A kind of intelligent detection system and method of arch morphology

The foot pressure distribution data is collected through pressure sensors, the foot axis is fitted using PCA method and the arch coefficient is calculated, which solves the inefficiency and expensive problems of foot arch morphology detection in the existing technology, and achieves an economical, portable and accurate foot arch Morphological detection.

CN115998285BActive Publication Date: 2024-12-06FUDAN UNIVERSITY +1
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202111220281.8
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2021-10-20
Publication Date
2024-12-06
Estimated Expiration
2041-10-20

AI Technical Summary

Technical Problem

The existing arch morphology detection methods are inefficient, expensive and difficult to use for daily screening, and cannot systematically and quantitatively correct arch abnormalities.

Method used

Using data acquisition, data preprocessing, data fitting and judgment components, the foot pressure distribution data is collected through pressure sensors, the foot axis is fitted using PCA method, and the foot arch coefficient is calculated to judge the arch shape, including data island extraction and filtering , interpolation processing and Graham scanning method and other technologies.

Benefits of technology

It realizes economical, portable and accurate arch morphology detection, reduces noise interference, and can accurately fit the plantar pressure distribution data in any direction, improving the efficiency and accuracy of the detection.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN115998285B_ABST
    Figure CN115998285B_ABST
Patent Text Reader

Abstract

The present invention provides an intelligent detection system for arch morphology, which is used for intelligently detecting arch data so as to judge the arch morphology, and is characterized in that it comprises: a data acquisition unit, which is used to collect original plantar pressure distribution data of a user; a data preprocessing unit, which performs data preprocessing on the original plantar pressure distribution data to obtain preprocessed plantar pressure distribution data and a plantar pressure distribution cloud map; a footprint drawing unit, which scans the plantar pressure distribution cloud map to obtain a convex envelope of the footprint; a data fitting unit, which fits the preprocessed plantar pressure distribution data based on a PCA method to obtain a foot axis; and a correction unit of the data fitting unit, which corrects the foot axis based on the foot axis to obtain a corrected foot axis; a foot division unit, which divides the forefoot, midfoot and hindfoot based on the foot axis or the corrected foot axis; and a judgment unit, which judges the arch morphology of the user.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention belongs to the field of intelligent detection, and relates to a system and method for intelligent detection of arch morphology. Background Art

[0002] The arch of the foot is a convex upward arch structure composed of the tarsal bones and metatarsal bones through ligaments and tendons. It makes the foot strong, light and elastic, and can withstand greater pressure and buffer the vibration caused by walking, running and jumping. It can also protect the blood vessels and nerves in the sole of the foot from compression. Due to the different arch structures, the foot is generally flat, normal and high arched.

[0003] Flat feet and high arches are two common arch deformities, which can cause different degrees of impairment of the arch's static support and dynamic shock absorption functions. In flat feet, the inner side of the arch is often overloaded and accompanied by calcaneal eversion, which can cause foot pain in severe cases and reduce a person's ability to walk and stand; high arches can cause excessive pressure on the forefoot and hindfoot, which can even cause damage to the plantar soft tissue and affect the stability of the ankle joint.

[0004] Currently, functional arch abnormalities are treated conservatively, and orthotic devices (arch pads, etc.) are often used to correct the patient's foot deformity by limiting abnormal inward and outward rotation of the heel, providing support to the arch, and changing the pressure distribution on the plantar surface.

[0005] The existing methods for detecting the arch morphology generally involve doctors judging the arch morphology based on the patient's self-reported symptoms and the measurement results of certain bony markers on the hospital's X-ray film. However, this method is limited by medical institutions and professionals, and is accompanied by certain radiation. It is an inefficient and expensive diagnostic method, and therefore difficult to use for daily arch abnormality screening. Another method is to collect the pressure distribution and three-dimensional morphological data of the foot to detect abnormalities, and then return them to normal as much as possible by designing appropriate devices. However, this method still has the defect that it often cannot achieve this goal systematically and quantitatively. Summary of the invention

[0006] In order to solve the above problems, an economical, portable and accurate intelligent detection method for arch morphology is provided. The present invention adopts the following technical solutions:

[0007] The present invention provides an intelligent detection system for arch morphology, which is used for intelligently detecting arch data so as to judge the arch morphology, and is characterized in that it comprises: a data acquisition unit, which is used to collect original plantar pressure distribution data of a user; a data preprocessing unit, which performs data preprocessing on the original plantar pressure distribution data to obtain preprocessed plantar pressure distribution data and a plantar pressure distribution cloud map; a footprint drawing unit, which scans the plantar pressure distribution cloud map to obtain a convex envelope of the footprint; a data fitting unit, which fits the preprocessed plantar pressure distribution data based on a PCA method to obtain a foot axis; a foot division unit, which divides the forefoot, midfoot and hindfoot based on the foot axis; and a judgment unit, which judges the arch morphology of the user.

[0008] The intelligent detection system for arch morphology provided by the present invention may also have such technical features, wherein the data fitting part also includes a correction unit, which roughly divides the forefoot and the hindfoot based on the foot axis, determines the widest medial and lateral positions of the forefoot and the hindfoot by the rough division, and connects the midpoints of the cross-sections of the widest medial and lateral positions to obtain a corrected foot axis.

[0009] An intelligent detection system for arch morphology provided by the present invention may also have such technical features, wherein the data preprocessing unit has a data island extraction unit and a data processing unit, the data island extraction unit extracts foot pressure data islands whose areas exceed the noise threshold based on the flooding algorithm for the original plantar pressure distribution data containing sensor hardware salt and pepper noise, and the data processing unit filters and interpolates the foot pressure data islands, and generates a plantar pressure distribution cloud map by performing secondary interpolation based on Lanczos.

[0010] The intelligent detection system for arch morphology provided by the present invention may also have such technical features, wherein the PCA method is to perform feature decomposition on the preprocessed plantar pressure distribution data based on the covariance matrix to obtain the corresponding data principal components, i.e., eigenvectors and weights, i.e., eigenvalues; the data fitting part has a feature decomposition unit, a geometric center determination unit, and a foot axis generation unit; the feature decomposition unit obtains the eigenvector C1 and eigenvalue corresponding to the preprocessed plantar pressure distribution data based on the PCA method; the eigenvector C1 indicates the direction of the principal component of the data, which is recorded as the C1 axis; the geometric center determination unit visualizes the preprocessed plantar pressure distribution data to form footprint data, takes the horizontal and vertical coordinates of each point in the footprint data as samples, and takes the average coordinates of all points as the geometric center of the footprint; the foot axis generation unit moves the center of the coordinate axis to the geometric center of the footprint, and continuously rotates the coordinate axis so that the variance of the footprint data on the C1 axis reaches the maximum value to obtain the long axis Ax of the footprint, and the long axis Ax is the foot axis.

[0011] The intelligent detection system for arch morphology provided by the present invention may also have such a technical feature, wherein the judgment unit judges the arch morphology of the user based on the arch coefficient, and the arch coefficient is the ratio of the width of the midfoot contacting part to the overall width of the midfoot, that is:

[0012] λ arch =l1 / l2

[0013] In the formula, λ arch is the arch coefficient, l1 is the width of the midfoot contacting part, l2 is the overall width of the midfoot, when λ arch When λ is greater than the specified coefficient flat-foot , indicating that the footprint at the arch is too wide, the judgment unit judges that the arch shape of the user is flat foot; when λ arch When λ is less than the specified coefficient high-arch , indicating that the footprint at the arch is too narrow or missing, the judgment unit judges that the user's arch shape is a high arch foot; in other cases, the judgment unit judges that the user's arch shape is a normal arch.

[0014] The present invention provides an intelligent detection method for arch morphology, which is used for intelligently detecting arch data so as to judge the arch morphology, and is characterized in that it comprises the following steps: step S1, collecting original plantar pressure distribution data of a user based on a pressure sensor; step S2, performing data preprocessing on the original plantar pressure distribution data to obtain preprocessed plantar pressure distribution data and a plantar pressure distribution cloud map; step S3, scanning the plantar pressure distribution cloud map based on a Graham scanning method to obtain a convex envelope of a footprint; step S4, fitting the preprocessed plantar pressure distribution data based on a PCA method to obtain a foot axis; step S5, dividing the forefoot, midfoot and hindfoot based on the foot axis; step S6, judging the arch morphology of the user based on the contact characteristics between the midfoot area and the ground.

[0015] The intelligent detection method for arch morphology provided by the present invention may also have such technical features, wherein step S4 also includes correcting the foot axis, and the correction includes the following steps: using the foot axis result obtained by fitting the PCA method as the approximate orientation of the foot, and roughly dividing the forefoot and hindfoot; determining the widest inner and outer positions of the forefoot and hindfoot; connecting the midpoints of the cross-sections of the widest positions of the forefoot and hindfoot to obtain a corrected foot axis.

[0016] The intelligent detection method for arch morphology provided by the present invention may also have such technical features, wherein, in step S2, data preprocessing includes the following steps: step S2-1, extracting foot pressure data islands whose areas exceed the noise threshold from the original plantar pressure distribution data containing sensor hardware salt and pepper noise based on the flooding algorithm; step S2-2, filtering and interpolating the extracted foot pressure data islands; step S2-3, improving the resolution of the foot pressure data islands based on Lanczos quadratic interpolation to obtain a plantar pressure distribution cloud map.

[0017] The intelligent detection method for arch morphology provided by the present invention may also have such technical features, wherein the PCA method is to perform feature decomposition on the preprocessed plantar pressure distribution data based on the covariance matrix to obtain the corresponding data principal components, i.e., eigenvectors and weights, i.e., eigenvalues, and step S4 includes the following sub-steps: step S4-1, based on the PCA method, obtain the eigenvector C1 and eigenvalue corresponding to the preprocessed plantar pressure distribution data, the eigenvector C1 indicates the direction of the data principal component, and the direction is recorded as the C1 axis; step S4-2, visualize the preprocessed plantar pressure distribution data to form footprint data, take the horizontal and vertical coordinates of each point in the footprint data as samples, and take the average coordinates of all points as the geometric center of the footprint; step S4-3, move the center of the coordinate axis to the geometric center of the footprint, and continuously rotate the coordinate axis so that the variance of the footprint data on the C1 axis reaches the maximum value to obtain the long axis Ax of the footprint, which is the foot axis.

[0018] The intelligent detection method for arch morphology provided by the present invention may also have such a technical feature, wherein, in step S6, the contact feature is represented by an arch coefficient, and the arch coefficient is the ratio of the width of the midfoot contacting part to the overall width of the midfoot, that is:

[0019] λ arch =l1 / l2

[0020] In the formula, λ arch is the arch coefficient, l1 is the width of the midfoot contacting part, l2 is the overall width of the midfoot, when λ arch When λ is greater than the specified coefficient flat-foot , indicating that the footprint at the arch is too wide, the user's arch shape is judged to be flat foot; when λ arch When λ is less than the specified coefficient high-arch , which means that the footprint at the arch is too narrow or missing, the user's arch shape is judged to be high arch foot; in other cases, the user's arch shape is judged to be normal arch.

[0021] Function and Effect of the Invention

[0022] According to the intelligent detection system and method for arch morphology of the present invention, firstly, a data preprocessing unit performs data preprocessing on the original plantar pressure distribution data collected by the user, reduces the outlier noise of the original plantar pressure distribution data containing the salt and pepper noise of the sensor hardware, and obtains high-quality data with the noise filtered out through filtering and quadratic interpolation; then, a data fitting unit performs feature decomposition on the preprocessed plantar pressure distribution data based on the PCA method to obtain the eigenvector fitting to obtain the foot axis. Since the foot axis obtained by the PCA method is insensitive to the direction of the coordinate axis, the plantar pressure distribution data collected in any direction can be accurately fitted; finally, a judgment unit judges the arch morphology of the user based on the contact characteristics between the divided midfoot area and the ground, namely, the arch coefficient.

[0023] The intelligent detection system and method of arch morphology of the present invention calculates the arch coefficient by fitting the pre-processed plantar pressure distribution data with the foot axis, and judges the arch morphology from the arch coefficient, which is economical, efficient and accurate. The collected foot pressure data not only intuitively reflects the contact between the foot and the ground when the foot is loaded, but also quantitatively reflects the distribution of plantar pressure. The judgment result is also highly consistent with the detection method of arch morphology in the prior art, and is more convenient. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 is a structural block diagram of an intelligent detection system for arch morphology in an embodiment of the present invention;

[0025] Figure 2 is a schematic diagram of preprocessing plantar pressure distribution data in an embodiment of the present invention;

[0026] Figure 3 is a flowchart of the intelligent detection system for arch morphology of an embodiment of the present invention;

[0027] Figure 4 is a flow chart of a method for intelligent detection of arch morphology according to an embodiment of the present invention;

[0028] Figure 5 is a visualized schematic diagram of the intelligent detection process of the arch morphology of the embodiment of the present invention;

[0029] Figure 6 is a schematic diagram comparing intelligent detection results of users with different foot orientations in an embodiment of the present invention;

[0030] Figure 7 Schematic diagram of the foot axis fitted in an embodiment of the present invention. DETAILED DESCRIPTION

[0031] In order to make the technical means, creative features, objectives and effects achieved by the present invention easy to understand, the following is a detailed description of an intelligent detection system and method for arch morphology of the present invention in combination with embodiments and drawings.

[0032] <Example 1>

[0033] Figure 1 It is a structural block diagram of the intelligent detection system for arch morphology in an embodiment of the present invention.

[0034] like Figure 1 As shown, the intelligent detection system 10 for arch morphology includes: a data acquisition unit 11, a data preprocessing unit 12, a footprint drawing unit 13, a data fitting unit 14, a foot division unit 15 and a judgment unit 16.

[0035] The data acquisition unit 11 is used to collect the original plantar pressure distribution data of the user.

[0036] In this embodiment, data is collected through a pressure sensor, and the user can stand still on the pressure sensor in any posture.

[0037] The data preprocessing unit 12 performs data preprocessing on the original plantar pressure distribution data to obtain preprocessed plantar pressure distribution data and a plantar pressure distribution cloud map.

[0038] The data preprocessing unit 12 includes a data island extracting unit 121 and a data processing unit 122 .

[0039] The data island extraction unit 121 extracts the foot pressure data islands whose areas exceed the noise threshold based on the flooding algorithm from the original foot pressure distribution data containing the sensor hardware salt and pepper noise.

[0040] The data processing unit 122 performs filtering and interpolation processing on the foot pressure data islands, and performs secondary interpolation based on Lanczos to complete the preprocessing of the original plantar pressure distribution data and generate a plantar pressure distribution cloud map.

[0041] Figure 2 It is a visualization schematic diagram of preprocessing plantar pressure distribution data in an embodiment of the present invention.

[0042] like Figure 2 As shown, from left to right are the original plantar pressure distribution data, the footprint data island and the plantar pressure distribution cloud map.

[0043] The footprint drawing unit 13 scans the sole pressure distribution cloud map to obtain the convex envelope of the footprint.

[0044] In this embodiment, the convex envelope of the footprint is obtained based on the Graham scanning algorithm. Figure 3As shown in a, the convex envelope is the convex polygon with the smallest area included in the entire footprint.

[0045] The data fitting unit 14 fits the preprocessed plantar pressure distribution data based on the PCA method to obtain the foot axis.

[0046] In this embodiment, the PCA method performs eigendecomposition on the preprocessed plantar pressure distribution data based on the covariance matrix to obtain the corresponding data principal components, namely eigenvectors, and weights, namely eigenvalues.

[0047] The data fitting unit 14 includes a feature decomposition unit 141 , a geometric center determination unit 142 , and a foot axis generation unit 143 .

[0048] The feature decomposition unit 141 obtains the feature vector C1 and feature value corresponding to the preprocessed plantar pressure distribution data based on the PCA method. The feature vector C1 indicates the direction of the main component of the data, and the direction is recorded as the C1 axis.

[0049] The geometric center determination unit 142 visualizes the preprocessed plantar pressure distribution data to form footprint data, takes the horizontal and vertical coordinates of each point in the footprint data as samples, and takes the average coordinates of all points as the geometric center of the footprint.

[0050] The foot axis generation unit 143 moves the center of the coordinate axis to the geometric center of the footprint, and continuously rotates the coordinate axis so that the variance of the footprint data on the C1 axis reaches the maximum value to obtain the long axis Ax of the footprint, which is the foot axis.

[0051] In this embodiment, the foot axis obtained by fitting based on the PCA method is insensitive to the direction of the coordinate axis, so the plantar pressure data collected in any direction can be accurately fitted, that is, the patient's standing position and the direction of both feet do not need to be limited during data collection.

[0052] The foot dividing unit 15 divides the foot into the forefoot, the midfoot, and the rearfoot based on the foot axis.

[0053] In this embodiment, once the foot axis is determined, the direction of the footprint is determined, thereby determining the positions of the forefoot, midfoot and hindfoot regions.

[0054] The determination unit 16 determines the shape of the arch of the user's foot.

[0055] The shape of the arch is closely related to the contact characteristics of the midfoot area with the ground. If the contact area accounts for too small a proportion or there is no contact area, it indicates a tendency towards high arches. Conversely, if the proportion is too large, it indicates a tendency towards flat feet.

[0056] In this embodiment, the width of the midfoot contacting part is determined by drawing a perpendicular bisector of the foot axis and calculating the length of the intersection of the straight line and the footprint. The overall width of the midfoot can be regarded as the projection of the midfoot on the horizontal plane, and the overall width of the midfoot can be approximated by the convex envelope of the footprint.

[0057] The ratio of the width of the midfoot contacting part to the overall width of the midfoot is used as the arch coefficient, and the judging unit 16 judges the arch shape of the user based on the arch coefficient. Specifically:

[0058] The arch coefficient is the ratio of the width of the midfoot contacting the ground to the overall width of the midfoot, that is:

[0059] λ arch =l1 / l2

[0060] In the formula, λ arch is the arch coefficient, l1 is the width of the midfoot contacting part, and l2 is the overall width of the midfoot.

[0061] The judgment rules are:

[0062] When arch When λ is greater than the specified coefficient flat-foot , indicating that the footprint at the arch is too wide, the determination unit 16 determines that the user's arch shape is flat foot;

[0063] When arch When λ is less than the specified coefficient high-arch , indicating that the footprint at the arch is too narrow or missing, the determination unit 16 determines that the user's arch shape is a high arch foot;

[0064] In other cases, the determination unit 16 determines that the arch shape of the user is a normal arch.

[0065] The specific working process of the intelligent detection system 10 for arch morphology in the above process is described below in conjunction with the accompanying drawings.

[0066] Figure 3 It is a working flow chart of the intelligent detection system for arch morphology of an embodiment of the present invention.

[0067] like Figure 3 As shown, in the first embodiment, the workflow of the intelligent detection system 10 for arch morphology includes the following steps:

[0068] Step A1, the data acquisition unit 11 collects the original plantar pressure distribution data of the user, and then proceeds to step A2;

[0069] Step A2, the data preprocessing unit 12 performs data preprocessing on the original plantar pressure distribution data to obtain preprocessed plantar pressure distribution data and a plantar pressure distribution cloud map, and then proceeds to step A3;

[0070] Step A3, the footprint drawing unit 13 scans the plantar pressure distribution cloud map based on the Graham scanning method to obtain the convex envelope of the footprint, and then proceeds to step A4;

[0071] Step A4, the data fitting unit 14 fits the preprocessed plantar pressure distribution data based on the PCA method to obtain the foot axis, and then proceeds to step A5;

[0072] Step A5, the foot division unit 15 divides the forefoot, the midfoot and the hindfoot based on the foot axis, and then proceeds to step A6;

[0073] In step A6, the determination unit 16 determines the arch shape of the user based on the contact characteristics between the midfoot area and the ground.

[0074] In the above step A2, the data preprocessing includes extracting the foot pressure data islands based on the flooding algorithm by the data island extraction unit 121 and filtering and interpolating the foot pressure data islands by the data processing unit 122 and performing secondary interpolation based on Lanczos.

[0075] Figure 4 It is a flow chart of the intelligent detection method of arch morphology according to an embodiment of the present invention.

[0076] like Figure 4 As shown, this embodiment provides a method for intelligent detection of arch morphology, comprising the following steps:

[0077] Step S1, collecting original plantar pressure distribution data of the user based on a pressure sensor.

[0078] Step S2, preprocessing the original plantar pressure distribution data to obtain preprocessed plantar pressure distribution data and a plantar pressure distribution cloud map.

[0079] The above step S2 also includes the following sub-steps:

[0080] Step S2-1, extracting foot pressure data islands whose areas exceed the noise threshold from the original plantar pressure distribution data containing sensor hardware salt and pepper noise based on a flooding algorithm;

[0081] Step S2-2, filtering and interpolating the extracted foot pressure data islands;

[0082] Step S2-3, based on Lanczos quadratic interpolation, the resolution of the foot pressure data island is improved to obtain a foot pressure distribution cloud map.

[0083] Step S3, scanning the plantar pressure distribution cloud map based on the Graham scanning method to obtain the convex envelope of the footprint.

[0084] Step S4, fitting the preprocessed plantar pressure distribution data based on the PCA method to obtain the foot axis.

[0085] The above step S4 also includes the following sub-steps:

[0086] Step S4-1, obtaining a characteristic vector C1 and a characteristic value corresponding to the preprocessed plantar pressure distribution data based on the PCA method, wherein the characteristic vector C1 indicates the direction of the principal component of the data, and the direction is recorded as the C1 axis;

[0087] Step S4-2, visualizing the preprocessed plantar pressure distribution data to form footprint data, taking the horizontal and vertical coordinates of each point in the footprint data as samples, and taking the average coordinates of all points as the geometric center of the footprint;

[0088] Step S4-3, move the center of the coordinate axis to the geometric center of the footprint, and continuously rotate the coordinate axis so that the variance of the footprint data on the C1 axis reaches the maximum value to obtain the long axis Ax of the footprint, which is the foot axis.

[0089] Step S5, dividing the forefoot, midfoot and hindfoot based on the foot axis.

[0090] Step S6, judging the arch shape of the user based on the contact characteristics between the midfoot area and the ground.

[0091] In the above step S6, the contact feature is represented by the arch coefficient, which is the ratio of the width of the midfoot contacting part to the overall width of the midfoot, that is:

[0092] λ arch =l1 / l2

[0093] In the formula, λ arch is the arch coefficient, l1 is the width of the midfoot contacting the ground, and l2 is the overall width of the midfoot. The specific judgment process is:

[0094] When arch When λ is greater than the specified coefficient flat-foot , which means that the footprint at the arch is too wide, and the user's arch shape is judged to be flat foot;

[0095] When arch When λ is less than the specified coefficient high-arch , which means that the footprint at the arch is too narrow or missing, and the user's arch shape is judged to be high arch foot;

[0096] In other cases, the user's arch shape is judged to be a normal arch.

[0097] Figure 5 It is a visualized schematic diagram of the intelligent detection process of the arch morphology of the embodiment of the present invention.

[0098] like Figure 5 As shown, from left to right are the convex envelope of the footprint obtained in step S3, the foot axis obtained by fitting in step S4, the division of the foot in step S5, and the width l1 of the midfoot contacting part and the overall width l2 of the midfoot calculated in step S6.

[0099] Figure 6 4 is a schematic diagram comparing intelligent detection results of users with different foot orientations in an embodiment of the present invention.

[0100] In this embodiment, the arch morphology of subject A and subject B is detected based on the intelligent detection method of arch morphology, such as Figure 6 As shown, the intelligent detection results of the two subjects when standing in a relaxed posture and standing with the foot axis parallel to each other are similar. Both can accurately determine the direction of the footprint and divide the positions of the forefoot, midfoot and hindfoot areas, so that the judgment results are not affected.

[0101] Example Function and Effect

[0102] According to the intelligent detection system and method for arch morphology provided in this embodiment, the data preprocessing unit first performs data preprocessing on the original plantar pressure distribution data collected by the user to reduce the outlier noise of the original plantar pressure distribution data containing sensor hardware salt and pepper noise, and obtains high-quality data with noise filtered out through filtering and secondary interpolation; then, the data fitting unit performs feature decomposition on the preprocessed plantar pressure distribution data based on the PCA method to obtain the eigenvector fitting to obtain the foot axis. Since the foot axis obtained by the PCA method is insensitive to the direction of the coordinate axis, the plantar pressure distribution data collected in any direction can be accurately fitted; finally, the judgment unit judges the user's arch morphology based on the contact characteristics between the divided midfoot area and the ground, that is, the arch coefficient.

[0103] The intelligent detection system and method of arch morphology of the present embodiment calculates the arch coefficient by fitting the pre-processed plantar pressure distribution data with the foot axis, and determines the arch morphology from the arch coefficient, which is economical, efficient and accurate. The collected foot pressure data not only intuitively reflects the contact between the foot and the ground when the foot is loaded, but also quantitatively reflects the distribution of plantar pressure. The judgment result is also highly consistent with the detection method of arch morphology in the prior art, and is more convenient.

[0104] <Example 2>

[0105] The difference between the second embodiment and the first embodiment is that the data fitting unit 14 further includes a correction unit 144, and step S4 of the intelligent detection method for arch morphology further includes correcting the foot axis.

[0106] For ease of description, in this modification example 1, the same symbols are given to the same structures as those in the embodiment and the same descriptions are omitted.

[0107] Since the PCA method cannot utilize the structural information between the points that constitute the footprint, although the first embodiment can fit the foot axis based on this method to generate the foot axis to determine the overall orientation of the footprint pattern, it has the possibility of further reducing the information entropy. Therefore, the second embodiment proposes a correction based on the foot axis fitting based on the PCA method to address this problem. Specifically:

[0108] The correction unit 144 uses the foot axis fitted in the first embodiment as the approximate direction of the foot, roughly divides the forefoot and the hindfoot, determines the widest medial and lateral positions of the forefoot and the hindfoot by the rough division, and connects the midpoints of the cross-sections of the widest medial and lateral positions to obtain a corrected foot axis.

[0109] Step S4 of the intelligent detection method for arch morphology also includes the following steps:

[0110] Firstly, the foot axis obtained by PCA fitting is used as the approximate orientation of the foot to roughly divide the forefoot and hindfoot.

[0111] Then, determine the widest medial and lateral locations of the forefoot and hindfoot.

[0112] Finally, the midpoints of the cross-sections at the widest positions of the forefoot and hindfoot were connected to obtain the corrected foot axis.

[0113] This correction method analyzes the metatarsal row and the two ends of the calcaneus based on the external features of the footprint (the widest part of the front and back feet), and estimates the position of the calcaneal center and the second metatarsophalangeal joint accordingly, thereby achieving automated foot axis fitting that is closer to clinical manual fitting results.

[0114] In the second embodiment, the foot axis obtained by fitting based on the PCA method in the first embodiment and the foot axis corrected based on the PCA method in the second embodiment are compared with the manually marked foot axis, wherein the footprint obtained by pressure collection basically and completely includes the projection of the heel and metatarsal area. The line connecting the center points of the heel and metatarsal area is taken as the foot axis direction of the manual marker, and extended to the edge of the footprint. The angle between the two foot axes and the bidirectional Hausdorff distance are calculated. Specifically:

[0115] The Hausdorff distance measures the similarity of two curves. The smaller the value, the higher the similarity. Let A and B be the axes obtained by two methods, and a and b be any point on the two axes. Then the bidirectional Hausdorff distance H(A,B) is:

[0116]

[0117] H(A,B)=max[h(A,B),h(B,A)]

[0118] Figure 7 Schematic diagram of the foot axis fitted in an embodiment of the present invention.

[0119] In the second embodiment, 9 subjects were recruited for the experiment. The foot axis line obtained by the PCA method and the foot axis line obtained by the PCA and correction method were used as shown in the figure. Figure 7 As shown, the straight line represents the foot axis obtained by fitting based on the PCA method, and the dotted line segment represents the foot axis obtained by fitting based on the PCA and correction method.

[0120] The results of the experiment showed that the results of the two methods of fitting the foot axis are very close to the foot axis determined by the clinical method, with an angle deviation of less than 2.5° and a bidirectional Hausdorff distance of less than 9mm, indicating that within the footprint, the distance between the obtained foot axis and any point on the reference axis and another axis does not exceed 9mm. It is noted that the foot axis obtained based on the modified PCA algorithm is closer to the clinically defined foot length axis, with an angle deviation of less than 1°. Moreover, the accuracy of the arch morphology results obtained by intelligent detection and judgment of the arch morphology of these 9 subjects reached more than 88%.

[0121] In summary, the intelligent detection method of arch morphology of the present invention has a high degree of consistency with the manual detection and judgment method, has a high accuracy rate, and has a high reference value for the detection of flat feet and high arches. In Example 2, the foot axis obtained by fitting based on the PCA method is also corrected. The method of this Example 2 uses the bony landmarks of the footprint to make targeted corrections, so that the accuracy of the foot axis is effectively improved. This algorithm combines the universality of PCA for pressure footprint fitting and the accuracy of determining the foot axis based on bony characteristics, so that the corrected foot axis is closer to the real foot axis manually marked, the angle deviation is smaller, and the plantar partition has higher reliability, so that the judgment result of the arch morphology is also more accurate.

Claims

1. An intelligent detection system for arch morphology, used for intelligently detecting arch data to judge the arch morphology, characterized in that: include: A data acquisition unit, used for collecting the original plantar pressure distribution data of the user; A data preprocessing unit performs data preprocessing on the original plantar pressure distribution data to obtain preprocessed plantar pressure distribution data and a plantar pressure distribution cloud map; A footprint drawing unit, scanning the plantar pressure distribution cloud map to obtain a convex envelope of the footprint; A data fitting unit, fitting the preprocessed plantar pressure distribution data based on the PCA method to obtain the foot axis; A foot division section, which divides the foot into a forefoot, a midfoot, and a hindfoot based on the foot axis; as well as The judging unit judges the arch shape of the user.

2. The intelligent detection system for arch morphology according to claim 1, characterized in that: in, The data fitting unit further includes a correction unit. The correction unit roughly divides the forefoot and the hindfoot based on the foot axis, determines the widest medial and lateral positions of the forefoot and the hindfoot by the rough division, and connects the midpoints of the sections of the widest medial and lateral positions to obtain a corrected foot axis.

3. The intelligent detection system for arch morphology according to claim 1, characterized in that: in, The data preprocessing unit comprises a data island extraction unit and a data processing unit. The data island extraction unit extracts the foot pressure data islands whose areas exceed the noise threshold based on the flooding algorithm from the original plantar pressure distribution data containing the sensor hardware salt and pepper noise. The data processing unit performs filtering and interpolation processing on the foot pressure data island, and generates the foot pressure distribution cloud map by performing secondary interpolation based on Lanczos.

4. The intelligent detection system for arch morphology according to claim 1, characterized in that: in, The PCA method is to perform eigendecomposition on the preprocessed plantar pressure distribution data based on the covariance matrix to obtain the corresponding data principal components, i.e., eigenvectors and weights, i.e., eigenvalues. The data fitting unit comprises a feature decomposition unit, a geometric center determination unit and a foot axis generation unit. The feature decomposition unit obtains the feature vector C1 and feature value corresponding to the preprocessed plantar pressure distribution data based on the PCA method, wherein the feature vector C1 indicates the direction of the principal component of the data, and the direction is recorded as the C1 axis. The geometric center determination unit visualizes the preprocessed plantar pressure distribution data to form footprint data, takes the horizontal and vertical coordinates of each point in the footprint data as samples, and takes the average coordinates of all points as the geometric center of the footprint. The foot axis generation unit moves the center of the coordinate axis to the geometric center of the footprint, and continuously rotates the coordinate axis so that the variance of the footprint data on the C1 axis reaches the maximum value to obtain the long axis Ax of the footprint, which is the foot axis.

5. The intelligent detection system for arch morphology according to claim 1, characterized in that: in, The judging unit judges the arch shape of the user based on the arch coefficient. The arch coefficient is the ratio of the width of the midfoot contacting portion to the overall width of the midfoot, that is: λ arch =l1 / l2 In the formula, λ arch is the arch coefficient, l1 is the width of the midfoot contacting part, l2 is the overall width of the midfoot, When arch When λ is greater than the specified coefficient flat-foot , indicating that the footprint at the arch is too wide, the determination unit determines that the arch shape of the user is flat foot; When arch When λ is less than the specified coefficient high-arch , indicating that the footprint at the arch is too narrow or missing, the determination unit determines that the arch shape of the user is a high arch foot; In other cases, the determination unit determines that the arch shape of the user is a normal arch.

6. A method for intelligently detecting the shape of a foot arch, for intelligently detecting the foot arch data to judge the shape of the foot arch, characterized in that: The following steps are involved: Step S1, collecting original plantar pressure distribution data of the user based on the pressure sensor; Step S2, performing data preprocessing on the original plantar pressure distribution data to obtain preprocessed plantar pressure distribution data and a plantar pressure distribution cloud map; Step S3, scanning the plantar pressure distribution cloud map based on the Graham scanning method to obtain the convex envelope of the footprint; Step S4, fitting the preprocessed plantar pressure distribution data based on the PCA method to obtain the foot axis; Step S5, dividing the forefoot, midfoot and hindfoot based on the foot axis; Step S6, judging the arch shape of the user based on the contact characteristics between the midfoot area and the ground.

7. The intelligent detection method for arch morphology according to claim 6, Features: Wherein, the step S4 also includes correcting the foot axis. The modification comprises the following steps: The foot axis result obtained by the PCA method is used as the approximate direction of the foot to roughly divide the forefoot and the hindfoot; determining the widest medial and lateral positions of the front foot and the rear foot; The midpoints of the cross-sections at the widest positions of the forefoot and the hindfoot are connected to obtain a corrected foot axis.

8. The intelligent detection method for arch morphology according to claim 6, characterized in that: in, In step S2, the data preprocessing includes the following sub-steps: Step S2-1, extracting foot pressure data islands whose areas exceed a noise threshold from the original foot pressure distribution data containing sensor hardware salt and pepper noise based on a flooding algorithm; Step S2-2, filtering and interpolating the extracted foot pressure data island; Step S2-3, improving the resolution of the foot pressure data island based on Lanczos quadratic interpolation to obtain the foot pressure distribution cloud map.

9. The intelligent detection method for arch morphology according to claim 6, Features: The PCA method is to perform feature decomposition on the pre-processed plantar pressure distribution data based on the covariance matrix to obtain the corresponding data principal components, i.e., eigenvectors and weights, i.e., eigenvalues. The step S4 comprises the following sub-steps: Step S4-1, obtaining a characteristic vector C1 and an characteristic value corresponding to the preprocessed plantar pressure distribution data based on the PCA method, wherein the characteristic vector C1 indicates the direction of the principal component of the data, and the direction is recorded as the C1 axis; Step S4-2, visualizing the preprocessed plantar pressure distribution data to form footprint data, taking the horizontal and vertical coordinates of each point in the footprint data as samples, and taking the average coordinates of all points as the geometric center of the footprint; Step S4-3, move the center of the coordinate axis to the geometric center of the footprint, and continuously rotate the coordinate axis so that the variance of the footprint data on the C1 axis reaches the maximum value to obtain the long axis Ax of the footprint, which is the foot axis.

10. The intelligent detection method of arch morphology according to claim 6, characterized in that: in, In step S6, the contact feature is represented by an arch coefficient. The arch coefficient is the ratio of the width of the midfoot contacting portion to the overall width of the midfoot, that is: λ arch =l1 / l2 In the formula, λ arch is the arch coefficient, l1 is the width of the midfoot contacting part, l2 is the overall width of the midfoot, When arch When λ is greater than the specified coefficient flat-foot , indicating that the footprint at the arch is too wide, and the user's arch shape is judged to be flat foot; When arch When λ is less than the specified coefficient high-arch , indicating that the footprint at the arch is too narrow or missing, and the user's arch shape is judged to be high arch foot; In other cases, the arch shape of the user is determined to be a normal arch.