Facial acupuncture positioning guidance system

Abstract

The present application belongs to the technical field of acupuncture treatment equipment, and discloses a positioning guidance system for facial acupuncture. A positioning guidance system for facial acupuncture includes: a diagnosis and treatment information acquisition module, which acquires the patient's treatment prescription and obtains the facial acupuncture points that the patient needs to receive acupuncture; an image information acquisition module, which acquires the facial image of the patient before acupuncture; an acupuncture point positioning module, which compares the facial image of the patient before acupuncture with the facial image with acupuncture points marked in the database, and obtains the position information of the facial acupuncture points that the patient needs to receive acupuncture; an acupuncture point marking module, which acquires the position information of the facial acupuncture points that the patient needs to receive acupuncture, and then sequentially irradiates the rays that can generate light spots on the patient's face to the facial acupuncture points to guide acupuncture treatment. The technical solution provided by the present application ensures that the corresponding acupuncture points can be accurately received during acupuncture, thereby enhancing the treatment effect.

Description

Technical Field

[0001] The present application relates to the technical field of acupuncture treatment, and in particular to a positioning guidance system for facial acupuncture. Background Art

[0002] Peripheral facial paralysis, also known as facial neuritis or idiopathic facial nerve paralysis, is mostly caused by nonspecific inflammation of the facial nerve on one side, with the main symptoms being loss of forehead wrinkles on the affected side, shallowing of nasolabial grooves, crooked corners of the mouth, and other damage to the facial expression muscles.

[0003] Acupoint stimulation is a good treatment for facial paralysis. Acupoints are generally stimulated by acupuncture. In addition, acupoints can also be stimulated by electric shock. Whether it is acupuncture or electric shock, it is necessary to accurately find the location of the patient's facial acupoints.

[0004] In clinical practice, it is common for a senior doctor to prescribe acupoint stimulation prescriptions for patients, and then the lower-level doctors perform acupuncture treatment according to the acupoint stimulation prescriptions. Because the facial conditions of each patient are different, the lower-level doctors are limited by experience and cannot accurately find the corresponding acupoints for acupuncture, resulting in unsatisfactory treatment results. Summary of the invention

[0005] The content of this application is used to introduce concepts in a brief form, which will be described in detail in the detailed implementation section below. The content of this application is not intended to identify the key features or essential features of the technical solution claimed for protection, nor is it intended to limit the scope of the technical solution claimed for protection.

[0006] As a first aspect of the present application, in order to solve the technical problem of inaccurate acupuncture point locations, the present application provides a facial acupuncture positioning guidance system, comprising:

[0007] The diagnosis and treatment information acquisition module obtains the patient's treatment prescription and the facial acupuncture points that the patient needs to receive;

[0008] An image information acquisition module is used to acquire the facial image of the patient before acupuncture;

[0009] The acupuncture point positioning module compares the facial image of the patient before acupuncture with the facial image with acupuncture points marked in the face database to obtain the location information of the facial acupuncture points that need to be acupunctured on the patient;

[0010] The acupuncture point marking module obtains the location information of the facial acupuncture points where the patient needs to be treated with acupuncture, and then irradiates the facial acupuncture points with a wavelength of visible light frequency in sequence to guide the acupuncture treatment.

[0011] In order to solve the problem of limited efficacy due to inaccurate acupuncture point positioning in acupuncture treatment, this application uses the diagnosis and treatment information acquisition module to first obtain the acupuncture points that need to be acupunctured, and then uses the acupuncture point positioning module to accurately match the facial image containing the preset acupuncture point marks from the database, thereby clarifying the exact location of the relevant acupuncture points of the current patient. Subsequently, the acupuncture point marking module emits a laser to project an indicator light spot on the patient's face, thereby ensuring that the staff can accurately find the corresponding acupuncture points when acupuncturing the patient, thereby improving the treatment effect.

[0012] Current acupoint guidance methods are mostly limited to providing approximate acupoint diagrams or video guidance to help medical staff identify acupoint locations. However, in actual operation, the acupoint acupuncture sequence and specific techniques still need to be judged by medical staff, which often leads to inaccurate acupuncture and thus affects the treatment effect. To address this problem, this application proposes the following technical solutions:

[0013] The diagnosis and treatment information acquisition module includes:

[0014] The prescription information input unit is used to input the patient's treatment prescription, and read the acupuncture points to be acupunctured, the acupuncture sequence of the acupuncture points, and the interval between acupuncture treatments from the treatment prescription;

[0015] An input unit is used to adjust the acupuncture points to be acupunctured, the acupuncture order of the acupuncture points, and the interval between acupuncture points;

[0016] The acupoint marking module marks the corresponding acupoints with light spots in sequence according to the acupuncture sequence and the interval between acupuncture.

[0017] The technical solution proposed in this application can automatically obtain the acupuncture points required for acupuncture, the acupuncture order and the interval time of each acupuncture point according to the preset treatment prescription before treatment. In addition, the solution is also provided with an input unit, allowing medical staff to flexibly adjust the above parameters according to actual conditions. Therefore, the subsequent acupuncture point marking module can accurately emit light spot marks according to the set sequence and time interval, guiding medical staff to perform acupuncture accurately and timely, reducing the impact of acupuncture point positioning errors, medical staff acupuncture errors and acupuncture time errors on the treatment effect.

[0018] Acupoint recognition essentially relies on determining the acupoints at corresponding distances based on the positions of the facial features. This process requires both accurate identification of facial features and accurate calculation of the distance between these features and the standard acupoint positions, such as the Yangbai acupoint being located 1 inch above the eyebrow. Current automatic acupoint recognition technology mostly relies on facial feature similarity matching. For example, the patient's facial image is compared with the Yangbai acupoint position marked in the database to find the area with the closest texture. However, this method overemphasizes the similarity of local features and ignores the differences in facial shapes and the diversity of facial textures between patients, resulting in large errors in directly locating acupoints by similarity. In response to the above problems, this application proposes the following technical solutions:

[0019] Furthermore, the image information acquisition module includes:

[0020] a camera for acquiring a facial image of the patient;

[0021] The marking line is attached straightly to the patient's face, and the length of the marking line is pre-recorded in the acupoint positioning module.

[0022] The technical solution proposed in this application adopts the method of setting marking lines on the patient's face to provide a length reference benchmark for the acquired image information. In this way, when identifying acupoints, the facial features can be used as stable reference points, combined with the precise scale information provided by the marking lines, to perform more accurate acupoint identification and positioning, thereby significantly improving the positioning accuracy of acupoints.

[0023] In order to further increase the positioning accuracy of acupoints, the present application provides the following technical solutions:

[0024] The acupoint positioning module includes:

[0025] A face shape definition unit, which predefines several standard face shapes of facial categories;

[0026] The image information storage unit is configured with a plurality of face shape databases, each face shape database corresponds to a standard face shape, and each face shape database stores standard image information corresponding to the standard face shape;

[0027] The image information preliminary screening unit is connected to the image information acquisition module signal, obtains the patient's facial image, and obtains the patient's standard face shape according to the facial image;

[0028] An image information segmentation unit, which selects a face shape database according to the patient's standard face shape, and segments the facial image into corresponding facial features based on the standard image information in the face shape database;

[0029] The acupoint positioning unit obtains the actual length of each pixel grid in the facial image on the patient's face, takes the patient's facial features as a reference, and uses the length of the acupoint from the boundary of the facial features as a basis to locate the position of each acupoint.

[0030] The technical solution proposed in this application is not to simply compare the patient's facial image with the database image directly to find the most similar facial image to complete the acupoint positioning. On the contrary, this solution first accurately captures the position of the patient's facial features, and then accurately locates the acupoints based on the positions of these facial features according to the established acupoint positioning rules. Compared with the traditional method that relies on acupoint similarity matching, this solution significantly improves the positioning accuracy and achieves more accurate and effective acupoint recognition. At the same time, compared with the prior art, the solution provided by this application can also reduce the difficulty of collecting and marking standard image information. Under the existing solution, it is necessary to mark the acupoints on each face image in the standard image information so that the acupoint recognition system can identify the corresponding acupoints, while in this solution, only the boundaries and contours of the facial features need to be marked, so the difficulty of marking is significantly reduced; at the same time, this solution can also avoid the influence of errors under different scales. Because each standard image information collected cannot be obtained at the same scale, errors will occur when actually locating acupoints because the scale of the standard image information is unclear. In this solution, only the boundaries of the facial features in the facial image are obtained using standard image information, and the length in the facial image is converted using markings, thereby increasing the accuracy of acupoint positioning.

[0031] Furthermore, the image information preliminary screening unit obtains the edge contour lines of the face in the patient's facial image, calculates the similarity between the patient's upper line segments, lower line segments and side line segments and each type of upper line segments, lower line segments and side line segments, respectively, and screens out the type of upper line segments, lower line segments and side line segments that are closest to the patient, thereby obtaining the facial category of the patient's face shape.

[0032] The technical solution proposed in this application adopts a method of directly comparing and analyzing the patient's facial contour from three different directions when determining the patient's face shape. This solution can accurately match the patient's face shape to the closest standard face shape, thereby providing a more accurate benchmark for subsequent acupoint positioning, significantly improving the accuracy of acupoint positioning.

[0033] In the process of face recognition, achieving accurate matching of the patient's face shape actually depends on a moderate fuzzy matching strategy. At present, edge contour matching usually relies on neural network models. Although it can provide high matching accuracy, it is accompanied by high resource consumption. In practical applications, given the uniqueness of each person's face shape, too high matching accuracy not only fails to significantly improve the accuracy of matching, but may cause overfitting of the model and affect the generalization ability. In addition, using Hausdorff distance for similarity matching may also make it difficult to achieve accurate alignment due to improper accuracy setting. Another common practice is to calculate the minimum bounding box (MBR) of the contour and match it by comparing the size, position and direction of the bounding box. Although this method has a small amount of calculation, it ignores the fine shape of the contour edge, thus lacking in the accuracy of face matching.

[0034] In view of the problem that it is difficult to strike a balance between simplicity and accuracy in current face shape matching technology, the present application proposes a technical solution to achieve matching of patient faces in a more efficient and accurate manner.

[0035] The image information preliminary screening unit sets a number of sampling points at equal intervals on the upper line segment, and fits the sampling points to a preset first multi-order function to obtain a first fitting parameter;

[0036] The image information preliminary screening unit sets a number of sampling points at equal intervals on the lower line segment, and fits the sampling points to a preset second multi-time function to obtain a second fitting parameter;

[0037] The image information preliminary screening unit sets a number of sampling points at equal intervals on the side line segment, fits the sampling points to a preset third multi-time function, and obtains a third fitting parameter;

[0038] The image information preliminary screening unit matches the face shape according to the similarity between the first fitting parameter, the second fitting parameter and the third fitting parameter and the first fitting parameter, the second fitting parameter and the third fitting parameter of each standard face shape.

[0039] In the technical solution proposed in this application, by flexibly adjusting the number of sampling points, effective management of the amount of calculation is achieved, and the computing resources can be optimized according to actual needs. After carefully selecting the appropriate number of sampling points, this solution can ensure the accuracy of face shape matching and achieve precise alignment. More importantly, this solution performs similarity assessment based on the fitting parameters of the sampling points, which essentially matches the subtle undulations of the contour edge, which significantly improves the accuracy in the facial contour matching process. In this way, this solution adopts a simple and efficient method to successfully achieve accurate matching of the patient's face shape, which not only controls the complexity of the calculation, but also ensures the high accuracy of the matching results.

[0040] Given that each patient's facial features are unique, and that patients with facial paralysis often have atypical changes in facial morphology, this greatly increases the difficulty of accurately locating acupoints based on existing facial features in the database. In order to overcome this challenge, this application proposes the following solutions:

[0041] The image information segmentation unit includes:

[0042] The facial features preliminary delineator is used to outline the preliminary boundaries of the facial features area, including eyes and nose;

[0043] The edge delineator extracts key points within the preliminary boundary, compares the key points with the standard image information in the face shape database, and determines whether the key points are the boundaries of the facial features. If so, the key points are recorded as edge points, and if not, the key points are deleted.

[0044] Collect all edge points and connect them to generate the boundaries of the facial features.

[0045] The technical solution proposed in this application abandons the traditional practice of simply comparing the patient's facial image with the image in the face database. On the contrary, this application uses an edge delineator to extract the key points of the face, and then compares the key points of the face with the standard image information in the database to determine whether the key points constitute the boundaries of the facial features. Subsequently, by orderly connecting these key points belonging to the boundaries, the contours of the facial features are accurately outlined. The core advantage of this solution is that it makes full use of the stability of the key points. These points can remain constant under different lighting, rotation angles and scale changes. These key points essentially correspond to the edge areas where facial features are significant. Therefore, only by comparing and analyzing these key points, the boundaries of the facial features can be efficiently and accurately located, significantly improving the accuracy of matching. Through this solution, the problem of acupoint positioning caused by facial deformation in patients with facial paralysis is effectively solved, providing strong support for precision medicine.

[0046] Furthermore, the acupoint marking module includes:

[0047] A number of lasers with adjustable angles for generating lasers with wavelengths in the visible light band;

[0048] The control unit is respectively connected to the acupoint positioning module and the diagnosis and treatment information acquisition module by signal, obtains the position of each acupoint on the patient's face, and controls the laser to project the light spot to the corresponding acupoint to form a light spot according to the acupuncture sequence of each acupoint recorded by the diagnosis and treatment information acquisition module. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] The drawings constituting a part of this application are used to provide a further understanding of this application, so that other features, purposes and advantages of this application become more obvious. The illustrative embodiment drawings and their descriptions of this application are used to explain this application and do not constitute an improper limitation on this application.

[0050] In addition, throughout the drawings, the same or similar reference numerals represent the same or similar elements. It should be understood that the drawings are schematic and that the components and elements are not necessarily drawn to scale.

[0051] In the attached picture:

[0052] Figure 1 Schematic diagram of the positioning guidance system for facial acupuncture.

[0053] Figure 2 Schematic diagram of the division of facial edge contours. DETAILED DESCRIPTION

[0054] Embodiments of the present application will be described in more detail below with reference to the accompanying drawings. Although certain embodiments of the present application are shown in the accompanying drawings, it should be understood that the present application can be implemented in various forms, and should not be construed as being limited to the embodiments set forth herein. On the contrary, these embodiments are provided to provide a more thorough and complete understanding of the present application. It should be understood that the drawings and embodiments of the present application are only for exemplary purposes and are not intended to limit the scope of protection of the present application.

[0055] It should also be noted that, for ease of description, only the parts related to the invention are shown in the drawings. In the absence of conflict, the embodiments and features in the embodiments of the present application can be combined with each other.

[0056] The present application will be described in detail below with reference to the accompanying drawings and in combination with embodiments.

[0057] Example 1: Reference Figure 1 A positioning and guiding system for facial acupuncture includes: a diagnosis and treatment information acquisition module, an image information acquisition module, an acupoint positioning module, and an acupoint marking module. The acupoint marking module is signal-connected to the diagnosis and treatment information acquisition module and the acupoint positioning module, and the image information acquisition module is signal-connected to the acupoint positioning module.

[0058] The diagnosis and treatment information acquisition module is used to obtain the patient's treatment prescription and obtain the facial acupuncture points where the patient needs acupuncture; the image information acquisition module is used to obtain the facial image of the patient before acupuncture; the acupuncture point positioning module is used to compare the facial image of the patient before acupuncture with the facial image with acupuncture points marked in the face shape database to obtain the position information of the facial acupuncture points where the patient needs acupuncture; the acupuncture point marking module is used to obtain the position information of the facial acupuncture points where the patient needs acupuncture, and then irradiate the facial acupuncture points with lasers with a wavelength of visible light frequency in sequence to guide the acupuncture treatment.

[0059] The above is the general technical solution of the positioning guidance system for facial acupuncture provided by this application. The following is a further introduction to the content of the technical solution:

[0060] The diagnosis and treatment information acquisition module is connected to the hospital's HIS system to obtain the patient's treatment prescription, or obtains the patient's treatment prescription by manual input. The treatment prescription in this solution is acupuncture treatment, including the acupuncture points, the acupuncture order of the acupuncture points, and the interval between acupuncture treatments.

[0061] Furthermore, the diagnosis and treatment information acquisition module includes: a prescription information entry unit and an input unit. The prescription information entry unit is used to enter the patient's treatment prescription, and read the acupuncture points to be acupunctured, the acupuncture sequence of the acupuncture points, and the interval between acupuncture points from the treatment prescription; the input unit is used to adjust the acupuncture points to be acupunctured, the acupuncture sequence of the acupuncture points, and the interval between acupuncture points.

[0062] The diagnosis and treatment information acquisition module is essentially an information processing terminal, such as a computer or a mobile phone. The staff can use the diagnosis and treatment information acquisition module to download the treatment prescription in the hospital HIS system, or directly input the treatment prescription using the input unit. The specific input method of the treatment prescription and the reading method of the treatment prescription are prior arts, and this application will not further introduce them.

[0063] The image information acquisition module acquires the facial image of the patient before acupuncture. Specifically, the patient will lie in a designated position before acupuncture, and at this time, a mobile camera or a fixed camera can be used to capture the facial image of the patient. In practice, the facial image is image information that has completed the face range segmentation. The solution of extracting the facial image along the facial edge contour is a prior art and will not be elaborated here.

[0064] The image information acquisition module includes: a camera and a marking line. The camera is fixed on the frame, or the staff holds the camera. Regardless of the camera fixing method, in this solution, the camera needs to be used to obtain the front face image of the patient. The marking line is a line of fixed length. When in use, the marking line is straightened and attached to the patient's face. The marking line needs to be away from the acupuncture points of the patient to avoid unclear identification of the acupuncture points in this part.

[0065] This application uses an image information acquisition module to acquire a frontal facial image of a patient, and a pre-recorded length of a marking line is also attached to the facial image. For example, the length of the marking line is 1 cm, which occupies 1000 pixels in the facial image, so the width of 1 pixel on the patient's face is 0.01 mm. In this way, when calculating the distance between any two points on the patient's face, it is only necessary to calculate how many pixels are between the two points.

[0066] In some embodiments, in order to reduce the amount of calculation, the acupoint positioning module pre-acupunctures the acupoints that need to be acupunctured in the diagnosis and treatment information acquisition module. In the acupoint positioning module, only the relevant acupoints are located, thereby reducing the amount of calculation and improving the system feedback efficiency.

[0067] Specifically: the acupoint positioning module includes a face shape definition unit, an image information storage unit, an image information preliminary screening unit, an image information segmentation unit and an acupoint positioning unit. Among them, the face shape definition unit divides the edge contour of the face into upper line segments, lower line segments and side line segments, and defines n1 types of upper line segments, n2 types of lower line segments, and n3 types of side line segments to generate N types of standard face shapes, N=n1×n2×n3.

[0068] In this scheme, each standard face shape is divided into three parts, and each part is classified and summarized. Therefore, after analyzing all standard face shapes, n1 types of upper line segments, n2 types of lower line segments, and n3 types of side line segments can be obtained. Thus, any one upper line segment, lower line segment, and side line segment selected from the n1 types of upper line segments, n2 types of lower line segments, and n3 types of side line segments and combined together is a standard face shape. Therefore, the number of standard face shapes N=n1×n2×n3.

[0069] The image information storage unit is configured with a plurality of face shape databases, each of which corresponds to a standard face shape, and each of which stores standard image information corresponding to the standard face shape. In this solution, the standard image information stored in the image information storage unit is a facial image that has been annotated, that is, the boundaries of each facial feature area are annotated. When storing the facial image of the patient, the facial image is classified according to the facial category of the standard face shape in this solution.

[0070] The image information preliminary screening unit is connected to the image information acquisition module signal, obtains the edge contour line of the face in the patient's facial image, calculates the similarity between the patient's upper line segments, lower line segments and side line segments with each type of upper line segments, lower line segments and side line segments, and screens out the type of upper line segments, lower line segments and side line segments that are closest to the patient, so as to obtain the facial category of the patient's face shape.

[0071] When locating acupoints, it is necessary to merge the image information of faces that are relatively similar to each other to make it easier to complete the acupoint location. Relatively speaking, the closer the face shapes of two people are, the closer the corresponding facial features and characteristics will be. In this way, when matching the facial features in the subsequent process, the accuracy will be higher, which increases the accuracy of the marking.

[0072] To this end, in this solution, the image information preliminary screening unit will preliminarily determine the patient's face shape. Specifically, the face shape determination method is as follows:

[0073] S1: Obtain the edge contour of the face in the face image.

[0074] In practice, the patient lies on a bed sheet with a single background. In the facial image thus obtained, the boundary between the patient's face and the background is clear, and the edge contour line can be directly obtained. In addition, the edge contour line can also be drawn manually. In other embodiments, an editable edge contour line can be automatically generated based on the difference between the skin color and the background color of most patients, and the staff can adjust the edge contour line.

[0075] S2: Divide the edge contour into an upper line segment, a lower line segment, and a side line segment.

[0076] The upper line segment corresponds to the forehead, the lower line segment corresponds to the chin, and the side line segments correspond to the sides of the face. Figure 2 As shown, Figure 2 The corresponding positions of the upper line segment, the lower line segment and the side line segment are marked.

[0077] The method for defining the edge contour line of the face is as follows: mark the corners of the eyes and the corners of the mouth in the facial image, and draw horizontal lines for each. The contour line between the two horizontal lines is the side line segment, the part above the two horizontal lines is the upper line segment, and the part below the two horizontal lines is the lower line segment.

[0078] S3: Calculate the similarity between the patient's upper line segments, lower line segments and side line segments and each type of upper line segments, lower line segments and side line segments, and select the type of upper line segments, lower line segments and side line segments that are closest to the patient, so as to obtain the patient's standard face shape.

[0079] Because the standard face shape is determined based on the type of the upper line segments, the type of the lower line segments, and the type of the side line segments, when determining the patient's standard face shape, it is only necessary to calculate the similarity between the upper line segments in the patient's facial image and the upper line segments of all standard face shapes, and filter out the type of upper line segments with the highest similarity, and so on, filter out the lower line segments and side line segments with the highest similarity, and then combine them to form the patient's standard face shape.

[0080] The above is the matching method of the patient's face shape. The most critical part in the matching process is the similarity calculation. The following is the specific plan for the similarity calculation.

[0081] The first step: the image information preliminary screening unit sets a number of sampling points at equal intervals on the upper line segment, fits the sampling points to a preset first multi-time function f1 (x), and obtains a first fitting parameter.

[0082] The first multi-time function is preset, and its expression is f1(x)=a1x m +a2x m-1 +a3x m-2 …+a m x 1 ; Among them, a1, a2, a3, ..., a m is the first fitting parameter, x is the variable, m is the number of terms of the first multi-time function, m is a positive integer. If m=3, the number of terms of the first multi-time function is 3. The more terms there are, the higher the accuracy of the subsequent fitting will be, and the corresponding amount of calculation will be greater.

[0083] The number of sampling points is also pre-set. The more sampling points there are, the greater the corresponding calculation amount and the greater the difficulty of calculation. For example, there are 50 sampling points (x, y). Both x and y in the sampling points are clear. We only need to make f1(x) as close to y as possible every time we substitute the sampling point x into f1(x).

[0084] The upper line segment is actually a line segment in the plane. Move the line segment into a plane rectangular coordinate system, and then select sampling points from it, retain the sampling points, and delete the rest. In this way, the remaining sampling points can be used to fit the first multi-time function f1 (x). After the fitting is completed, a1, a2, a3, ..., a in the first multi-time function f1 (x) can be calculated. m The value of the first fitting parameter of this type.

[0085] The above is the method for obtaining the first fitting parameter. The principles of the subsequent second fitting parameter and the third fitting parameter are the same.

[0086] When moving a line segment into a plane rectangular coordinate system, a point on the left side of the line segment is used as a reference point. It only needs to be that the reference point is located at the same point in the plane rectangular coordinate system.

[0087] Step 2: The image information preliminary screening unit sets a number of sampling points at equal intervals on the lower line segment, and fits the sampling points to a preset second multi-time function f2 (x) to obtain second fitting parameters.

[0088] The second multi-time function is preset, and its expression is f2(x)=b1x j +b2x j-1 +b3x j-2 …+b j x 1 ; Among them, b1, b2, b3, ..., bm is the second fitting parameter, x is the variable, j is the number of terms of the second multi-time function, and j is a positive integer.

[0089] The third step: the image information preliminary screening unit sets a number of sampling points at equal intervals on the side line segment, and fits the sampling points to a preset third multi-time function f3 (y) to obtain third fitting parameters.

[0090] The third multi-time function is preset, and its expression is f3(y)=c1y o +c2y o-1 +c3y o-2 …+c o y 1 ; where c1, c2, c3, …, c o is the third fitting parameter, y is the variable, o is the number of terms of the third multi-time function, and o is a positive integer. The difference between the third multi-time function and the second multi-time function is that the value of the ordinate is used as the independent variable for fitting, while the first and second multi-time functions both use the abscissa as the independent variable.

[0091] According to the above method, a corresponding set of three fitting parameters can be obtained, and the set of these fitting parameters actually describes the shape changes of the edges of each region in the facial image.

[0092] Step 4: The image information preliminary screening unit matches the face shape according to the similarity between the first fitting parameter, the second fitting parameter and the third fitting parameter and the first fitting parameter, the second fitting parameter and the third fitting parameter of each standard face shape.

[0093] In this solution, several standard face shapes are pre-configured, and for each standard face shape, the corresponding first fitting parameter, second fitting parameter and third fitting parameter are calculated using the method of the first to third steps. Therefore, the similarity of the first fitting parameter, the second fitting parameter and the third fitting parameter can be used as the basis for calculating the similarity in S3.

[0094] For example, we obtain the facial image of patient T, divide the facial contour in the facial image of patient T into upper line segments, lower line segments and side line segments, and then calculate the first fitting parameter of the upper line segment, the second fitting parameter of the lower line segment and the third fitting parameter of the side line segment, calculate the similarity between the first fitting parameter of patient T and the first fitting parameter of the upper line segment of each standard face shape, and obtain the closest upper line segment, and then select the lower line segment and side line segment with the highest similarity, so as to complete the matching of the standard face shape. In this way, the matching of facial images is completed by using the above scheme.

[0095] The similarity is calculated by cosine similarity. The first fitting parameter, the second fitting parameter and the third fitting parameter are sequences arranged in sequence, so that the cosine similarity can be calculated.

[0096] The image information segmentation unit is mainly used to identify the facial features of the patient in order to segment the facial features of the patient from the facial image. Specifically, the image information segmentation unit includes: a preliminary facial features delineator and an edge delineator. The preliminary facial features delineator is used to delineate the preliminary boundaries of the facial features area, which includes the eyes, nose, mouth, etc. This part is mainly manually marked. The staff marks the approximate edges of the facial features such as the eyes and nose in the image information. This part does not need to be precisely outlined, but only needs to roughly distinguish the positions of key areas such as the eyes, nose, and mouth.

[0097] Then the edge delineator performs actual edge recognition of the facial features based on this part of the edge. Specifically, the edge delineator extracts key points within the preliminary boundary, compares the key points with the standard image information in the face database, and determines whether the key points are the boundaries of the facial features; if the key points match a feature point with high similarity in the face database, then the key points are used as edge points, and if no feature point with high similarity can be matched, then the key points are not used as edge points; collect all edge points and connect them to generate the boundaries of the facial features.

[0098] In the solution provided by the present application, the edge delineator will extract key points within the preliminary boundary, and then determine whether the key points are edge points. After obtaining all the edge points, the edges of the corresponding facial features can be obtained. The function of the edge delineator is to accurately delineate the edges of the facial features for subsequent acupoint positioning.

[0099] The key point extraction method and key point matching method include the following steps:

[0100] SO1: The edge delineator obtains the pixel values ​​within each preliminary boundary and converts the pixel values ​​into grayscale values.

[0101] This step mainly converts the image information with color into grayscale information. In some implementations, a Gaussian filter may also be used for filtering.

[0102] SO2: Randomly select a pixel point P from the preliminary boundary i , i represents the index of the pixel point, i is a positive integer; i Generate a circular judgment area with a preset radius as the center.

[0103] In this solution, the diameter of the circular judgment area is set to 3 pixels.

[0104] SO3: Pre-set threshold t and compare pixel Pi and the grayscale values ​​of the other pixels in the circular judgment area to determine the pixel P i Is it a candidate corner point? The following two conditions are true for the pixel point P i Satisfy any one, pixel point P i is a candidate corner point;

[0105] The first condition is: if there are h consecutive pixels in the circular judgment area that are larger than pixel P i The gray value of the plus the threshold t, then P i is a candidate corner point;

[0106] The second condition is: if there are h consecutive pixels smaller than pixel P in the circular judgment area i The gray value minus the threshold t, then P i is a candidate corner point.

[0107] In fact, screening candidate corner points (extreme points) is to screen pixel points P i Is it the point with the smallest pixel value or the point with the largest pixel value in the circular area? If it is the point with the smallest pixel value, then there must be h pixels with grayscale values ​​greater than the pixel P. i On the contrary, there are h pixels with gray values ​​less than pixel P i In order to further increase the stability of the candidate corner points in this scheme, the threshold t is further increased, so that the difference between the pixel values ​​of the candidate corner points and the surrounding pixels must be greater, and h is an integer greater than zero.

[0108] SO4: Traverse all the pixels within the preliminary boundary to obtain the candidate corner point set A. Each point in the candidate corner point set A is a key point.

[0109] In SO4, non-maximum suppression is performed on each candidate corner point. That is, the response values ​​of the candidate corner point and its surrounding pixels are compared (the response value can be calculated based on the grayscale difference). If the response value of the candidate corner point is not the local maximum, it is removed from the candidate corner point set. In this solution, the response value is the grayscale value.

[0110] In SO1~SO4, the key points have been collected. The subsequent steps require the use of appropriate methods to describe each key point. Specifically:

[0111] SO5: Pre-configure a rectangular window. For each key point in the candidate corner point set A, determine a rectangular window with the key point as the center, and randomly select m pairs of pixels in the rectangular window.

[0112] For each pair of pixels (p1, p2), compare their grayscale values ​​I(p1) and I(p2). If I(p1) > I(p2), the corresponding descriptor bit is 1; otherwise, it is 0 to generate an m-bit binary descriptor, which is the descriptor of the key point.

[0113] For ease of understanding, the descriptor bits and descriptors are explained as follows: By comparing the grayscale values ​​of each pair of pixels (p1, p2), a binary bit (0 or 1) is obtained. This binary bit is called the "descriptor bit". By obtaining all the "descriptor bits", the descriptor can be obtained.

[0114] In this solution, the rectangular window is pre-configured and has a fixed size. Since the size of the rectangular window is fixed, a rectangular window can be configured with each candidate corner point as the center, and then m pairs of pixels are selected in the rectangular window. The size of m can be 128 or 256.

[0115] Compare the key points with the standard image information in the face shape database. In the above scheme, all the key points within the preliminary boundary have been extracted, and then the key points are compared with the corresponding boundary points (feature points) representing the facial features in the face shape database. If the similarity of a key point exceeds a predetermined threshold, it indicates that the key point is an edge point of the facial features.

[0116] For example, a key point F is extracted from the preliminary area of ​​the eye in the facial image. If the similarity between the key point F and a feature point on the edge of the eye in the face shape database exceeds a preset value, then the key point F can be identified as an edge point. Otherwise, if there is no key point in the face shape data whose similarity with the key point F exceeds the preset value, then the key point F is not an edge point.

[0117] In this way, the above operation can be completed by ensuring that the facial image in the face shape data can accurately mark the corresponding edge contour. For the calculation of similarity, cosine similarity is used in this solution. The specific calculation formula will not be further elaborated here.

[0118] Of course, when judging whether a key point is an edge point, in addition to setting a similarity threshold, you can also set a quantity threshold. For example, if a key point F is extracted from the preliminary area of ​​the eyes of a facial image, it is necessary to be able to match 5 feature points from the face shape database with a similarity exceeding the threshold before the key point F can be identified as a boundary point.

[0119] The acupoint positioning unit obtains the actual length of each pixel grid in the facial image on the patient's face, and locates the position of each acupoint based on the patient's facial features and the distance from the acupoint to the boundary of the facial features. For example, according to medical records, a certain acupoint is located one inch below the eye. In this solution, the outline of the eye can be accurately located, and the side length of each pixel grid can be accurately obtained, so the position of one inch below the eye can be accurately located.

[0120] The above scheme has already defined the boundaries of each facial feature area and the length of each pixel grid in the facial image. Therefore, the location of the acupoints can be completed according to the pre-entered rules. The specific calculation method is not listed here.

[0121] Through the above scheme, the position of each acupuncture point in the patient's facial image and the acupuncture order of each acupuncture point have been clearly understood.

[0122] The acupoint marking module includes a bracket, a laser and a control unit. The bracket is set above the patient's face, the laser is set on the bracket, there are multiple lasers, and the angle of the laser is adjustable.

[0123] The arrangement structure of the laser and the adjustment structure of the laser angle are prior arts and will not be further elaborated in this solution.

[0124] After the laser emits the laser, a light spot can be formed on the patient's face, thus guiding the staff to perform acupuncture on the patient according to the treatment prescription. Specifically: the control unit is connected to the acupoint positioning module and the diagnosis and treatment information acquisition module respectively, obtains the location of each acupoint on the patient's face, and controls the laser to project the light spot to the corresponding acupoint to form a light spot according to the acupuncture sequence of each acupoint recorded by the diagnosis and treatment information acquisition module.

[0125] In practice, because the angle of the laser adjustment is not accurate, and the position of the patient's face relative to the laser will also be offset to a certain extent, in fact, the laser first irradiates the patient's face with a light spot, and then determines whether to continue adjusting the laser angle based on whether the light spot is located on the corresponding acupuncture point.

[0126] To this end, this solution also provides the following methods:

[0127] The control unit obtains the acupuncture point that needs to be acupunctured according to the acupuncture sequence of each acupuncture point, and then controls the corresponding laser to adjust to the corresponding angle according to the position of the acupuncture point, and then emits the laser; it determines the relative position between the light spot of the current laser irradiation on the patient's face and the acupuncture point position, and further adjusts the angle of the laser according to the relative position until the light spot coincides with the acupuncture point position. The acupuncture point position has been calculated by the acupuncture point positioning module, and the position of the light spot can be directly identified by RBG color. The specific method of identifying the light spot will not be discussed here.

[0128] The control unit adjusts the laser spot color to green when adjusting the laser, and adjusts the laser spot color to red when the laser spot coincides with the acupuncture point position.

[0129] The solution provided by the present application can prevent the staff from misjudging the current spot position as the acupuncture point position during the laser adjustment process, resulting in incorrect acupuncture position, thereby affecting the treatment effect. In this way, the facial acupuncture positioning guidance system provided by the present application can accurately guide the acupuncture process.

[0130] In a further solution, the control unit can use the color of the laser to remind the staff to stop the acupuncture according to the pre-recorded acupuncture time.

[0131] For example: the treatment prescription states that acupuncture point No. 1 needs to be inserted for 50 seconds. After the laser spot is irradiated on acupuncture point No. 1 for 50 seconds, the laser will immediately turn into yellow light and sound an alarm. At this time, the staff can follow the prompts and accurately pull out the corresponding needle.

[0132] In this way, the present solution can guide the entire acupuncture process, avoid errors made by staff during the acupuncture process, and increase the accuracy of acupuncture.

[0133] The above description is only some preferred embodiments of the present application and an explanation of the technical principles used. Those skilled in the art should understand that the scope of the invention involved in the embodiments of the present application is not limited to the technical solutions formed by a specific combination of the above technical features, but should also cover other technical solutions formed by any combination of the above technical features or their equivalent features without departing from the above invention concept. For example, the above features are replaced with (but not limited to) technical features with similar functions disclosed in the embodiments of the present application to form a technical solution.

Claims

1. A positioning guidance system for facial acupuncture, characterized in that: include: The diagnosis and treatment information acquisition module obtains the patient's treatment prescription and the facial acupuncture points that the patient needs to receive; An image information acquisition module is used to acquire the facial image of the patient before acupuncture; The acupuncture point positioning module compares the facial image of the patient before acupuncture with the facial image with acupuncture points marked in the face database to obtain the location information of the facial acupuncture points that need to be acupunctured on the patient; The acupuncture point marking module obtains the location information of the facial acupuncture points where the patient needs to be treated, and then irradiates the facial acupuncture points with a wavelength of visible light frequency in sequence to guide acupuncture treatment; in: The image information acquisition module includes: a marking line, which is attached straightly to the patient's face, and the length of the marking line is pre-recorded in the acupoint positioning module; The acupoint positioning module includes: A face shape definition unit divides the edge contour of the face into upper line segments, lower line segments and side line segments, and defines n1 types of upper line segments, n2 types of lower line segments, and n3 types of side line segments to generate standard face shapes of N face categories, N=n1×n2×n3; The image information storage unit is configured with a plurality of face shape databases, each face shape database corresponds to a standard face shape, and each face shape database stores standard image information corresponding to the standard face shape; The image information preliminary screening unit is connected to the image information acquisition module signal, obtains the edge contour line of the face in the patient's facial image, calculates the similarity between the patient's upper line segment, lower line segment and side line segment and each type of upper line segment, lower line segment and side line segment, and screens out the type of upper line segment, lower line segment and side line segment that are closest to the patient, so as to obtain the patient's standard face shape; An image information segmentation unit, which selects a face shape database according to the patient's standard face shape, and segments the facial image into corresponding facial features based on the standard image information in the face shape database; The acupoint positioning unit obtains the actual length of each pixel grid in the facial image on the patient's face, takes the patient's facial features as a reference, and uses the length of the acupoint from the boundary of the facial features as a basis to locate the position of each acupoint.

2. The positioning guidance system for facial acupuncture according to claim 1, characterized in that: The diagnosis and treatment information acquisition module includes: The prescription information input unit is used to input the patient's treatment prescription, and read the acupuncture points to be acupunctured, the acupuncture order of the acupuncture points, and the interval between acupuncture points from the treatment prescription; An input unit, used to adjust the acupuncture points to be acupunctured, the acupuncture order of the acupuncture points, and the interval between acupuncture points; The acupoint marking module marks the corresponding acupoints with light spots in sequence according to the acupuncture sequence and the interval time of acupuncture.

3. The positioning guidance system for facial acupuncture according to claim 1, characterized in that: The image information acquisition module includes: A camera is used to obtain facial images of the patient.

4. The facial acupuncture positioning guidance system according to claim 1, characterized in that: The method of defining the edge contour of the face is: Mark the corners of the eyes and mouth in the facial image, and draw horizontal lines respectively. The contour line of the part between the two horizontal lines is the side line segment, the part above the two horizontal lines is the upper line segment, and the part below the two horizontal lines is the lower line segment.

5. The facial acupuncture positioning guidance system according to claim 4, characterized in that: The image information preliminary screening unit sets a number of sampling points at equal intervals on the upper line segment, and fits the sampling points to a preset first multi-order function to obtain a first fitting parameter; The image information preliminary screening unit sets a number of sampling points at equal intervals on the lower line segment, and fits the sampling points to a preset second multi-time function to obtain a second fitting parameter; The image information preliminary screening unit sets a number of sampling points at equal intervals on the side line segment, fits the sampling points to a preset third multi-time function, and obtains a third fitting parameter; The image information preliminary screening unit matches the face shape according to the similarity between the first fitting parameter, the second fitting parameter and the third fitting parameter and the first fitting parameter, the second fitting parameter and the third fitting parameter of each standard face shape.

6. The positioning guidance system for facial acupuncture according to claim 3, characterized in that: The image information segmentation unit includes: The facial features preliminary delineator is used to outline the preliminary boundaries of the facial features area, including eyes and nose; The edge delineator extracts key points within the preliminary boundary, compares the key points with the standard image information in the face shape database, and determines whether the key points are the boundaries of the facial features. If so, the key points are recorded as edge points, and if not, the key points are deleted. Collect all edge points and connect them to generate the boundaries of the facial features.

7. The facial acupuncture positioning guidance system according to claim 1, characterized in that: The acupoint marking module includes: A number of lasers with adjustable angles for generating lasers with wavelengths in the visible light band; The control unit is respectively connected to the acupoint positioning module and the diagnosis and treatment information acquisition module by signal, obtains the position of each acupoint on the patient's face, and controls the laser to project the light spot to the corresponding acupoint to form a light spot according to the acupuncture sequence of each acupoint recorded by the diagnosis and treatment information acquisition module.

Images