Method and system for constructing head and neck curve classification model

Abstract

The invention discloses a head and neck curve classification model construction method and system, and relates to the field of medical health and industrial product design, and the method comprises the steps: employing a laser range finder to measure the nod coordinate data of the rear side of the neck of the sagittal plane of a human body, and obtaining a head and neck two-dimensional point set; fitting the two-dimensional point set by using a four-order B spline fitting method to obtain a head and neck sample curve; performing clustering analysis on the plurality of sample curves to obtain classification data of the sample curves; and performing hierarchical fitting on each type of sample curves, constructing a representative curve model, and designing the pillow curvature by using the representative curve model. According to the method, the head and neck curve information can be more accurately obtained, and then the curve classification model is constructed, so that a standardized reference is provided for subsequent pillow curvature design, and a more suitable pillow is designed.

Description

Technical Field

[0001] The present invention relates to the technical field of medical health and industrial product design, and more particularly to a method and system for constructing a head and neck curve classification model. Background Art

[0002] As people's pace of life accelerates, cervical spine problems are becoming increasingly common. In order to provide a better sleeping experience and cervical spine protection, it is necessary to design a suitable pillow based on the curvature of the individual's head and neck. However, most pillows on the market are standardized and difficult to meet personalized needs. Therefore, it is particularly important to develop a method to accurately measure the curvature of the head and neck and design the shape of the pillow accordingly.

[0003] However, the existing technology obtains the human head and neck curve information mainly through image recognition or measuring instruments with physical telescopic structures. The image recognition method is easily affected by the hair contour of the person being measured and cannot truly reflect the curve characteristics of the posterior side of the sagittal plane of the head and neck; and the newly formed curve is not smooth and cannot be directly applied to the model design of products such as pillows; at the same time, due to individual differences, each person's head and neck curve is different, and the directly obtained curve data information cannot be directly applied to the design of the market-standard pillow model. The measuring instrument method based on the physical telescopic structure can only obtain the head and neck curve modeling information due to the telescopic structure of the physical connecting rod, which must be further converted into a digital signal for subsequent processing. In terms of the currently used data fitting, the method of using a cubic linear regression equation cannot adjust the curve by partition, and cannot well match the physiological curve characteristics of the human body.

[0004] Therefore, how to more accurately obtain the head and neck curve information and then construct a curve classification model is a problem that technical personnel in this field need to solve urgently. Summary of the invention

[0005] In view of this, the present invention provides a method and system for constructing a head and neck curve classification model, which can quickly and accurately obtain the coordinate points of the head and neck, form a smooth curve that conforms to the characteristics of the cervical curve, and can classify curves with different characteristics and fit the curves of different classifications to form a representative curve model, so as to design a pillow that is suitable for a certain type of head and neck physiological curvature characteristics.

[0006] In order to achieve the above object, the present invention provides the following technical solutions:

[0007] The present invention discloses a method for constructing a head and neck curve classification model, and the specific steps are as follows:

[0008] The laser rangefinder is used to measure the coordinate data of each point on the back of the neck in the sagittal plane of the human body to obtain a two-dimensional point set of the head and neck;

[0009] The two-dimensional point set is fitted using a fourth-order B-spline fitting method to obtain a head and neck sample curve;

[0010] Performing cluster analysis on the plurality of sample curves to obtain classification data of the sample curves;

[0011] Perform hierarchical fitting on each type of sample curve to build a representative curve model.

[0012] Furthermore, the laser rangefinder has a plurality of laser rangefinder probes, which are arranged in a row; the coordinate data is determined by measuring the horizontal distance from each point on the neck to the corresponding laser rangefinder probe, as well as the position information of each laser rangefinder probe.

[0013] Furthermore, the fourth-order B-spline fitting method includes:

[0014] Based on the coordinate values ​​of the two-dimensional point set, construct a node vector corresponding to the two-dimensional point set;

[0015] The fourth-order B-spline basis function is obtained by recursively calculating the linear B-spline basis function;

[0016] Using the fourth-order B-spline basis function to represent the fitting value of each node in the node vector, a linear equation system is established;

[0017] Solve the linear equations using the least squares method to determine the unknown coefficients of each data point;

[0018] Based on the undetermined coefficients, a fourth-order B-spline fitting curve is constructed.

[0019] Furthermore, the node vector is represented as u=[u 0 ,u 1 ,...,u k-1 ,u k ,...,u n ,...,u n+k-1 ], where n is the number of data points in the two-dimensional point set, k is the order of the B-spline curve; the starting node is u 0 ,u 1 ,...,u k-1 , the end node is u n ,...,u n+k-1 , the internal node is represented by u j , j=k,k+1,...,n-1;

[0020] The starting node and the ending node are both repeated nodes.

[0021] Furthermore, the recursive definition of the fourth-order B-spline basis function is expressed as:

[0022]

[0023] Where x represents the coordinate value from the shoulder to the head, N j,4 (x) is the fourth-order B-spline basis function; N j,3 (x) is the third-order B-spline basis function, and through the second-order B-spline basis function N j,2 (x) is calculated recursively, and the calculation formula is:

[0024]

[0025] The linear B-spline basis function is expressed as:

[0026]

[0027] Furthermore, the linear equations are expressed as:

[0028]

[0029] Among them, y i ′ represents the fitting value of each data point; c j is the unknown coefficient, corresponding to the weight of the control point of the B-spline curve; N j,4 (x i ) means that in x i The value of the j-th B-spline basis function at point .

[0030] Furthermore, the formula for solving the linear equations using the least squares method is:

[0031]

[0032] middle, y i Represents the actual value of each data point;

[0033] The fourth-order B-spline fitting curve C(x) is expressed as:

[0034]

[0035] Furthermore, the cluster analysis includes:

[0036] Extracting the peak points and the trough points of each of the sample curves;

[0037] Based on the peak point and the trough point, construct a feature vector corresponding to the sample curve;

[0038] The hierarchical clustering method is used to cluster the characteristic vectors to determine the classification data of each sample curve.

[0039] Furthermore, the hierarchical fitting includes:

[0040] The sample curves of each category are averaged to obtain the representative curves of each category of sample curves;

[0041] Each representative curve is fitted again, and the fitting error is minimized by optimizing the B-spline control point parameters to obtain the representative curve model.

[0042] The present invention also discloses a head and neck curve classification model construction system, comprising:

[0043] Coordinate acquisition module: Use a laser rangefinder to measure the coordinate data of each point on the back of the neck in the sagittal plane of the human body to obtain a two-dimensional point set of the head and neck;

[0044] Curve fitting module: using a fourth-order B-spline fitting method to fit the two-dimensional point set to obtain a head and neck sample curve;

[0045] Cluster analysis module: performing cluster analysis on a plurality of the sample curves to obtain classification data of the sample curves;

[0046] Hierarchical fitting module: Perform hierarchical fitting on each type of sample curve to build a representative curve model.

[0047] Through the above technical solutions, it can be known that compared with the prior art, the present invention discloses a method and system for constructing a head and neck curve classification model. By measuring with a laser rangefinder, high-precision head and neck coordinate data can be obtained, thereby improving the accuracy of curve fitting; the fourth-order B-spline fitting method is adopted, which can more flexibly handle complex curve shapes, provide good smoothness and high approximation accuracy, and minimize the fitting error by optimizing the B-spline control point parameters, which can improve the generalization ability and applicability of the model; through cluster analysis, different sample curves can be effectively identified and classified, thereby providing an objective basis for analyzing the head and neck characteristics of different individuals, hierarchical fitting of various sample curves after clustering, and representative curve models can be extracted, which is helpful to realize the standardized description of different head and neck morphologies. The present invention can obtain head and neck curve information more accurately, and then construct a curve classification model, thereby providing a standardized reference for subsequent pillow curvature design, and designing a more suitable pillow. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0049] Figure 1It is a schematic diagram of the overall process of an embodiment of the present invention.

[0050] Figure 2 A schematic diagram of creating a human head and neck curve dataset according to an embodiment of the present invention.

[0051] Figure 3 Schematic diagram of two-dimensional coordinates of an embodiment of the present invention. DETAILED DESCRIPTION

[0052] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0053] The embodiment of the present invention discloses a method for constructing a head and neck curve classification model. Figure 1 As shown, the specific steps are as follows:

[0054] The laser rangefinder is used to measure the coordinate data of each point on the back of the neck in the sagittal plane of the human body to obtain a two-dimensional point set of the head and neck;

[0055] The four-order B-spline fitting method is used to fit the two-dimensional point set to obtain the head and neck sample curve;

[0056] Perform cluster analysis on multiple sample curves to obtain classification data of the sample curves;

[0057] Perform hierarchical fitting on each type of sample curve to build a representative curve model.

[0058] In a specific embodiment, the laser rangefinder has a plurality of laser rangefinder probes, and n laser rangefinder probes are arranged in a row; coordinate data is determined by measuring the horizontal distance from each point on the neck to the corresponding laser rangefinder probe, as well as the position information of each laser rangefinder probe.

[0059] Specifically, Figure 2 As shown in FIG, a rangefinder with n laser rangefinder probes combined into a row is used to align the head and neck on the posterior side of the sagittal plane of the human body. The distance from each point on the neck to the laser head is measured by the laser rangefinder. Combined with the position information of the laser head, a point set in a two-dimensional coordinate system is formed, as shown in FIG. Figure 3 As shown in the figure, x represents the coordinate value along the direction from shoulder to head, y represents the coordinate value of the head and neck curve measurement point corresponding to the x coordinate, and the direction of the y coordinate is perpendicular to the direction of the x coordinate and points to the extracted neck curve; the extracted head and neck data points are parameterized, and the coordinate values ​​are given according to the xy coordinate system through the points to form a two-dimensional coordinate system. The point set [x i,y i ], where i = 1, 2, ..., n. Define these points as p 1 ,p 2 ,...,p n Among them, x i : The horizontal coordinate of the data point represents a certain position information of the head and neck; i : The vertical coordinate of the data point represents the measurement value of the laser head to the head and neck curve under the corresponding horizontal coordinate.

[0060] In a specific embodiment, the fourth-order B-spline fitting method includes:

[0061] Based on the coordinate values ​​of the two-dimensional point set, construct the node vector corresponding to the two-dimensional point set;

[0062] The fourth-order B-spline basis function is obtained by recursively calculating the linear B-spline basis function;

[0063] The fourth-order B-spline basis function is used to represent the fitting value of each node in the node vector, and a linear equation system is established;

[0064] Solve the linear equations using the least squares method to determine the unknown coefficients for each data point;

[0065] Based on the unknown coefficients, a fourth-order B-spline fitting curve is constructed.

[0066] In a specific embodiment, the node vector is represented as u=[u 0 ,u 1 ,...,u k-1 ,u k ,...,u n ,...,u n+k-1 ], where n is the number of data points in the two-dimensional point set and k is the order of the B-spline curve. The starting node is u 0 ,u 1 ,...,u k-1 These nodes are the control points of the initial part of the curve. For the fourth-order B-spline (k = 4), these nodes take the same value (in p 1 The corresponding horizontal axis value is referenced), in order to create a natural boundary condition at the starting point of the curve, that is, the first and second order derivatives of the curve at the starting point are zero. The end node is u n ,...,u n+k-1 , these nodes are the control points for the end of the curve. Similar to the starting nodes, these nodes are repeated (in p n The corresponding abscissa value is a reference) to ensure that the first and second derivatives of the curve are zero at the end point, thus creating a natural boundary condition at the end point. The internal node is represented by u j , j = k, k+1, ..., n-1; internal node uj Corresponding to the horizontal coordinate x of the collected data point i , where i represents the index of the data point, that is, how many data points we need to fit, and j represents the index of the control point or basis function, that is, how many control points or basis functions are involved in the calculation when constructing the B-spline curve.

[0067] Both the start node and the end node are repeated nodes.

[0068] In a specific embodiment, the recursive definition of the fourth-order B-spline basis function is expressed as:

[0069]

[0070] Where x represents the coordinate value from the shoulder to the head, N j,4 (x) is the fourth-order B-spline basis function; N j,3 (x) is the third-order B-spline basis function, and through the second-order B-spline basis function N j,2 (x) is calculated recursively, and the calculation formula is:

[0071]

[0072] Similarly, N j,2 (x) is a second-order B-spline basis function, which can be obtained by using the linear B-spline basis function N j,1 (x) is calculated recursively;

[0073] The linear B-spline basis function is expressed as:

[0074]

[0075] From N j,1 (x) Start the calculation and gradually use the recursive formula to calculate higher-order basis functions until the required fourth-order B-spline basis function N is obtained. j,4 (x).

[0076] In a specific embodiment, the B-spline basis function is used to represent the fitting value of each data point, and the linear equation system is expressed as:

[0077]

[0078] Among them, y i ′ represents the fitting value of each data point; c j is the unknown coefficient, corresponding to the weight of the control point of the B-spline curve; N j,4 (x i ) means that in x i The value of the j-th B-spline basis function at point .

[0079] In a specific embodiment, the unknown coefficient c is solved by minimizing the sum of squared errors of all data points. j , minimize the sum of squares of errors, that is, use the least squares method to solve the undetermined coefficients. The formula for solving the linear equations using the least squares method is:

[0080]

[0081] middle, y i Represents the actual value of each data point;

[0082] The fourth-order B-spline fitting curve C(x) is expressed as:

[0083]

[0084] In a specific embodiment, cluster analysis includes:

[0085] Extract the peak points and trough points of each sample curve;

[0086] Based on the peak points and trough points, construct the feature vector corresponding to the sample curve;

[0087] The hierarchical clustering method is used to cluster each eigenvector and determine the classification data of each sample curve.

[0088] Specifically, the characteristic vectors include: vertical spacing: the vertical distance between the peak and the trough (difference in y coordinates); horizontal spacing: the horizontal distance between consecutive peaks or troughs (difference in x coordinates); maximum slope: the maximum absolute value of the first-order derivative of the curve between the peak and trough points.

[0089] In a specific embodiment, hierarchical fitting includes:

[0090] The sample curves of each category are averaged to obtain the representative curves of each category of sample curves;

[0091] Each representative curve was fitted again, and the fitting error was minimized by optimizing the B-spline control point parameters to obtain a representative curve model, which was used to design the curvature of the pillow.

[0092] The specific average processing is as follows:

[0093] 1. Data alignment: Align sample curves of the same class to the same scale or coordinate system.

[0094] 2. Determine parameters: Determine the spacing parameters, which will be used to compare data points on different curves.

[0095] 3. Curve data sampling: Sample data points from each curve at the same spacing parameter value. This means getting a data point from each curve at the same "position point".

[0096] 4. Calculate the average: For each sample point, calculate the average of the values ​​of all curves at that point. This can be an arithmetic average or a weighted average, depending on the specific situation.

[0097] 5. Construct a representative curve: Use these averages as data points to construct a new curve that represents the "average" shape of the original set of curves.

[0098] The re-fitting is specifically as follows: In order to make the representative curve smoother, further processing is required, and B-spline fitting is used.

[0099] An embodiment of the present invention further discloses a head and neck curve classification model construction system, comprising:

[0100] Coordinate acquisition module: Use a laser rangefinder to measure the coordinate data of each point on the back of the neck in the sagittal plane of the human body to obtain a two-dimensional point set of the head and neck;

[0101] Curve fitting module: Use the fourth-order B-spline fitting method to fit the two-dimensional point set to obtain the head and neck sample curve;

[0102] Cluster analysis module: perform cluster analysis on multiple sample curves to obtain classification data of the sample curves;

[0103] Hierarchical fitting module: Perform hierarchical fitting on each type of sample curve to build a representative curve model.

[0104] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.

[0105] The above description of the disclosed embodiments enables one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for constructing a head and neck curve classification model, characterized in that: The specific steps are as follows: The laser rangefinder is used to measure the coordinate data of each point on the back of the neck in the sagittal plane of the human body to obtain a two-dimensional point set of the head and neck; The two-dimensional point set is fitted using a fourth-order B-spline fitting method to obtain a head and neck sample curve; Performing cluster analysis on the plurality of sample curves to obtain classification data of the sample curves; Perform hierarchical fitting on each type of sample curve to build a representative curve model.

2. The method for constructing a head and neck curve classification model according to claim 1, characterized in that: The laser rangefinder has a plurality of laser rangefinder probes arranged in a row; the coordinate data is determined by measuring the horizontal distance from each point of the neck to the corresponding laser rangefinder probe and the position information of each laser rangefinder probe.

3. The method for constructing a head and neck curve classification model according to claim 1, characterized in that: The fourth-order B-spline fitting method comprises: Based on the coordinate values ​​of the two-dimensional point set, construct a node vector corresponding to the two-dimensional point set; The fourth-order B-spline basis function is obtained by recursively calculating the linear B-spline basis function; Using the fourth-order B-spline basis function to represent the fitting value of each node in the node vector, a linear equation system is established; Solve the linear equations using the least squares method to determine the unknown coefficients of each data point; Based on the undetermined coefficients, a fourth-order B-spline fitting curve is constructed.

4. The method for constructing a head and neck curve classification model according to claim 3, characterized in that: The node vector is represented as u=[u0,u1,...,u k-1 ,u k ,...,u n ,...,u n+k-1 ], where n is the number of data points in the two-dimensional point set, k is the order of the B-spline curve; the starting nodes are u0, u1, ..., u k-1 , the end node is u n ,...,u n+k-1 , the internal node is represented by u j , j=k,k+1,...,n-1; The starting node and the ending node are both repeated nodes.

5. The method for constructing a head and neck curve classification model according to claim 4, characterized in that: The recursive definition of the fourth-order B-spline basis function is expressed as: Where x represents the coordinate value from the shoulder to the head, N j,4 (x) is the fourth-order B-spline basis function; N j,3 (x) is the third-order B-spline basis function, and through the second-order B-spline basis function N j,2 (x) is calculated recursively, and the calculation formula is: The linear B-spline basis function is expressed as:

6. A method for constructing a head and neck curve classification model according to claim 5, characterized in that: The linear equation system is expressed as: Among them, y i ′ represents the fitting value of each data point; c j is the unknown coefficient, corresponding to the weight of the control point of the B-spline curve; N j,4 (x i ) means that in x i The value of the j-th B-spline basis function at point .

7. The method for constructing a head and neck curve classification model according to claim 6, characterized in that: The formula for solving the linear equations using the least squares method is: Among them, y i Represents the actual value of each data point; The fourth-order B-spline fitting curve C(x) is expressed as:

8. The method for constructing a head and neck curve classification model according to claim 1, characterized in that: The cluster analysis comprises: Extracting the peak points and the trough points of each of the sample curves; Based on the peak point and the trough point, construct a feature vector corresponding to the sample curve; The hierarchical clustering method is used to cluster the characteristic vectors to determine the classification data of each sample curve.

9. The method for constructing a head and neck curve classification model according to claim 1, characterized in that: The hierarchical fitting comprises: The sample curves of each category are averaged to obtain the representative curves of each category of sample curves; Each representative curve is fitted again, and the fitting error is minimized by optimizing the B-spline control point parameters to obtain the representative curve model.

10. A head and neck curve classification model construction system, characterized in that: include: Coordinate acquisition module: Use a laser rangefinder to measure the coordinate data of each point on the back of the neck in the sagittal plane of the human body to obtain a two-dimensional point set of the head and neck; Curve fitting module: using a fourth-order B-spline fitting method to fit the two-dimensional point set to obtain a head and neck sample curve; Cluster analysis module: performing cluster analysis on a plurality of the sample curves to obtain classification data of the sample curves; Hierarchical fitting module: Perform hierarchical fitting on each type of sample curve to build a representative curve model.

Images