Pre-foot-rotation estimation device, pre-foot-rotation estimation method, and storage medium

Abstract

The present invention provides a technique for easily discriminating a pre-rotation type, which includes a pre-rotation estimation device, a pre-rotation estimation method, and a storage medium. In a foot pre-pronation estimation device (50), a pre-pronation estimation unit (72) estimates the degree of pronation of a foot on the basis of a comparison between a feature quantity of a predetermined site and a reference value. A feature amount extraction unit (60) extracts feature amounts of a plurality of locations including at least the length of the foot width and the amount of depression in the arch shape from a top view image. The pronation estimation unit (72) estimates the degree of pronation of the foot on the basis of the comparison between the characteristic quantities of a plurality of locations including at least the length of the foot width and the amount of depression in the arch shape and the respective reference values. The feature quantities of the plurality of locations further include the foot length and the amount of protrusion in the outer contour in the vicinity of the toenail. A precession estimation unit (72) estimates which of a first type in which the degree of precession is higher than a predetermined reference and a second type in which the degree of precession is equal to or lower than the predetermined reference is satisfied as the type classified according to the degree of precession of the foot.

Description

Technical Field

[0001] The present invention relates to a foot pronation estimation device, a foot pronation estimation method and a storage medium, and more particularly to a technology for analyzing the state of a foot based on an image of a foot shape. Background Art

[0002] Some running shoes are sometimes classified according to the type of pronation (internal rotation type). As a classification method, they are often divided into runners who overpronate (over pronation) and runners who do not (natural pronation, under pronation). By using running shoes that are suitable for their own pronation type, runners can maintain moderate pronation and reduce the risk of accidents.

[0003] Here, there is known a technique for measuring the footprint or the pressure distribution of the foot to estimate the arch height index or the flexibility of the customer's foot and select a type of shoe suitable for the customer (for example, see Patent Document 1).

[0004] [Prior art literature]

[0005] [Patent Document]

[0006] [Patent Document 1] International Publication No. 2005 / 006905 Summary of the invention

[0007] [Problems to be solved by the invention]

[0008] To select running shoes according to the pronation type, it is necessary to be able to recognize which pronation type one has. However, to determine the pronation type, it is usually necessary to shoot the running on the treadmill from behind to analyze the pace, and it is necessary to use a facility with equipment for such shooting or analysis. Alternatively, in the pronation type determination technology in Patent Document 1, although it can be inferred based on the footprint or the pressure distribution of the foot, it requires the use of a measuring machine such as an optical sensor or a pressure sensor, so it may not be easy for ordinary runners to determine.

[0009] On the other hand, since shoe manufacturers classify running shoes according to pronation types, it is ideal to avoid type mismatch such as runners who do not overpronate using shoes designed for overpronation. In particular, as online shoe sales have become more common, it is desirable to establish a method that can easily determine the pronation type without dedicated equipment.

[0010] The present disclosure has been made in view of the above situation, and an object of the present invention is to provide a technology for easily distinguishing the type of pronation.

[0011] [Technical means to solve the problem]

[0012] In order to solve the above-mentioned problem, a foot pronation estimation device according to an embodiment of the present disclosure includes: an image acquisition unit, which acquires at least a top-view image of the foot as an image representing the contour shape of the foot of a measurement subject when viewed from above; a feature quantity extraction unit, which extracts a feature quantity of a specified part representing the top-view contour shape of the foot from the top-view image; and a pronation estimation unit, which estimates the degree of pronation of the foot based on a comparison between the feature quantity of the specified part and a reference value.

[0013] Another embodiment of the present disclosure is a method for estimating pronation of a foot. The method includes the following steps: obtaining at least a top view image of a foot as an image representing the contour shape of a foot of a measurement subject when viewed from above; extracting a feature value of a predetermined portion representing the top view contour shape of the foot from the top view image by image processing; and estimating the degree of pronation of the foot based on a comparison process of the feature value of the predetermined portion with a reference value.

[0014] [Effects of the Invention]

[0015] According to the present disclosure, it is possible to provide a technique for easily distinguishing the type of pronation. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 It is a diagram showing the basic structure of the foot pronation estimation system.

[0017] Figure 2 This is a functional block diagram showing each structure of the pronation estimation device.

[0018] Figure 3 This is a diagram showing a method for measuring foot length and foot width.

[0019] Figure 4 It is a diagram showing a method for measuring the amount of inner depression and the amount of outer protrusion.

[0020] Figure 5 is a graph showing the relationship between pronation type and shoe type.

[0021] Figure 6 This is a diagram showing the comparison results between the classification results of the subjects obtained by the decision tree analysis and the classification results of the pronation types based on the three-dimensional eversion angle.

[0022] Figure 7 This is a tree diagram schematically showing the result of a decision tree analysis of foot pronation estimation based on three factors: foot length, foot width, and medial concavity.

[0023] Figure 8 The flowchart shows the procedure of the pronation estimation process using the pronation estimation algorithm based on the three elements of foot length, foot width, and medial concavity.

[0024] Fig. 9 This is a tree diagram schematically showing the result of a decision tree analysis of foot pronation estimation based on three factors: foot width, medial concavity, and lateral protrusion.

[0025] Fig.10 The flowchart shows the procedure of the pronation estimation process using the pronation estimation algorithm based on the three elements of the foot width, the medial concavity, and the lateral protrusion.

[0026] [Explanation of Symbols]

[0027] 10: Measurement subject

[0028] 30: Top view image

[0029] 31: Foot length

[0030] 32: Foot width

[0031] 38: Amount of medial concavity

[0032] 39: Outer protrusion

[0033] 44: Arch shape

[0034] 45: The outer shape near the little toe

[0035] 50: Foot pronation prediction device

[0036] 54: Image acquisition unit

[0037] 60: Feature extraction unit

[0038] 72: Pronation Prediction

[0039] 74: Type determination unit

[0040] 80: Storage

[0041] 110: Type 1

[0042] 111: Type 2

[0043] 120: Pronation type

[0044] 121: Shoe type DETAILED DESCRIPTION

[0045] Hereinafter, based on the embodiment, Figure 1 In the embodiments and modifications, the same or equivalent components are denoted by the same reference numerals, and duplicate descriptions are omitted as appropriate.

[0046] In the present embodiment, a feature quantity representing the top view contour shape of the foot is extracted from a top view image of the foot of the measurement subject at the seat, and the degree of pronation of the foot is estimated based on a comparison with a reference value. The top view image is an image representing the contour shape of the measurement subject's foot when viewed from above. The top view image includes the contour shape of the entire foot, such as the toe shape or the heel shape, so it is possible to extract a feature quantity of the foot part represented by the foot length or the foot width from the top view image. As a method for estimating the pronation type based on the feature quantity of the foot part, the following method is adopted: a decision tree is used to pre-analyze the relationship between the feature quantity of the foot part measured for multiple subjects and the pronation type, and an estimation algorithm based on the analysis result is used to perform the estimation.

[0047] Figure 1 The basic structure of the foot pronation estimation system 100 is shown. The foot pronation estimation system 100 includes a device for acquiring a top view image of a foot, such as an information terminal 16 or a three-dimensional foot prosthesis measuring device 18, and a foot pronation estimation device 50. For example, it is assumed that a runner who is considering purchasing running shoes classified according to the pronation type is set as a measurement subject 10, and the measurement subject 10 uses the information terminal 16 or the three-dimensional foot prosthesis measuring device 18 to determine the pronation type of his or her own foot. The measurement subject 10 acquires a top view image of each foot, and determines the pronation type of each left and right foot.

[0048] The information terminal 16 may be a portable information terminal such as a smart phone or a tablet terminal, or a personal computer. The information terminal 16 includes, for example, a combination of hardware such as a camera module, a distance sensor, a microprocessor, a touch panel, a memory, and a communication module. The information terminal 16 may be a general-purpose information terminal such as a smart phone or a tablet terminal prepared by the measurement subject 10 or a shoe sales store. In the information terminal 16, a website provided by the pronation estimation device 50 is accessed via a web browser, or information provided by the pronation estimation device 50 is displayed on a screen through application software running on the information terminal 16.

[0049] When using the information terminal 16, the measurement subject 10 uses a camera built into the information terminal 16 such as a mobile phone terminal to capture the feet of the measurement subject 10 at the seat to obtain an image. For example, in the case of a method of capturing the entire foot from directly above the foot, even if the heel is fully captured, it is difficult to capture the entire foot contour shape in one image because the ankle is captured. Therefore, the premise may be that the image is captured in such a way that at least the contours of the toes and heels, and the contours of the left and right sides from the forefoot to the midfoot are captured.

[0050] In the case where the information terminal 16 has a three-dimensional scanner function using a technology such as LiDAR, it is also possible to scan the periphery of the foot to generate a three-dimensional pattern of the foot shape, and obtain a top-view image as a two-dimensional image from the three-dimensional pattern. Even in the case where the information terminal 16 does not have a three-dimensional scanner function based on LiDAR, etc., it is also possible to generate a three-dimensional pattern of the foot shape through image synthesis processing such as photogrammetry, and obtain a top-view image as a two-dimensional image from the three-dimensional pattern. The person being measured 10 can also place his foot on a dedicated measurement mat 12, use the camera function of the information terminal 16 and the foot shape acquisition application to scan his own foot shape, and generate a three-dimensional pattern of the foot shape.

[0051] When the three-dimensional foot shape measuring device 18 is used, the person 10 to be measured may generate a three-dimensional shape of the foot using the three-dimensional foot shape measuring device 18 installed in a shoe store, for example, and obtain a top view image as a two-dimensional image from the three-dimensional pattern. The three-dimensional foot shape measuring device 18 obtains three-dimensional data of the foot shape by laser measurement. The measurement value as the three-dimensional measurement result of the foot shape scanned by the three-dimensional foot shape measuring device 18 may be sent from the three-dimensional foot shape measuring device 18 to the foot pronation estimation device 50, and the foot pronation estimation device 50 generates a top view image as a two-dimensional image from the three-dimensional pattern.

[0052] In a modified example, a side view image may be acquired instead of a top view image of the foot of the measurement subject 10. Alternatively, both a top view image and a side view image of the foot of the measurement subject 10 may be acquired. Alternatively, a top view image or a side view image of the foot of the measurement subject 10 when standing may be acquired.

[0053] The pronation estimation device 50 is connected to a plurality of information terminals 16 or three-dimensional foot shape measuring devices 18 via network lines such as the Internet or a local area network (LAN) and communication methods such as wireless communication. The pronation estimation device 50 can also be implemented as a server that estimates the pronation type of the foot based on the measurement data sent from a plurality of information terminals 16 or three-dimensional foot shape measuring devices 18. The pronation estimation device 50 can include a single server computer or a combination of multiple server computers. In addition, the "pronation estimation device" mentioned in the claims can refer to the pronation estimation system 100 as a whole or the pronation estimation device 50. Moreover, in the present embodiment, most of the characteristic functions contained in the "pronation estimation device" mentioned in the claims are implemented in the form possessed by the pronation estimation device 50, so the pronation estimation device 50 is substantially equivalent to the "pronation estimation device" in the claims. However, in a modified example, it can also be implemented in the following manner, that is, the "foot pronation estimation device" mentioned in the claims is constituted by the collaboration of the foot pronation estimation device 50 and the information terminal 16, for example, it is constituted so that the information terminal 16 performs processing such as acquiring overhead images or extracting feature quantities of foot parts, outputting estimation results of foot pronation types, etc.

[0054] Figure 2 2 is a functional block diagram showing the components of the pronation estimation device 50 . Figure 2 In the figure, regarding the pronation estimation device 50, functional blocks that can be realized by the linkage of various hardware structures and software structures are described. Those skilled in the art should understand that these functional blocks can be realized in various forms by hardware only, software only, or a combination thereof. The pronation estimation device 50, for example, includes a combination of hardware such as a microprocessor, a memory, a display, and a communication module. The pronation estimation device 50 can be a server computer constructed and managed by a shoe manufacturer. Among them, in the pronation estimation device 50, a program with the following functions is running. The pronation estimation device 50 includes a communication unit 52, an image acquisition unit 54, a feature extraction unit 60, an evaluation unit 70, a storage unit 80, an output unit 90, and an analysis and processing unit 99 as its functions. The communication unit 52 of the pronation estimation device 50 is connected to the information terminal 16 or the three-dimensional foot shape measuring device 18 via a network.

[0055] The image acquisition unit 54 acquires the overhead image sent from the information terminal 16 or the three-dimensional foot shape measuring device 18 via the communication unit 52. The image acquisition unit 54 acquires an image representing the overhead contour shape of the feet of the measurement subject 10 at the seat as the overhead image. The image acquisition unit 54 acquires overhead images for each of the left and right feet. However, it is not necessary to acquire images of both left and right feet, and an image of either left or right foot may be acquired. In a modified example, the image acquisition unit 54 may also acquire overhead images and side images of the feet of the measurement subject 10. Moreover, in a modified example, the image acquisition unit 54 may also acquire overhead images or side images of the feet in standing.

[0056] The image acquisition unit 54 includes a measurement value acquisition unit 56 and an image generation unit 58. When the three-dimensional measurement value of the foot shape is transmitted from the information terminal 16 or the three-dimensional foot shape measuring device 18 instead of the overhead image, the measurement value acquisition unit 56 acquires the three-dimensional measurement value via the communication unit 52. In this case, the image generation unit 58 generates a three-dimensional pattern based on the three-dimensional measurement value, and generates the overhead image as a two-dimensional image from the three-dimensional pattern. The technology of generating a three-dimensional pattern from the three-dimensional measurement value is known, so its description is omitted.

[0057] The feature extraction unit 60 extracts feature values ​​of multiple foot parts from the top view image. The feature values ​​of the multiple foot parts are at least any one of (1) foot length, (2) length of foot width, (3) amount of depression in the shape of the arch (hereinafter referred to as "inner depression"), and (4) amount of protrusion in the outer shape near the little toe (hereinafter referred to as "outer protrusion"). The feature extraction unit 60 includes a foot length calculation unit 62, a foot width calculation unit 64, a depression calculation unit 66, and a protrusion calculation unit 68. The foot length calculation unit 62 calculates the foot length based on the top view image. The foot width calculation unit 64 calculates the foot width based on the top view image. The depression calculation unit 66 calculates the inner depression based on the top view image. The protrusion calculation unit 68 calculates the outer protrusion based on the top view image.

[0058] The pronation estimation unit 72 estimates the degree of pronation of the foot based on a comparison between a feature value including at least one of the foot length, the foot width, the amount of medial concavity, and the amount of lateral protrusion and a reference value. The pronation estimation unit 72 estimates which of the multiple pronation types the foot of the measurement subject 10 corresponds to using a foot pronation estimation algorithm predetermined in the design stage by a decision tree analysis described later. The pronation estimation unit 72 estimates the pronation type for each of the left and right feet of the measurement subject 10.

[0059] The storage unit 80 includes a reference storage unit 82, a pronation type storage unit 84, and a shoe type storage unit 86. The reference storage unit 82 stores reference values ​​related to feature quantities of a plurality of foot parts determined by decision tree analysis described later.

[0060] The pronation type storage unit 84 stores a plurality of pronation types, the plurality of pronation types including at least a first type in which the pronation degree is higher than a predetermined reference and a second type in which the pronation degree is less than the predetermined reference, as types classified according to the pronation degree of the foot. The predetermined reference mentioned here is, for example, "natural internal rotation" in which the heel moderately turns inward when landing.

[0061] As shoe types classified according to the form of pronation of the foot, the shoe type storage unit 86 is divided into a first shoe type for coping with a pronation degree higher than a specified benchmark, and a second shoe type for coping with a pronation degree below a specified benchmark, to store information on shoe product styles.

[0062] The evaluation unit 70 includes a pronation estimation unit 72, a type determination unit 74, and a shoe selection unit 76. The pronation estimation unit 72 estimates the pronation type of the foot of the measurement subject 10 using a foot pronation estimation algorithm described later. The type determination unit 74 determines a shoe type recommended to the measurement subject 10 from a plurality of shoe types based on the pronation degree estimated by the pronation estimation unit 72. The shoe selection unit 76 selects a shoe product style corresponding to the shoe type determined by the type determination unit 74 from the shoe type storage unit 86.

[0063] The output unit 90 includes a result output unit 92 and a recommendation output unit 94. The result output unit 92 transmits information on the pronation type of the foot of the measurement subject 10 determined by the pronation estimation unit 72 or information on the type of shoes suitable for the foot of the measurement subject 10 determined by the type determination unit 74 to the information terminal 16 via the communication unit 52. The recommendation output unit 94 transmits information on the style of the shoe product recommended to the measurement subject 10 selected by the shoe selection unit 76 to the information terminal 16 via the communication unit 52.

[0064] The analysis and processing unit 99 determines a pronation estimation algorithm in the design stage of the pronation estimation device 50, and the pronation estimation algorithm is used to estimate whether the degree of pronation of the foot corresponds to the first type or the second type based on multiple feature quantities measured from the images of the feet of multiple subjects. The analysis and processing unit 99 determines the pronation estimation algorithm by decision tree analysis, especially regression tree. The decision tree analysis is performed in advance at a stage before the time point of estimating the pronation of the measurement subject 10, for example, in the design stage of the pronation estimation system 100. The judgment process and judgment threshold of each feature quantity determined by the decision tree analysis are adopted as the pronation estimation algorithm in the pronation estimation process in the pronation estimation unit 72. The judgment threshold of each feature quantity determined by the decision tree analysis is stored in the reference storage unit 82 as a judgment reference value of each feature quantity in the pronation estimation process. Candidates of the feature quantity of each foot part used for the estimation of pronation are foot length, foot width, medial concavity, and lateral protrusion, all of which are feature quantities that can be extracted from the top view image.

[0065] The decision tree analysis in the design stage is implemented according to the following process. First, the three-dimensional eversion angle of the heel is detected from the image obtained by taking a picture of the subject running barefoot using a motion capture system, and the actual pronation type is determined. In particular, when the eversion angle of the heel is greater than the specified threshold value X, it is determined to be excessive internal rotation or the first type. In addition, the following method can also be used, that is, the two-dimensional eversion angle of the heel is detected from the rear-shot image of the subject running barefoot on a treadmill. Moreover, even if it is not barefoot, as long as the shoes have no internal rotation inhibition function, it is also possible to run in the state of wearing shoes. This pronation type determination is implemented for n subjects, and they are classified according to the pronation type. On the other hand, the feature quantities of multiple foot parts are measured as the decision tree analysis object data of these subjects. As the feature quantities of the foot parts, for example, the foot length, foot width, medial concavity, and lateral protrusion are extracted from the overhead image of the foot.

[0066] A decision tree analysis is performed with the classification of whether or not n subjects meet the condition of overpronation as the target variable and two or more elements of foot length, foot width, medial concavity, and lateral protrusion as explanatory variables. While changing the combination of elements set as explanatory variables, such as a combination of two elements, a combination of three elements, and a combination of four elements among the four elements of foot length, foot width, medial concavity, and lateral protrusion, the decision tree analysis is performed on all combination patterns. In the decision tree analysis, the maximum depth of the node is set to 5, and the element of the explanatory variable and its threshold value that minimize the loss function, that is, the square error, of each branched node are regressively determined.

[0067] The classification result of the subject obtained by the decision tree analysis is compared with the classification result of the pronation type based on the three-dimensional eversion angle. The combination of feature quantities when the comparison result is 100% consistent with the classification results of all subjects can be considered as an effective explanatory variable that can estimate the pronation of the foot with high accuracy. The branching conditions and the order of branches in the decision tree analysis that can estimate the pronation of the foot with high accuracy are adopted in the pronation estimation algorithm used in the pronation estimation unit 72. The threshold value of each feature quantity determined as a branching condition is stored in the reference storage unit 82 as the reference value of each feature quantity in the pronation estimation algorithm. In the reference storage unit 82, not only the reference value of each feature quantity can be stored, but also the pronation estimation algorithm itself can be stored. For example, in the case where an algorithm or a reference value of each feature quantity that can more accurately determine the pronation type of the foot is determined by decision tree analysis with the help of subsequent additional experiments, the information stored in the reference storage unit 82 can also be rewritten. As a result, the pronation estimation algorithm or the reference value of each feature quantity can be updated, so that the estimation accuracy can be easily improved.

[0068] Figure 3 The feature extraction unit 60 sets a frame 21 surrounding the outline of the foot 20 in the overhead image 30. The frame 21 is a rectangular area set as the minimum rectangle circumscribing the entire outline of the foot 20. The points of contact between the outline of the foot 20 and the frame 21 are the toe side endpoint 24, the heel side endpoint 25, the inner side endpoint 28, and the outer side endpoint 29.

[0069] The sole part of the foot other than the toes is mainly divided into a forefoot 40, a midfoot 41, and a rearfoot 42. The area of ​​the midfoot 41, especially the inner foot side, is the arch 43. The arch shape 44 refers to the contour shape of the inner foot side (also called the inside of the foot) from the forefoot 40 to the midfoot 41, and mainly refers to the contour shape of the inner foot side of the arch 43. The arch shape 44 is generally a shape with a concave portion. The outer contour shape 45 near the little toe mainly refers to the contour shape of the outer foot side (also called the outside of the foot) from the forefoot 40 to the midfoot 41, and mainly refers to the contour shape of the outer foot side of the forefoot 40. The outer contour shape 45 near the little toe is generally a shape of a protruding portion.

[0070] The foot length calculation unit 62 calculates the distance between the upper side 22 passing through the toe side endpoint 24 and the lower side 23 passing through the heel side endpoint 25 in the frame 21, that is, the length of the left side 26 passing through the inner side endpoint 28 as the foot length 31. The inner side endpoint 28 is a position of a predetermined ratio in the left side 26, for example, 72% of the foot length 31 or a position of 28% from the upper side 22. The outer side endpoint 29 is a position of a predetermined ratio in the right side 27, for example, 63% of the foot length 31 or a position of 37% from the upper side 22. The foot width calculation unit 64 calculates the length of a straight line connecting the inner side endpoint 28 and the outer side endpoint 29 as the foot width 32.

[0071] Figure 4 The method for measuring the amount of medial concavity and the amount of lateral protrusion is shown. The amount of medial concavity 38 calculated by the concavity calculation unit 66 is the distance between the width of the imaginary line passing through the most protruding position in the arch shape 44 and the imaginary line passing through the most concave position. That is, the amount of medial concavity 38 is the distance between the left side 26 of the medial end point 28 and the medial reference line 36 passing through the most concave point 33, which is the most concave position in the arch shape 44. The medial reference line 36 is an imaginary line parallel to the left side 26. The amount of medial concavity 38 takes a negative value in the case of a concavity, and takes a positive value in the case of a protruding shape without a concavity. Here, the imaginary line passing through the most concave point 33 and orthogonal to the medial reference line 36 is set as the foot width direction line 35. In addition, the intersection of the contour line on the outer foot side of the midfoot 41 and the foot width direction line 35 is set as the outer reference point 34. An imaginary line passing through the outer reference point 34 and parallel to the right side 27 is referred to as an outer reference line 37 .

[0072] The outer protrusion amount 39 calculated by the protrusion amount calculation unit 68 is the distance between the width of the imaginary line passing through the most protruding position in the outer contour shape 45 near the little toe and the imaginary line passing through the predetermined reference point. That is, the outer protrusion amount 39 is the distance between the right side 27 passing through the outer end point 29 and the outer reference line 37 passing through the outer reference point 34. The outer protrusion amount 39 takes a positive value in the case of protrusion, and conversely takes a negative value in the case of no protrusion but concavity.

[0073] Figure 5Indicates the relationship between the pronation type and the shoe type. As the pronation type 120, the pronation type storage unit 84 stores "natural pronation" in which the heel moderately, that is, the valgus angle is within the threshold value X when landing, "excessive pronation" in which the heel greatly turns inward and the valgus angle exceeds the threshold value X, and "insufficient pronation" in which the heel turns inward slightly or turns outward slightly. The situation in which the heel turns inward is also called "supination", but here it is classified in the form included in the "insufficient pronation" with a small turning inward. Based on the "natural pronation", the situation in which the pronation degree of the foot is higher than the "natural pronation" is set as the first type 110, and the situation in which the pronation degree of the foot is less than the "natural pronation" is set as the second type 111.

[0074] As the shoe type 121, the shoe type storage unit 86 stores an "inward rotation response pattern" corresponding to the first type 110 and a "cushion pattern" corresponding to the second type 111 as shoe manufacturing patterns. The "inward rotation response pattern" is also called a "stability pattern". In order to suppress pronation, these patterns may also be provided with a hard member on the inner side of the heel, or have a mechanism for holding the arch of the foot, or have a function of suppressing the inward torsion of the midfoot.

[0075] Figure 6 The comparison results of the classification results of the subjects obtained by the decision tree analysis and the classification results of the pronation type based on the three-dimensional eversion angle are shown. In the first analysis, the decision tree analysis was performed based on all four factors, namely, the foot length, foot width, medial concavity, and lateral protrusion. In the first analysis, the pronation type was inferred with 100% accuracy, which was consistent with the classification results of the pronation type based on the three-dimensional eversion angle. In the second analysis, the decision tree analysis was performed based on the two factors of medial concavity and lateral protrusion. In the second analysis, the pronation type was also inferred with 100% accuracy, which was consistent with the classification results of the pronation type based on the three-dimensional eversion angle.

[0076] In the third analysis, the decision tree analysis was performed based on the three elements of foot length, medial concavity, and lateral protrusion. In the third analysis, the classification result of the pronation type based on the three-dimensional eversion angle was partially inconsistent, and the estimation accuracy was less than 100%. In the fourth analysis, the decision tree analysis was performed based on the three elements of foot width, medial concavity, and lateral protrusion. In the fourth analysis, it was also consistent with the classification result of the pronation type based on the three-dimensional eversion angle, and the pronation type was estimated with 100% accuracy.

[0077] In the fifth analysis, a decision tree analysis was performed based on the three elements of foot length, foot width, and medial concavity. In the fifth analysis, the pronation type was inferred with 100% accuracy, which was consistent with the classification result of the pronation type based on the three-dimensional eversion angle. In the sixth analysis, a decision tree analysis was performed based on the three elements of foot length, foot width, and lateral protrusion. In the sixth analysis, the pronation type was inferred with 100% accuracy, which was consistent with the classification result of the pronation type based on the three-dimensional eversion angle.

[0078] In the seventh analysis, the decision tree analysis was performed based on the two factors of foot length and medial concavity. In the seventh analysis, the classification result of the pronation type based on the three-dimensional eversion angle was consistent, and the pronation type was inferred with 100% accuracy. In the eighth analysis, the decision tree analysis was performed based on the two factors of foot length and lateral protrusion. In the eighth analysis, the classification result of the pronation type based on the three-dimensional eversion angle was partially inconsistent, and the inference accuracy was less than 100%.

[0079] In the ninth analysis, the decision tree analysis was performed based on the two factors of foot width and medial concavity. In the ninth analysis, the classification result of the pronation type based on the three-dimensional eversion angle was consistent, and the pronation type was inferred with 100% accuracy. In the tenth analysis, the decision tree analysis was performed based on the two factors of foot width and lateral protrusion. In the tenth analysis, the classification result of the pronation type based on the three-dimensional eversion angle was partially inconsistent, and the inference accuracy was less than 100%.

[0080] According to the above experiments, it is known that, in particular, when a decision tree analysis is performed with a combination of elements including foot width and medial concavity, the pronation type can be inferred with high accuracy. On the other hand, the decision tree analysis with an inference accuracy less than 100% is an analysis performed with a combination of elements including lateral protrusion. Therefore, as an ideal combination, in the present embodiment, a foot pronation inference algorithm based on the three elements of (1) foot length, foot width, medial concavity or (2) foot width, medial concavity, and lateral protrusion is adopted, which at least include foot width and medial concavity. However, in a modified example, a foot pronation inference algorithm based on a combination of the two elements of foot width and medial concavity is also effective, and a foot pronation inference algorithm based on a combination of the three elements of foot length, foot width, and lateral protrusion is also effective. Furthermore, a foot pronation inference algorithm based on all four elements of foot length, foot width, medial concavity, and lateral protrusion is also effective.

[0081] Figure 7 This is a tree diagram schematically showing the result of a decision tree analysis of foot pronation estimation based on three factors: foot length, foot width, and medial concavity. Figure 7In the example, the maximum depth of the node is set to 5 through decision tree analysis to classify the measured values ​​of the sample number n of 11. In N10 as a routing node, "leg length" is regressively determined as an explanatory variable and "predetermined value a" is determined as a threshold, and whether the leg length is less than a becomes a branch condition.

[0082] When the foot length is less than a, the leaf node N11 with a square error of 0 and a sample number n of 1 is classified as insufficient internal rotation according to its valgus angle. The branch destination when the foot length exceeds a is node N12. In node N12 with a sample number n of 10, "foot width" is regressively determined as an explanatory variable and "predetermined value b" is determined as a threshold, and whether the foot width is less than b becomes a branch condition.

[0083] The branch destination when the foot width is less than b is node N13. In node N13 with a sample number n of 2, "foot width" is regressively determined as the explanatory variable and "predetermined value c" is determined as the threshold, and whether the foot width is less than c becomes the branch condition. When the foot width is less than c, the leaf node N14 with a square error of 0 and a sample number n of 1 is classified as excessive pronation because its valgus angle exceeds the threshold value X. When the foot width exceeds c, the leaf node N15 with a square error of 0 and a sample number n of 1 is also classified as excessive pronation because its valgus angle exceeds the threshold value X.

[0084] In node N12, when the foot width exceeds b, the branch destination is node N16. In node N16 with a sample number n of 8, "foot width" is regressively determined as the explanatory variable and "prescribed value d" is determined as the threshold, and whether the foot width is less than d becomes the branch condition. The branch destination when the foot width is less than d is node N17. In node N17 with a sample number n of 3, "medial concavity" is regressively determined as the explanatory variable and "prescribed value e" is determined as the threshold, and whether the medial concavity is less than e becomes the branch condition. When the medial concavity is less than e, leaf node N18 with a square error of 0 and a sample number n of 2 is classified as natural inward rotation based on the average of its valgus angle. When the medial concavity exceeds e, leaf node N19 with a square error of 0 and a sample number n of 1 is also classified as natural inward rotation based on its valgus angle.

[0085] In node N16, when the foot width exceeds d, the branch destination is node N20. In node N20 with a sample number n of 5, the "medial concavity" is regressively determined as the explanatory variable and the "predetermined value f" is determined as the threshold value, and whether the medial concavity is less than f becomes the branch condition. When the medial concavity is less than f, the leaf node N21 with a square error of about 0.9 and a sample number n of 4 is classified as excessive pronation because its average valgus angle exceeds the threshold value X. When the medial concavity exceeds f, the leaf node N22 with a square error of 0 and a sample number n of 1 is classified as natural pronation according to its valgus angle.

[0086] By using decision tree analysis for classification, even when samples with greatly deviated valgus angles are included, such as in leaf node N11, feature quantities and thresholds suitable for classification can be regressively determined, thereby enabling appropriate classification without affecting the accuracy of overall pronation estimation.

[0087] Follow through Figure 7 The foot pronation estimation algorithm used in the pronation estimation unit 72 is determined by the process determined by the decision tree analysis. Here, in the estimation process performed by the pronation estimation unit 72, not all of excessive internal rotation, natural internal rotation, and insufficient internal rotation are determined, but only the foot pronation estimation algorithm used in the pronation estimation unit 72 is determined. Figure 5 As shown in FIG. 1 , it is estimated whether it is the first type or the second type. Therefore, it is not necessary to distinguish between natural internal rotation and insufficient internal rotation. In addition, in the case of the same classification as node N13 or node N17, the estimation process is not required. As a result, the foot pronation estimation algorithm used in the pronation estimation unit 72 is changed from Figure 7 The process based on decision tree analysis is simplified as shown in the figure below. The predetermined value a, predetermined value b, predetermined value d, and predetermined value f, which are threshold values ​​of branch conditions, are determined by decision tree analysis and become part of the pronation estimation algorithm, and are each stored in the reference storage unit 82 as a reference value for each feature value.

[0088] Figure 8 The flowchart shows the procedure of the pronation estimation process using the pronation estimation algorithm based on the three elements of foot length, foot width, and medial concavity. Figure 8 In the example, the equivalent of Figure 7The branch judgment of nodes N10, N12, N16, and N20 in the decision tree analysis is used as a foot pronation estimation algorithm. The image acquisition unit 54 acquires a top view image of the foot of the measurement subject 10 (S10), and the feature extraction unit 60 extracts the foot length, foot width, and medial concavity as feature quantities of the foot part (S11). If the foot length is less than the specified value a (Y of S12), the pronation estimation unit 72 estimates it as the second type (S17). In the case where the foot length exceeds the specified value a (N of S12), if the foot width is less than the specified value b (Y of S13), it is estimated to be the first type (S16). In S13, if the foot width exceeds the specified value b (N of S13) and if the foot width is less than the specified value d (Y of S14), it is estimated to be the second type (S17). In S14, when the foot width exceeds the specified value d (N of S14) and the amount of medial concavity is less than the specified value f (Y of S15), it is estimated as the first type (S16). When the amount of medial concavity exceeds the specified value f (N of S15), it is estimated as the second type (S17). Based on the result of the estimation (S16, S17) of which of the first type and the second type is met, the output unit 90 outputs the pronation type (S18), and the pronation estimation process is completed. In the above process, the processing of S12, S13, S14, and S15 is equivalent to Figure 7 The branch decision processing of the node N10, the node N12, the node N16, and the node N20.

[0089] Fig. 9 This is a tree diagram schematically showing the result of a decision tree analysis of foot pronation estimation based on three factors: foot width, medial concavity, and lateral protrusion. Fig. 9 In the example, the maximum depth of the node is set to 5 through decision tree analysis to classify the measured values ​​with the number of samples n being 11. In N30 as a routing node, "foot width" is regressively determined as an explanatory variable and "predetermined value b" is determined as a threshold, and whether the foot width is less than b becomes a branch condition.

[0090] The branch destination when the foot width is less than b is node N31. In node N31 with a sample number n of 2, "foot width" is regressively determined as the explanatory variable and "predetermined value c" is determined as the threshold, and whether the foot width is less than c becomes the branch condition. When the foot width is less than c, the leaf node N32 with a square error of 0 and a sample number n of 1 is classified as excessive pronation because its valgus angle exceeds the threshold value X. When the foot width exceeds c, the leaf node N33 with a square error of 0 and a sample number n of 1 is also classified as excessive pronation because its valgus angle exceeds the threshold value X.

[0091] In routing node N30, when the foot width exceeds b, the branch destination is node N34. In node N34 with a sample number n of 9, the "inner concavity" is regressively determined as the explanatory variable and the "predetermined value g" is determined as the threshold value, and whether the inner concavity is less than g becomes the branch condition.

[0092] The branch destination when the medial concavity is less than g is node N35. In node N35 with a sample number n of 2, the "medial concavity" is regressively determined as the explanatory variable and the "predetermined value h" is determined as the threshold value, and whether the medial concavity is less than h becomes the branch condition. When the medial concavity is less than h, the leaf node N36 with a square error of 0 and a sample number n of 1 is classified as natural internal rotation according to its valgus angle. When the medial concavity exceeds h, the leaf node N37 with a square error of 0 and a sample number n of 1 is classified as insufficient internal rotation according to its valgus angle.

[0093] In node N34, when the amount of medial concavity exceeds g, the branch destination is node N38. In node N38 where the number of samples n is 7, the "foot width" is regressively determined as the explanatory variable and the "prescribed value i" is determined as the threshold, and whether the foot width is less than i becomes the branch condition. The branch destination when the foot width is less than i is node N39. In node N39 where the number of samples n is 3, the "outer protrusion amount" is regressively determined as the explanatory variable and the "prescribed value j" is determined as the threshold, and whether the outer protrusion amount is less than j becomes the branch condition. When the outer protrusion amount is less than j, the leaf node N40 with a square error of 0 and a sample number n of 1 is classified as natural internal rotation according to its valgus angle. When the outer protrusion amount exceeds j, the leaf node N41 with a square error of 0 and a sample number n of 1 is also classified as natural internal rotation according to its valgus angle.

[0094] In node N38, the branch destination when the foot width exceeds i is node N42. In node N42 with a sample number n of 5, "foot width" is regressively determined as the explanatory variable and "predetermined value k" is determined as the threshold, and whether the foot width is less than k becomes the branch condition. When the foot width is less than k, the leaf node N43 with a square error of about 0.9 and a sample number n of 4 is classified as excessive pronation because its average valgus angle exceeds the threshold value X. When the foot width exceeds k, the leaf node N44 with a square error of 0 and a sample number n of 1 is classified as natural pronation according to its valgus angle.

[0095] Follow through Fig. 9 The foot pronation estimation algorithm used in the pronation estimation unit 72 is determined by the process determined by the decision tree analysis. Here, in the estimation process performed by the pronation estimation unit 72, not all of excessive internal rotation, natural internal rotation, and insufficient internal rotation are determined, but only the foot pronation estimation algorithm used in the pronation estimation unit 72 is determined. Figure 5As shown in FIG. 1 , it is estimated whether it is the first type or the second type. Therefore, it is not necessary to distinguish between natural internal rotation and insufficient internal rotation. In addition, in the case of the same classification as node N31 or node N39, the estimation process is not required. As a result, the foot pronation estimation algorithm used in the pronation estimation unit 72 is changed from Fig. 9 The process based on decision tree analysis is simplified as shown in the figure below. Moreover, the branch processing based on the lateral protrusion amount, that is, the judgment result of node N39, is natural internal rotation, which is omitted in the foot pronation estimation algorithm in the figure below. Therefore, the lateral protrusion amount is not referenced in the foot pronation estimation process, which is essentially equivalent to the foot pronation estimation process based on the two elements of foot width and medial depression. The specified value b, specified value g, specified value i, and specified value k that serve as the threshold value of the branch condition are specific values ​​determined by decision tree analysis and become part of the foot pronation estimation algorithm, and each is stored in the reference storage unit 82 as a reference value for each feature amount.

[0096] Fig.10 The flowchart shows the procedure of the pronation estimation process using the pronation estimation algorithm based on the three elements of the foot width, the medial concavity, and the lateral protrusion. Fig.10 In the example, the equivalent of Fig. 9 The branch judgment of nodes N30, N34, N38, and N42 in the decision tree analysis is used as a foot pronation estimation algorithm. The image acquisition unit 54 acquires a top view image of the foot of the measurement subject 10 (S20), and the feature extraction unit 60 extracts the foot width, the medial depression amount, and the lateral protrusion amount as the feature amount of the foot part (S21). In a modified example, the feature extraction unit 60 can also extract only the foot width and the medial depression amount as the specification of the feature amount of the foot part. If the foot length is less than the specified value b (Y of S22), the pronation estimation unit 72 estimates it as the first type (S26). When the foot length exceeds the specified value b (N of S22) and if the medial depression amount is less than the specified value g (Y of S23), it is estimated to be the second type (S27). In S23, when the amount of medial depression exceeds the prescribed value g (N of S23) and if the foot width is less than the prescribed value i (Y of S24), it is inferred to be the second type (S27). In S24, when the foot width exceeds the prescribed value i (N of S24) and if the foot width is less than the prescribed value k (Y of S25), it is inferred to be the first type (S26), and when the foot width exceeds the prescribed value k (N of S25), it is inferred to be the second type (S27). Based on the result of the inference (S26, S27) of which of the first type and the second type is met, the output unit 90 outputs the pronation type (S28), and the foot pronation inference processing ends. In the above process, the processing of S22, S23, S24, and S25 is equivalent to Fig. 9Branch decision processing of node N30, node N34, node N38, and node N42.

[0097] The above embodiment is described. Those skilled in the art will appreciate that the embodiment is for illustration only, and that various modifications may exist for the combination of the constituent elements or the processing steps described above, and such modifications are also within the scope of the present invention.

[0098] In a modified example, in addition to the top view image of the foot, a side view image may be used as an auxiliary, and a part of the feature values ​​of multiple foot parts may be extracted from the side view image. In another modified example, the pronation estimation unit 72 may also obtain information related to the previous disease of the measurement subject 10, and auxiliary reference to the information related to the previous disease in the estimation process of the pronation type.

[0099] In another variation, the ratio of the number of shoe product styles corresponding to the first type and the number of shoe product styles corresponding to the second type may be adjusted according to the distribution of the pronation types based on the estimated results of the pronation types of a large number of measurement subjects 10. In another variation, a training method for the tibialis posterior muscle or the trunk or buttocks may be suggested based on the estimated results of the pronation types of the measurement subjects 10. In another variation, not only the shoe product style corresponding to the estimated results of the pronation types may be recommended, but also the design of the insole corresponding to the estimated results of the pronation types may be suggested.

[0100] Furthermore, if the above-described embodiment is generalized, the following aspects are obtained.

[0101] (1) A foot pronation estimation device according to an embodiment of the present disclosure comprises:

[0102] an image acquisition unit that acquires at least a top-view image of a foot as an image representing a contour shape of the foot of the measurement subject when viewed from top;

[0103] a feature quantity extraction unit that extracts a feature quantity of a predetermined portion representing a top-view contour shape of a foot from the top-view image; and

[0104] The pronation estimation unit estimates the degree of pronation of the foot based on a comparison between the characteristic amount of the predetermined part and a reference value.

[0105] According to the foot pronation estimation device of (1) above, it is possible to provide a technique for easily determining the pronation type of a measurement subject.

[0106] (2) The pronation estimation device according to (1), wherein:

[0107] The feature extraction unit extracts feature values ​​of a plurality of locations from the overhead image, wherein the feature values ​​of the plurality of locations include at least the length of the foot width and the amount of depression in the shape of the arch of the foot.

[0108] The pronation estimation unit estimates the degree of pronation of the foot based on comparison between feature quantities of a plurality of locations including at least the length of the foot width and the amount of concavity of the arch shape and respective reference values.

[0109] (3) The pronation estimation device according to (2), wherein:

[0110] The characteristic quantities of the plurality of parts further include leg length.

[0111] (4) The pronation estimation device according to (2), wherein:

[0112] The feature quantities of the plurality of locations further include a protrusion amount in the outer contour shape near the little toe.

[0113] (5) The pronation estimation device according to any one of (1) to (4) may further include:

[0114] a storage unit storing a plurality of pronation types, the plurality of pronation types including at least a first type in which the pronation degree is higher than a predetermined reference and a second type in which the pronation degree is lower than the predetermined reference, as types classified according to the pronation degree of the foot;

[0115] The pronation estimation unit estimates which of the plurality of pronation types the body portion corresponds to based on a comparison between the feature amount of the predetermined portion and a reference value.

[0116] (6) The pronation estimation device according to any one of (1) to (5) may further include:

[0117] a storage unit storing a plurality of shoe types, the plurality of shoe types including at least a first shoe type for coping with a case where the degree of pronation is higher than a predetermined reference and a second shoe type for coping with a case where the degree of pronation is lower than a predetermined reference, as shoe types classified according to forms for coping with pronation of a foot; and

[0118] The type determination unit determines a shoe type recommended to the measurement subject from among the plurality of shoe types based on the pronation degree estimated by the pronation estimation unit.

[0119] (7) The pronation estimation device according to any one of (1) to (6) may further include:

[0120] a storage unit storing respective reference values ​​related to the feature quantities of the plurality of locations,

[0121] The pronation estimation unit estimates the pronation degree based on a comparison between the extracted feature amount and the stored reference value.

[0122] (8) The pronation estimation device according to any one of (1) to (7), wherein:

[0123] The image acquisition unit acquires, as the bird's-eye view image, an image showing a top-view contour shape of a foot of a measurement subject at a seat.

[0124] (9) The pronation estimation device according to any one of (1) to (8), wherein:

[0125] The image acquisition unit further acquires a side view image of the foot of the measurement subject,

[0126] The feature amount extraction unit extracts the feature amount of the predetermined portion from at least one of the top view image and the side view image.

[0127] (10) A method for estimating pronation of a foot according to an embodiment of the present disclosure includes the following steps:

[0128] acquiring at least a top-view image of a foot as an image representing a contour shape of the foot of the measurement subject when viewed from top;

[0129] extracting a feature quantity of a predetermined portion representing a top-view contour shape of a foot from the top-view image by image processing; and

[0130] The degree of pronation of the foot is estimated based on a comparison process between the characteristic amount of the predetermined part and a reference value.

[0131] (11) A computer-readable storage medium according to an embodiment of the present disclosure stores a pronation estimation program to enable a computer to perform the following functions:

[0132] acquiring at least a top-view image of a foot as an image representing a contour shape of the foot of the measurement subject when viewed from top;

[0133] Extracting a feature value of a predetermined portion representing a top-view contour shape of a foot from the top-view image; estimating a degree of pronation of the foot based on a comparison between the feature value of the predetermined portion and a reference value; and outputting information representing a result of the estimation.

Claims

1. A foot pronation estimation device, characterized in that: include: an image acquisition unit that acquires at least a top-view image of a foot as an image representing a contour shape of the foot of the measurement subject when viewed from top; a feature quantity extraction unit that extracts a feature quantity of a predetermined portion representing a top-view contour shape of a foot from the top-view image; as well as The pronation estimation unit estimates the degree of pronation of the foot based on a comparison between the characteristic amount of the predetermined part and a reference value.

2. The foot pronation estimation device according to claim 1, characterized in that: The feature extraction unit extracts feature values ​​of a plurality of locations from the overhead image, wherein the feature values ​​of the plurality of locations include at least the length of the foot width and the amount of depression in the shape of the arch of the foot. The pronation estimation unit estimates the degree of pronation of the foot based on comparison between feature quantities of a plurality of locations including at least the length of the foot width and the amount of concavity of the arch shape and respective reference values.

3. The foot pronation estimation device according to claim 2, characterized in that: The characteristic quantities of the plurality of parts further include leg length.

4. The foot pronation estimation device according to claim 2, characterized in that: The feature quantities of the plurality of locations further include a protrusion amount in the outer contour shape near the little toe.

5. The foot pronation estimation device according to claim 1, characterized in that: Also includes: a storage unit storing a plurality of pronation types, the plurality of pronation types including at least a first type in which the pronation degree is higher than a predetermined reference and a second type in which the pronation degree is lower than the predetermined reference, as types classified according to the pronation degree of the foot; The pronation estimation unit estimates which of the plurality of pronation types the body portion corresponds to based on a comparison between the feature amount of the predetermined portion and a reference value.

6. The pronation estimation device according to claim 1, characterized in that: Also includes: a storage unit storing a plurality of shoe types, the plurality of shoe types including at least a first shoe type for coping with a case where the degree of pronation is higher than a predetermined reference and a second shoe type for coping with a case where the degree of pronation is lower than a predetermined reference, as shoe types classified according to forms for coping with pronation of a foot; and The type determination unit determines a shoe type recommended to the measurement subject from among the plurality of shoe types based on the pronation degree estimated by the pronation estimation unit.

7. The foot pronation estimation device according to claim 1, characterized in that: Also includes: a storage unit storing a reference value related to the feature quantity of the predetermined part, The pronation estimation unit estimates the pronation degree based on a comparison between the extracted feature amount and the stored reference value.

8. The foot pronation estimation device according to claim 1, characterized in that: The image acquisition unit acquires, as the bird's-eye view image, an image showing a top-view contour shape of a foot of a measurement subject at a seat.

9. The pronation estimation device according to claim 1, characterized in that: The image acquisition unit further acquires a side view image of the foot of the measurement subject, The feature amount extraction unit extracts the feature amount of the predetermined portion from at least one of the top view image and the side view image.

10. A method for estimating foot pronation, characterized in that: The process includes: acquiring at least a top-view image of a foot as an image representing a contour shape of the foot of the measurement subject when viewed from top; extracting a feature quantity of a predetermined portion representing a top-view contour shape of a foot from the top-view image by image processing; as well as The degree of pronation of the foot is estimated based on a comparison process between the characteristic amount of the predetermined part and a reference value.

11. A computer-readable storage medium having a foot pronation estimation program stored thereon, characterized in that: Enable the computer to perform the following functions: acquiring at least a top-view image of a foot as an image representing a contour shape of the foot of the measurement subject when viewed from top; extracting a feature quantity of a predetermined portion representing a top-view contour shape of a foot from the top-view image; estimating the degree of pronation of the foot based on a comparison between the characteristic amount of the predetermined part and a reference value; as well as Information indicating the result of the estimation is output.

Images