embodiment, updating the facial acupuncture point map includes:

[0020] Based on the preset occlusion algorithm, determine whether the target acupoint in the facial acupoint map is occluded;

[0021] and / or, based on a preset jitter anomaly algorithm, determining whether a target acupoint in the facial acupoint map has an abnormality;

[0022] In the case where the target acupoint in the facial acupoint map is blocked or abnormal, the target acupoint in the facial acupoint map is discarded.

[0023] In this implementation, a preset occlusion algorithm is used to determine whether a target acupoint is obscured and, if so, discard the acupoint. If a jitter anomaly is detected at a target acupoint, the algorithm uses this jitter anomaly algorithm to determine and discard the acupoint. This effectively eliminates inaccurate data caused by occlusion or anomalies, ensuring the accuracy of the acupoint atlas. Ultimately, only clear and stable acupoint information is retained, effectively improving the accuracy of the robot's acupoint recognition and manipulation.

[0024] In an optional embodiment, determining whether a target acupoint in the facial acupoint map is blocked includes:

[0025] Obtaining a first point, where the first point is a key point in the facial acupoint atlas where the target acupoint is mapped;

[0026] Determine two adjacent key points on the same side as the first point in the facial acupuncture point map;

[0027] Determine the vector product between the first point and two adjacent key points;

[0028] When the vector product is greater than 0, it is determined that the target acupuncture point corresponding to the first point is not blocked;

[0029] When the vector product is less than 0, it is determined that the target acupuncture point corresponding to the first point is blocked.

[0030] In this embodiment, the occlusion condition of the target acupoint is determined by vector multiplication, and the target acupoint is discarded if it is occluded, which can effectively ensure the accuracy of the robot's acupoint recognition and operation.

[0031] In an optional embodiment, determining whether a target acupoint in the facial acupoint map has an abnormality includes:

[0032] Determine the current two-dimensional coordinate information of the current key point in the facial acupoint map, wherein the current key point is the key point mapped to the target acupoint;

[0033] Compare the current two-dimensional coordinate information with the two-dimensional coordinate information of the previous frame to obtain a first distance;

[0034] Comparing the two-dimensional coordinate information of the previous frame with the two-dimensional coordinate information of the previous frame to obtain a second distance; wherein the key point corresponding to the current two-dimensional coordinate information is consistent with the key point corresponding to the two-dimensional coordinate information of the previous frame and the key point corresponding to the two-dimensional coordinate information of the previous frame;

[0035] When the difference between the first distance and the second distance is less than a preset condition, it is determined that the target acupuncture point corresponding to the current key point is normal;

[0036] When the difference between the first distance and the second distance is greater than or equal to a preset condition, it is determined that the target acupuncture point corresponding to the current key point is abnormal.

[0037] The vibration anomaly detection method provided in this embodiment can reduce the abnormal points where vibration occurs to a certain extent, thereby ensuring the accuracy of robot acupoint recognition and operation.

[0038] In an optional embodiment, mapping the target acupuncture points to the key point map includes:

[0039] Determine the reference position according to the positioning rules of the target acupuncture points on the face;

[0040] Determine the reference position of the reference part on the key point map;

[0041] Based on the reference position and positioning rules, the target position of the target acupuncture point on the key point map is determined, and the target position is used to map the target acupuncture point.

[0042] In this embodiment, mapping the target acupoints onto the key point atlas can provide high-precision spatial positioning and consistent standardized reference, which helps to achieve accurate acupoint identification and operation.

[0043] In a second aspect, the present invention provides a facial acupuncture point spatial positioning system for an intelligent robot, the system comprising:

[0044] The acquisition module is used to obtain the facial image captured by the depth camera;

[0045] A generation module, configured to identify all key points of the face based on the facial image and generate a key point map; wherein each key point on the key point map includes corresponding three-dimensional point coordinates;

[0046] A mapping module is used to map the target acupoints to the key point map to obtain the facial acupoint map;

[0047] Update module, used to update facial acupuncture point map;

[0048] The vector module is used to determine the acupoint normal of the target acupoint based on the updated facial acupoint map. The acupoint normal is used to guide the intelligent robot to perform directional operations.

[0049] In a third aspect, the present invention provides a computer device comprising: a memory and a processor, the memory and the processor being communicatively connected to each other, the memory storing computer instructions, and the processor executing the computer instructions to execute the facial acupoint spatial positioning method applied to an intelligent robot according to the above-mentioned first aspect or any corresponding embodiment thereof.

[0050] In a fourth aspect, the present invention provides a computer-readable storage medium storing computer instructions for enabling a computer to execute the method for spatial positioning of facial acupoints applied to an intelligent robot according to the first aspect or any corresponding embodiment thereof.

[0051] In a fifth aspect, the present invention provides a computer program product comprising computer instructions for causing a computer to execute the method for spatial positioning of facial acupuncture points applied to an intelligent robot according to the first aspect or any corresponding embodiment thereof.

[0052] It should be noted that the facial acupoint spatial positioning system, computer device, computer-readable storage medium, and computer program product provided by the present invention correspond to the aforementioned facial acupoint spatial positioning method for an intelligent robot. Therefore, for the beneficial effects of the facial acupoint spatial positioning system, computer device, computer-readable storage medium, and computer program product for an intelligent robot, please refer to the description of the corresponding beneficial effects of the facial acupoint spatial positioning method for an intelligent robot above, and will not be repeated here. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the specific embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0054] Figure 1 is a flow chart of a method for spatially locating facial acupuncture points applied to an intelligent robot according to an embodiment of the present invention;

[0055] Figure 2 is a schematic diagram of a process for generating a key point map according to an embodiment of the present invention;

[0056] Figure 3 is a schematic diagram of a facial acupuncture point map after mapping according to an embodiment of the present invention;

[0057] Figure 4 is a front view of acupoint normal determination according to an embodiment of the present invention;

[0058] Figure 5 is a side view of acupoint normal determination according to an embodiment of the present invention;

[0059] Figure 6 is a schematic diagram of acupoint occlusion determination according to an embodiment of the present invention;

[0060] Figure 7 is a structural block diagram of a facial acupoint spatial positioning system applied to an intelligent robot according to an embodiment of the present invention;

[0061] Figure 8 Schematic diagram of the hardware structure of a computer device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0062] To make the purpose, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are 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 making creative efforts shall fall within the scope of protection of the present invention.

[0063] According to an embodiment of the present invention, an embodiment of a method for spatially locating facial acupuncture points applied to an intelligent robot is provided. It should be noted that the steps shown in the flowcharts of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowcharts, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0064] In this embodiment, a method for spatial positioning of facial acupuncture points applied to an intelligent robot is provided, which can be executed by an intelligent robot device such as an acupuncture robot or a massage robot including a control system. Figure 1 FIG. 1 is a flow chart of a method for spatially locating facial acupuncture points applied to an intelligent robot according to an embodiment of the present invention. Figure 1 As shown, the process includes the following steps:

[0065] Step S101: Acquire a facial image captured by a depth camera.

[0066] When the device boots up, it first uses its built-in depth vision camera to capture the user's frontal face from an appropriate distance in front of them. After capturing the frontal face image, real-time capture begins, continuously capturing images of the user's face. After each capture, the depth vision camera's internal parameters, the camera's left eye image, the camera's right eye image, and the camera's depth image are captured.

[0067] Step S102: Identify all key points of the face based on the facial image and generate a key point map; wherein each key point on the key point map includes corresponding three-dimensional point coordinates.

[0068] Specifically, refer to Figure 2 As shown, the MediaPipe facial recognition algorithm can be used to identify the left and right eye visual images separately, obtaining facial keypoint information in each of these images. This facial recognition algorithm can not only identify all keypoints in the frontal face, but also in the current facial image in real time. Because facial images are acquired using a depth vision camera, each keypoint not only contains 2D coordinate information, but also has its depth information determined based on the point cloud information. This means that each keypoint includes corresponding 3D point coordinates.

[0069] Step S103: Map the target acupoints to the key point map to obtain a facial acupoint map.

[0070] In this embodiment, when there is no occlusion in the facial image, Mediapipe identifies a total of 478 facial key points. Each key point is mapped to the desired target acupoint according to the acupoint positioning rules. In this embodiment, taking Jingming, Chengqi, Tongziliao, Taiyang, Sizhukong, and Zanzhu as examples, the mapped facial acupoint map is as follows: Figure 3 In addition, in addition to mapping the desired target acupoints, other acupoints can also be mapped at the same time.

[0071] Step S104: updating the facial acupoint map.

[0072] Since the 478 key points identified above represent all the key points of the face calculated by artificial intelligence, and in actual applications, the user's face may rotate and some key points may be occluded, a judgment should be made based on the actual situation. Specifically, based on the mapped facial acupoint atlas, the desired target acupoint is selected from the key points, and an occlusion algorithm is used to determine whether the target acupoint is occluded. If so, the target acupoint is discarded and calculation continues for other target acupoints. Furthermore, a jitter anomaly algorithm can be used to determine whether the target acupoint is abnormal. If so, the target acupoint is discarded and calculation continues for other target acupoints. After the judgment is completed, an updated facial acupoint atlas is obtained.

[0073] Step S105: Determine the acupoint normal of the target acupoint based on the updated facial acupoint map. The acupoint normal is used to guide the intelligent robot to perform directional operation.

[0074] The updated facial acupoint map only includes unobstructed points. This part of the points is a layer of point cloud on the surface of the human face that is not obstructed. It lacks information about the inner surface pointing to the outer surface. For a certain target acupoint, the robot can operate from all directions of the target acupoint, which is not suitable for practical applications. In practical applications, since the user's face will rotate, some key points will be obstructed, and it is difficult for the intelligent robot to accurately perform the corresponding operation. To this end, the acupoint normal of the unobstructed target acupoint can be calculated, and the acupoint normal can be used to determine which direction of the vertex is the outer surface of the human face. In the present invention, the direction of the acupoint normal is the direction from the inner surface to the outer surface. In practical applications, the negative direction of the acupoint normal can be used as the operation direction of acupuncture, acupoint massage, etc. By determining the acupoint normal of the target acupoint and combining it with the three-dimensional coordinates of the target acupoint, the spatial positioning of the target acupoint on the face can be achieved.

[0075] This embodiment specifically relates to a spatial positioning method for facial acupoint recognition. A depth vision camera can be used to obtain the physical distance between the face in a scene and the camera, providing three-dimensional data for spatial positioning. A facial recognition algorithm is then used to identify key facial features and map these key features to the desired target acupoints. Finally, the robot's operating direction is determined by the acupoint normal of the target acupoint. In this embodiment, the facial acupoint spatial positioning method based on depth vision and a facial recognition algorithm can not only accurately identify and locate target facial acupoints, but also provide accurate operating directions through normal calculation, resulting in high positioning accuracy and operability.

[0076] In addition, the present invention combines depth cameras and machine vision, and can realize real-time recognition of facial acupoints based on the captured video stream using artificial intelligence, thereby automatically obtaining the spatial position of facial acupoints. This method is not limited by the image quality of feature matching, and does not require separate training of the corresponding facial acupoint recognition model, which can effectively reduce development costs.

[0077] In some optional embodiments, determining the acupoint normal of the target acupoint based on the facial acupoint atlas includes:

[0078] Determine a preset number of key points that surround the target acupuncture point.

[0079] Multiple triangles are constructed with the target acupuncture points as common vertices and the key points as vertices.

[0080] The normal vector of each triangle is determined, wherein each normal vector takes the target acupuncture point as a starting point.

[0081] Based on the normal vectors of all triangles, the acupoint normal of the target acupoint is determined.

[0082] The following is an example, refer to Figure 4 and Figure 5 As shown, Figure 4 For the front view, Figure 5 It is a side view. The three-dimensional spatial coordinates of the target acupuncture point can be calculated by the two-dimensional coordinates of the target acupuncture point and the point cloud coordinate set obtained by the depth camera, and then the key points around the target acupuncture point are obtained according to the three-dimensional spatial coordinates of the target acupuncture point. For example, the target acupuncture point is O, and the surrounding key points are determined to include five key points A, B, C, D, and E, which are then surrounded by triangles OAB, OBC, OCD, ODE, and OEA respectively. For triangle OAB, take vector OA and vector OB, and cross-product vector OA with vector OB to obtain the normal vector OA' of triangle OAB. According to the above method, the normal vectors OB', OC', OD', and OE' of triangles OBC, OCD, ODE, and OEA are obtained respectively. Since the size of each triangle has different effects on its contribution to the acupuncture point normal, the surface area of ​​each triangle can be used as the weight value of the acupuncture point normal contribution. The weight value of small triangles is small, and the weight value of large triangles is large. Because the cross product of two vectors is itself part of the triangle area calculation, they can be added directly without normalization, effectively weighting the triangle area. The normal vector OO' of target acupoint O is then calculated as OA' + OB' + OC' + OD' + OE'. After further normalization, this becomes the normal vector of target acupoint O, or the target acupoint's acupoint normal.

[0083] An acupoint normal is a vector perpendicular to the facial surface. In this embodiment, the outward direction from the facial surface is assumed to be positive. The acupoint normal refers to a vector line extending outward from a target acupoint on the face. In practical applications, the negative direction of the acupoint normal can be used as the operating direction for acupuncture needle insertion, acupoint massage, and other procedures.

[0084] In this embodiment, by analyzing the acupoint normal of the target acupoint, the operation direction of the robot can be accurately calculated to ensure the accuracy of the robot operation and thus ensure the treatment effect.

[0085] In some optional implementations, updating the facial acupoint map based on the current facial image includes:

[0086] Based on the preset occlusion algorithm, determine whether the target acupuncture point in the current facial image is occluded.

[0087] And / or, based on a preset jitter anomaly algorithm, determining whether there is an abnormality in the target acupuncture point in the current facial image.

[0088] In the case that the target acupoint in the current facial image is blocked or abnormal, the target acupoint in the facial acupoint atlas is discarded.

[0089] In this embodiment, the occlusion algorithm and the jitter anomaly algorithm may be performed simultaneously, or the occlusion algorithm may be performed first and then the jitter anomaly algorithm, or only the occlusion algorithm or only the jitter anomaly algorithm may be performed.

[0090] In this embodiment, a preset occlusion algorithm is used to determine whether a target acupoint is obscured and, if so, discard the acupoint. If a jitter anomaly is detected at a target acupoint, the acupoint is discarded based on the preset jitter anomaly algorithm. This effectively eliminates inaccurate data caused by occlusion or anomalies, ensuring the accuracy of the acupoint atlas. Ultimately, only clear and stable acupoint information is retained, effectively improving the accuracy of the robot's acupoint recognition and manipulation.

[0091] In some optional implementations, determining whether a target acupuncture point in the current facial image is blocked includes:

[0092] Obtaining a first point, where the first point is a key point in the facial acupoint atlas where the target acupoint is mapped;

[0093] Determine two adjacent key points on the same side as the first point in the facial acupuncture point map;

[0094] Based on the current facial image, determining a current position of the first point in the current facial image, and using a point corresponding to the current position as the second point;

[0095] Determine the vector product between the second point and two adjacent key points;

[0096] When the vector product is greater than 0, it is determined that the target acupuncture point corresponding to the first point is not blocked;

[0097] When the vector product is less than 0, it is determined that the target acupuncture point corresponding to the first point is blocked.

[0098] The following is an example, refer to Figure 6 As shown in the figure. After obtaining the keypoint map based on the face recognition algorithm, if the occlusion status of target point c on the right face is determined, the adjacent points a and b to the left of target point c in the keypoint map are selected. Normally, target point c is to the right of adjacent points a and b, forming vectors ac and bc. If the cross product of vectors ac and bc is greater than 0, point c is not occluded.

[0099] When the face turns left, the target point c shifts to point c', which is to the left of adjacent points a and b, forming vectors ac' and bc'. If the cross product of ac' and bc' is less than 0, point c' is occluded and is discarded.

[0100] Similarly, we can determine the occlusion status of the key points on the left face.

[0101] In this embodiment, the occlusion condition of the target acupoint is determined by vector multiplication, and the target acupoint is discarded if it is occluded, which can effectively ensure the accuracy of the robot's acupoint recognition and operation.

[0102] In some optional implementations, determining whether there is an abnormality in the target acupuncture point in the current facial image includes:

[0103] Determine the current two-dimensional coordinate information of the current key point in the current facial image, wherein the current key point is the key point mapped with the target acupuncture point;

[0104] Compare the current two-dimensional coordinate information with the two-dimensional coordinate information of the previous frame to obtain a first distance;

[0105] Comparing the two-dimensional coordinate information of the previous frame with the two-dimensional coordinate information of the previous frame to obtain a second distance; wherein the key point corresponding to the current two-dimensional coordinate information is consistent with the key point corresponding to the two-dimensional coordinate information of the previous frame and the key point corresponding to the two-dimensional coordinate information of the previous frame;

[0106] When the difference between the first distance and the second distance is less than a preset condition, it is determined that the target acupuncture point corresponding to the current key point is normal;

[0107] When the difference between the first distance and the second distance is greater than or equal to a preset condition, it is determined that the target acupuncture point corresponding to the current key point is abnormal.

[0108] Due to the real-time acquisition of facial images, face recognition methods using continuous video frames are prone to fluctuations in recognition results. This can manifest as key points in the recognition results not matching the actual facial points, or even appearing to flicker back and forth. Therefore, this embodiment uses a jitter anomaly algorithm to discard abnormal acupoints.

[0109] The following example illustrates that the frame step size of this algorithm can be appropriately increased or decreased depending on the degree of fluctuation in the recognition algorithm results. A step size of 1 selects 123 frames, a step size of 2 selects 135 frames, and so on. In this embodiment, the frame step size selected is 1, which means that the current frame, the previous frame, and the previous previous frame are selected for comparison.

[0110] Key points are acquired in real time using the face recognition algorithm. The occlusion status is determined using the occlusion point rejection algorithm. The 2D coordinates of the current point are obtained and stored at the end of the queue. The current 2D coordinates (x2, y2) are compared with the 2D coordinates (x1, y1) of the previous frame at the end of the queue. The first distance is calculated using the distance formula: L1 = √[(x1-x2)² + (y1-t2)²]. Additionally, the 2D coordinates of the current point in the previous frame (x1, y1) are compared with the 2D coordinates of the previous frame (x0, y0), and the second distance is calculated using the distance formula: L0 = √[(x1-x0)² + (y1-t0)²]. When both distances L0 and L1 are less than 1% of the image size (the minimum of the image length and width), the current point is considered normal; otherwise, it is an abnormal value.

[0111] The vibration abnormality detection method provided in this embodiment can reduce the abnormal points where vibration occurs to a certain extent, thereby ensuring the accuracy of robot acupoint recognition and operation.

[0112] In some optional embodiments, mapping the target acupuncture points to the key point map includes:

[0113] The reference location is determined based on the positioning rules of the target acupoint on the face. For example, the Chengqi point is located directly below the pupil and at the lower edge of the eye socket. Therefore, the reference location for the Chengqi point is the pupil or the eye socket.

[0114] Determine the reference position of the reference part on the key point map. Still taking the above example, the position of the pupil on the key point map can be further determined and used as the reference position.

[0115] Based on the reference position and positioning rules, the target position of the target acupuncture point on the key point map is determined, and the target position is used to map the target acupuncture point.

[0116] In this embodiment, mapping the target acupoints onto the key point map can provide high-precision spatial positioning and consistent standardized reference, which helps to achieve accurate acupoint identification and operation.

[0117] This embodiment also provides a facial acupuncture point spatial positioning system for an intelligent robot. This system is used to implement the above-mentioned embodiments and preferred embodiments, and details already described will not be repeated. As used below, the term "module" may refer to a combination of software and / or hardware that implements a predetermined function. Although the systems described in the following embodiments are preferably implemented in software, implementation using hardware, or a combination of software and hardware, is also possible and contemplated.

[0118] This embodiment provides a facial acupuncture point spatial positioning system for an intelligent robot. Figure 7 As shown, the system includes:

[0119] Acquisition module 201, used to acquire a facial image captured by a depth camera;

[0120] A generating module 202 is configured to identify all key points of the face based on the facial image and generate a key point map, wherein each key point on the key point map includes corresponding three-dimensional point coordinates;

[0121] A mapping module 203 is used to map the target acupoints onto a key point map to obtain a facial acupoint map;

[0122] An updating module 204 is used to update the facial acupuncture point atlas;

[0123] The vector module 205 is used to determine the acupoint normal of the target acupoint based on the updated facial acupoint map. The acupoint normal is used to guide the intelligent robot to perform directional operations.

[0124] In some optional implementations, the vector module 205 includes:

[0125] A vector unit is used to determine a preset number of key points surrounding the target acupuncture point; construct multiple triangles with the target acupuncture point as a common vertex and the key points as vertices; determine the normal vector of each triangle, wherein each normal vector starts from the target acupuncture point; and determine the acupuncture point normal of the target acupuncture point based on the normal vectors of all triangles.

[0126] In an optional implementation, the updating module 204 includes:

[0127] An updating unit is configured to determine, based on a preset occlusion algorithm, whether a target acupoint in the facial acupoint map is occluded; and / or, based on a preset jitter anomaly algorithm, to determine whether a target acupoint in the facial acupoint map is abnormal; and to discard the target acupoint in the facial acupoint map if the target acupoint is occluded or abnormal.

[0128] Among them, judging whether the target acupoint in the facial acupoint map is blocked includes: obtaining a first point, which is one of the key points in the facial acupoint map where the target acupoint is mapped; determining two adjacent key points on the same side of the first point in the facial acupoint map; determining the vector product between the first point and the two adjacent key points; when the vector product is greater than 0, determining that the target acupoint corresponding to the first point is not blocked; when the vector product is less than 0, determining that the target acupoint corresponding to the first point is blocked. Determining whether there is an abnormality in a target acupoint in a facial acupoint map includes: determining current two-dimensional coordinate information corresponding one-to-one to a current key point in the facial acupoint map, wherein the current key point is a key point mapped with a target acupoint; comparing the current two-dimensional coordinate information with the two-dimensional coordinate information of a previous frame to obtain a first distance; comparing the two-dimensional coordinate information of the previous frame with the two-dimensional coordinate information of the previous frame to obtain a second distance; wherein, the key point corresponding to the current two-dimensional coordinate information is consistent with the key point corresponding to the two-dimensional coordinate information of the previous frame and the key point corresponding to the two-dimensional coordinate information of the previous frame; when the difference between the first distance and the second distance is less than a preset condition, determining that the target acupoint corresponding to the current key point is normal; when the difference between the first distance and the second distance is greater than or equal to the preset condition, determining that the target acupoint corresponding to the current key point is abnormal.

[0129] In an optional implementation, the mapping module 203 includes:

[0130] The mapping unit is used to determine the reference part according to the positioning rules of the target acupuncture point on the face; determine the reference position of the reference part on the key point map; based on the reference position and the positioning rules, determine the target position of the target acupuncture point on the key point map, and the target position is used to map the target acupuncture point.

[0131] The facial acupoint spatial positioning system applied to the intelligent robot in this embodiment is presented in the form of functional units, where the units refer to ASIC circuits, processors and memories that execute one or more software or fixed programs, and / or other devices that can provide the above functions.

[0132] The further functional description of each of the above modules and units is the same as that of the above corresponding embodiments and will not be repeated here.

[0133] The embodiment of the present invention also provides a computer device having the above Figure 7 The facial acupuncture point spatial positioning system applied to intelligent robots is shown.

[0134] See also Figure 8 , Figure 8 is a structural diagram of a computer device provided by an optional embodiment of the present invention, such as Figure 8As shown, the computer device includes: one or more processors 10, memory 20, and interfaces for connecting various components, including high-speed interfaces and low-speed interfaces. Various components utilize different buses to communicate with each other and can be installed on a common mainboard or installed in other ways as needed. The processor can process the instructions executed in the computer device, including instructions stored in the memory or on the memory to display the graphical information of the GUI on an external input / output device (such as, a display device coupled to the interface). In some optional embodiments, if necessary, multiple processors and / or multiple buses can be used together with multiple memories and multiple memories. Equally, multiple computer devices can be connected, and each device provides part of the necessary operations (for example, as a server array, a group of blade servers, or a multi-processor system). Figure 8 A processor 10 is taken as an example.

[0135] The processor 10 may be a central processing unit, a network processor, or a combination thereof. The processor 10 may further include a hardware chip. The hardware chip may be an application-specific integrated circuit, a programmable logic device, or a combination thereof. The programmable logic device may be a complex programmable logic device, a field programmable gate array, a general purpose array logic, or any combination thereof.

[0136] The memory 20 stores instructions that can be executed by at least one processor 10, so that the at least one processor 10 executes the method shown in the above embodiment.

[0137] The memory 20 may include a program storage area and a data storage area, wherein the program storage area may store an operating system and application programs required for at least one function; the data storage area may store data created based on the use of the computer device, etc. In addition, the memory 20 may include a high-speed random access memory, and may also include a non-transient memory, such as at least one disk storage device, a flash memory device, or other non-transient solid-state storage device. In some optional embodiments, the memory 20 may optionally include a memory remotely located relative to the processor 10, and these remote memories may be connected to the computer device via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0138] The memory 20 may include a volatile memory, such as a random access memory; the memory may also include a non-volatile memory, such as a flash memory, a hard disk or a solid-state drive; the memory 20 may also include a combination of the above types of memory.

[0139] The computer device further includes a communication interface 30 for the computer device to communicate with other devices or a communication network.

[0140] The embodiment of the present invention also provides a computer-readable storage medium. The above-mentioned method according to the embodiment of the present invention can be implemented in hardware, firmware, or implemented as a computer code that can be recorded in a storage medium, or implemented as a computer code that is originally stored in a remote storage medium or a non-temporary machine-readable storage medium and downloaded through a network and will be stored in a local storage medium, so that the method described herein can be stored in such software processing on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. Among them, the storage medium can be a magnetic disk, an optical disk, a read-only storage memory, a random access memory, a flash memory, a hard disk or a solid-state drive, etc.; further, the storage medium can also include a combination of the above-mentioned types of memory. It can be understood that a computer, a processor, a microprocessor controller or programmable hardware includes a storage component that can store or receive software or computer code. When the software or computer code is accessed and executed by a computer, a processor or hardware, the method shown in the above embodiment is implemented.

[0141] A portion of the present invention may be applied as a computer program product, such as a computer program instruction, which, when executed by a computer, can call or provide the method and / or technical solution according to the present invention through the operation of the computer. Those skilled in the art should understand that the form in which the computer program instruction exists in a computer-readable medium includes, but is not limited to, a source file, an executable file, an installation package file, etc. Accordingly, the way in which the computer program instruction is executed by the computer includes, but is not limited to: the computer directly executes the instruction, or the computer compiles the instruction and then executes the corresponding compiled program, or the computer reads and executes the instruction, or the computer reads and installs the instruction and then executes the corresponding installed program. Here, the computer-readable medium may be any available computer-readable storage medium or communication medium that can be accessed by the computer.

[0142] Although the embodiments of the present invention have been described with reference to the accompanying drawings, those skilled in the art may make various modifications and variations without departing from the spirit and scope of the present invention. Such modifications and variations are all within the scope defined by the appended claims.

Claims

1. A facial acupuncture point spatial positioning method applied to an intelligent robot, characterized in that: The method comprises: Get the facial image captured by the depth camera; Based on the facial image, all key points of the face are identified to generate a key point map; wherein each key point on the key point map includes corresponding three-dimensional point coordinates; Mapping the target acupuncture points onto the key point map to obtain a facial acupuncture point map; Updating the facial acupuncture point atlas; Based on the updated facial acupoint atlas, the acupoint normal of the target acupoint is determined, and the acupoint normal is used to guide the intelligent robot to perform directional operation.

2. The method according to claim 1, characterized in that The step of determining the acupoint normal of the target acupoint based on the facial acupoint atlas comprises: Determine a preset number of key points that surround the target acupuncture point; Constructing a plurality of triangles with the target acupuncture points as common vertices and the key points as vertices; Determine a normal vector of each triangle, wherein each normal vector takes the target acupuncture point as a starting point; The acupoint normal of the target acupoint is determined based on the normal vectors of all triangles.

3. The method according to claim 1, characterized in that The updating of the facial acupuncture point atlas comprises: Based on a preset occlusion algorithm, determining whether the target acupoint in the facial acupoint map is occluded; and / or, based on a preset jitter abnormality algorithm, determining whether the target acupoint in the facial acupoint map has an abnormality; When the target acupoint in the facial acupoint map is blocked or abnormal, the target acupoint in the facial acupoint map is discarded.

4. The method according to claim 3, characterized in that The determining whether the target acupoint in the facial acupoint atlas is blocked includes: Acquire a first point, where the first point is one of the key points in the facial acupoint atlas where the target acupoint is mapped; Determine two adjacent key points on the same side of the first point in the facial acupuncture point map; Determine a vector product between the first point and two adjacent key points; When the vector product is greater than 0, it is determined that the target acupuncture point corresponding to the first point is not blocked; When the vector product is less than 0, it is determined that the target acupuncture point corresponding to the first point is blocked.

5. The method according to claim 3, characterized in that: The determining whether the target acupoint in the facial acupoint atlas is abnormal includes: Determine the current two-dimensional coordinate information of the current key point in the facial acupuncture point atlas, wherein the current key point is the key point mapped with the target acupuncture point; Compare the current two-dimensional coordinate information with the previous frame of two-dimensional coordinate information to obtain a first distance; Compare the previous frame of two-dimensional coordinate information with the previous frame of two-dimensional coordinate information to obtain a second distance; wherein the key point corresponding to the current two-dimensional coordinate information is consistent with the key point corresponding to the previous frame of two-dimensional coordinate information and the key point corresponding to the previous frame of two-dimensional coordinate information; When the difference between the first distance and the second distance is less than a preset condition, determining that the target acupuncture point corresponding to the current key point is normal; When the difference between the first distance and the second distance is greater than or equal to the preset condition, it is determined that the target acupuncture point corresponding to the current key point is abnormal.

6. The method according to claim 1, characterized in that Mapping the target acupuncture points onto the key point map includes: Determining a reference position according to the positioning rule of the target acupuncture point on the face; Determine the reference position of the reference part on the key point map; Based on the reference position and the positioning rule, the target position of the target acupuncture point on the key point map is determined, and the target position is used to map the target acupuncture point.

7. A facial acupuncture point spatial positioning system applied to an intelligent robot, characterized in that: The system comprises: An acquisition module, used to acquire a facial image captured by a depth camera; A generating module, configured to identify all key points of the face based on the facial image and generate a key point map; wherein each key point on the key point map includes corresponding three-dimensional point coordinates; A mapping module, used for mapping the target acupuncture points to the key point map to obtain a facial acupuncture point map; An updating module, used for updating the facial acupuncture point atlas; The vector module is used to determine the acupoint normal of the target acupoint based on the updated facial acupoint map, and the acupoint normal is used to guide the intelligent robot to perform directional operation.

8. A computer device, characterized in that: include: A memory and a processor, wherein the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the facial acupuncture point spatial positioning method for an intelligent robot as described in any one of claims 1 to 6 by executing the computer instructions.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable the computer to execute the facial acupuncture point spatial positioning method applied to an intelligent robot as described in any one of claims 1-6.

10. A computer program product, characterized in that The method comprises computer instructions, wherein the computer instructions are used to enable a computer to execute the facial acupuncture point spatial positioning method applied to an intelligent robot according to any one of claims 1 to 6.

Images