JP2015107141A - Posture determination device, posture determination system, and program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a posture determination device capable of accurately evaluating a posture of a person in a sitting state, a posture determination system, and a program.SOLUTION: Terminal equipment 4 includes: a position acquisition part 30 that acquires positional information on a predetermined region of a person's body, estimated from a photographed image obtained by photographing the person in a sitting state; a pressure acquisition part 34 that acquires information on pressure applied to a seating surface on which the person sits; and an evaluation part 44 that evaluates the person's posture in the sitting state on the basis of the acquired positional information and the acquired pressure information.

Description

  The present invention relates to an attitude determination device, an attitude determination system, and a program.

  Patent Document 1 discloses a three-dimensional space based on a distance image acquisition unit that acquires a distance image in which a luminance value at each pixel indicates a distance from a camera to an object, and the distance image acquired by the distance image acquisition unit. A coordinate calculation unit that calculates the coordinates of the object, and a person region that is a region of an object that occupies a volume equal to or greater than a predetermined volume threshold in a three-dimensional space based on the coordinates of the object calculated by the coordinate calculation unit. A posture detection device comprising: a person region determination unit to be determined; and a posture detection unit that detects a posture of a subject person by applying a skeleton model obtained by modeling a human skeleton to the person region determined by the person region determination unit Is disclosed. Patent Document 2 discloses a method and apparatus for calculating a posture deviation value of a patient.

JP 2012-120647 A JP-T-2004-512919

  An object of the present invention is to provide a posture determination device, a posture determination system, and a program capable of accurately evaluating the posture of a seated person.

  According to a first aspect of the present invention, there is provided position acquisition means for acquiring position information of a predetermined part of a person's body, which is estimated from a photographed image of a person in a sitting state, and a seat on which the person is seated. Evaluation that evaluates the posture of the person in the sitting state based on pressure acquisition means for acquiring pressure information on the surface, position information acquired by the position acquisition means, and pressure information acquired by the pressure acquisition means A posture determination apparatus having means.

  The present invention according to claim 2 further includes inclination detection means for detecting the inclination of the body of the person based on the position information acquired by the position acquisition means, and the evaluation means is detected by the inclination detection means. The posture determination device according to claim 1, wherein the posture of the person in the sitting state is evaluated based on the tilt and the pressure information acquired by the pressure acquisition unit.

  According to a third aspect of the present invention, the position acquisition means indicates at least a two-dimensional coordinate position of a predetermined part of the person's body, which is estimated from a photographed image taken from the lateral direction of the person. The information is acquired, and the detection unit detects the degree of inclination of the person in the front-rear direction of the predetermined part based on the two-dimensional coordinate value of the predetermined part acquired by the position acquisition unit. 2. The posture determination device according to 2.

  According to a fourth aspect of the present invention, the position acquisition unit calculates a position on a three-dimensional coordinate of a predetermined part of the person's body, which is estimated from a photographed image photographed from the front or back direction of the person. And detecting the degree of inclination of the person in the front-rear direction of the predetermined part based on the three-dimensional coordinate value of the predetermined part acquired by the position acquisition unit. Item 3. The attitude determination device according to Item 2.

  The present invention according to claim 5 is characterized in that the evaluation means evaluates the degree of appropriateness of the posture lower than the inclination of the person's body in the forward direction with respect to the inclination of the person's body in the forward direction. Is an attitude determination device.

  The present invention according to claim 6 is characterized in that the evaluation means evaluates according to a deviation from a reference position with respect to the position of the center of gravity calculated from the pressure information acquired by the pressure acquisition means. It is an attitude | position determination apparatus of description.

  According to a seventh aspect of the present invention, in the posture determination apparatus according to the sixth aspect, the evaluation means evaluates the degree of appropriateness of the posture with respect to the shift of the position of the center of gravity to the front of the seat surface lower than the shift to the rear of the seat surface. It is.

  The present invention according to claim 8 is the posture determination according to any one of claims 1 to 7, wherein the evaluation means evaluates according to a deviation from a reference value of the pressure distribution on the seating surface acquired by the pressure acquisition means. Device.

  The present invention according to claim 9 is the posture determination apparatus according to any one of claims 1 to 8, wherein the evaluation means evaluates according to the symmetry of the pressure distribution on the seating surface acquired by the pressure acquisition means.

  The present invention according to claim 10 is the posture determination apparatus according to any one of claims 1 to 9, further comprising display means for displaying an evaluation result by the evaluation means.

  The present invention according to claim 11 includes a photographing device that photographs a person in a seated state, a measuring device that measures pressure applied to a seating surface on which the person is seated, and a posture determination device, wherein the posture determination device includes: Position acquisition means for acquiring position information of a predetermined part of the body of the person estimated from a captured image captured by the imaging apparatus, and the person based on the position information acquired by the position acquisition means An inclination detecting means for detecting the inclination of the body, a pressure acquiring means for acquiring pressure information measured by the measuring device, an inclination detected by the inclination detecting means, and a pressure information acquired by the pressure acquiring means And an evaluation unit that evaluates the posture of the person in the seated state.

  According to a twelfth aspect of the present invention, there is provided a position acquisition step of acquiring position information of a predetermined part of a body of the person estimated from a photographed image obtained by photographing a person in a sitting state, and a seat on which the person is seated A program for causing a computer to execute a pressure acquisition step for acquiring pressure information applied to a surface, and an evaluation step for evaluating the posture of the person in the sitting state based on the acquired position information and the acquired pressure information. .

  According to the first aspect of the present invention, the posture of a seated person can be accurately evaluated as compared with the case where the present configuration is not provided.

  According to the second aspect of the present invention, posture disturbance due to the inclination of the body of a person in a seated state can be reflected in the evaluation.

  According to the third aspect of the present invention, even if the photographed image is an image photographed from the side of the person, the posture of the person in the seated state can be accurately evaluated as compared with the case where the present configuration is not provided.

  According to the fourth aspect of the present invention, even when the photographed image is an image photographed from the front or back of the person, the posture of the seated person can be accurately evaluated as compared with the case where the present configuration is not provided. it can.

  According to the fifth aspect of the present invention, it is possible to perform a relatively lower evaluation on the forward bending posture than on the warping posture.

  According to the sixth aspect of the present invention, the seating position on the seat surface can be reflected in the posture evaluation.

  According to the seventh aspect of the present invention, a relatively low evaluation can be performed when the seating position is shallower than the reference position than when the seating position is deeper than the reference position.

  According to the eighth aspect of the present invention, it is possible to accurately evaluate the posture of a seated person as compared to the case where the present configuration is not provided.

  According to the present invention of claim 9, the posture of a person in a seated state can be accurately evaluated as compared with the case where the present configuration is not provided.

  According to the tenth aspect of the present invention, the evaluation result can be visually confirmed.

  According to the present invention of claim 11, it is possible to accurately evaluate the posture of a person in a seated state as compared with the case where the present configuration is not provided.

  According to the twelfth aspect of the present invention, the posture of a seated person can be accurately evaluated as compared with the case where the present configuration is not provided.

It is a schematic diagram which shows an example of a structure of the attitude | position determination system 2 which concerns on embodiment of this invention. It is a figure which shows an example of the hardware constitutions of the terminal device 4 which concerns on embodiment of this invention. It is a block diagram which shows the function structure of the terminal device 4 implement | achieved by running a program. It is a schematic diagram explaining the location where the inclination detection part 32 detects an inclination. It is a figure explaining an example of the inclination detection at the time of image | photographing from the front or back of the person 12, (a) is a schematic diagram which illustrated the estimated position of each site | part on the person 12, (b) is a head It is a schematic diagram shown about the distance of the part A and the shoulder center P, (c) is a schematic diagram explaining calculation of inclination, (d) is a calculation formula of inclination. It is a figure explaining an example of the inclination detection at the time of image | photographing from the side of the person 12, (a) is a schematic diagram explaining calculation of inclination, (b) is a calculation formula of inclination. It is a table | surface which shows an example about an angle conversion value and a deduction coefficient. It is a schematic diagram explaining the determination method of the deduction number Gr_Level in the evaluation part 44. FIG. It is a schematic diagram which shows about the seating posture used as a reference | standard, the seating posture of the back of the posture judgment object, and the shallow seating posture where the pelvis has fallen, (a) shows the seating posture used as a reference, (b) shows the seating of the back The posture is shown, and (c) shows a shallow sitting posture with the pelvis collapsed. It is the schematic diagram shown about the pressure distribution example in the correct attitude | position which is an attitude | position used as the reference | standard of attitude | position determination, and the calculation of the statistic based on this pressure distribution example. It is the schematic diagram shown about the calculation of the statistic based on the example of the pressure distribution in the sitting posture of a dorsum and the pressure distribution example. It is a table | surface which shows an example of the classification | category specified by the symmetry determination part 40. FIG. It is a schematic diagram which shows an example of the pressure distribution at the time of dividing | segmenting an area | region into the 4th quadrant from the 1st quadrant as a seat surface division at the time of determining symmetry. It is a schematic diagram shown about the histogram of a 1st quadrant and a 2nd quadrant, (a) shows the histogram of the x direction and y direction in a 1st quadrant, (b) shows the histogram of the x direction and the y direction in a 2nd quadrant. Show. 6 is a schematic diagram illustrating an example of a screen displayed on the UI unit 22 by the display unit 46. FIG. 5 is a flowchart illustrating the operation of the posture determination system 2. It is a flowchart which shows an example of acquisition operation (S10) of standard data in posture judging system 2. It is a flowchart which shows an example of the attitude | position determination operation | movement (S20) in the attitude | position determination system 2. FIG.

Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.
FIG. 1 is a schematic diagram illustrating an example of a configuration of a posture determination system 2 according to an embodiment of the present invention. The posture determination system 2 according to the present invention includes a terminal device 4, a photographing device 6, and a pressure measuring device 8. The photographing device 6 and the pressure measuring device 8 can communicate with each other by wire or wirelessly. It is connected to the.

  The terminal device 4 is a device having a function as a computer, and is, for example, a tablet PC (personal computer), a PDA (Personal Digital Assistant), a smartphone, or the like.

  The photographing device 6 is a camera that photographs a person in a seated state. In the present embodiment, the photographing device 6 is provided integrally with the terminal device 4 and photographs the person 12 seated on the chair 10 from the front of the person 12. Note that the photographing device 6 may be provided separately from the terminal device 4. The person 12 seated on the chair 10 may be photographed from the back of the person 12. In the present embodiment, the photographing device 6 is a camera equipped with a distance image sensor and is configured to photograph a distance image. Therefore, three-dimensional data is obtained from the photographing device 6 as photographing data. In the present embodiment, the image capturing device 6 only needs to be able to capture a distance image, and the distance image capturing method does not matter. For example, the distance image capturing method may be a TOF (Time Of Flight) method or a pattern irradiation method. As will be described later, when photographing from the side of the person 12, the photographing device 6 is configured as a camera for photographing a two-dimensional image and does not have to photograph a distance image.

  In the present embodiment, the photographing device 6 is provided to photograph the upper body of the seated person 12 including, for example, the buttocks, but the photographing range is not limited thereto. For example, it may be provided to photograph the body above the waist, or may be provided to photograph a range including the lower part than the buttocks.

  The pressure measuring device 8 is a measuring device that measures a pressure applied to a seating surface on which a person is seated. The pressure measuring device 8 is configured in a sheet shape, for example, and is laid on the seat surface, and detects the pressure applied to the seat surface by the seating. In the present embodiment, the pressure measuring device 8 is provided on the seat surface of the chair 10 and detects the pressure at each point (for example, 16 × 16 256 points) in the seat surface region.

FIG. 2 is a diagram illustrating an example of a hardware configuration of the terminal device 4.
As shown in FIG. 2, the terminal device 4 transmits and receives data between the CPU 14, the memory 16, the storage unit 18 such as a hard disk drive (HDD), the imaging device 6, and the pressure measurement device 8 wirelessly or by wire. The communication interface (IF) 20 includes a user interface (UI) unit 22 such as a touch panel that receives input made by an operator and displays information to be displayed. These components are connected to each other via a control bus 24. It is connected. Note that the communication IF 20 may transmit / receive data to / from other devices via a network (not shown). Further, the UI unit 22 is not limited to a configuration using a touch panel, and may be configured from a pointing device or a keyboard and a display.

  The CPU 14 executes processing based on a program stored in the memory 16 or the storage unit 18 and controls the operation of the terminal device 4. In the present embodiment, the CPU 14 has been described as reading and executing a program stored in the memory 16 or the storage unit 18. However, the program is stored in a storage medium such as a CD-ROM and provided to the CPU 14. It is also possible to provide it to the CPU 14 via the communication IF 20.

  FIG. 3 is a block diagram illustrating a functional configuration of the terminal device 4 realized by executing the program.

  As illustrated in FIG. 3, the terminal device 4 includes a position acquisition unit 30, an inclination detection unit 32, a pressure acquisition unit 34, a centroid position detection unit 36, a statistic calculation unit 38, and a symmetry determination unit 40. A reference data storage unit 42, an evaluation unit 44, and a display unit 46.

  The position acquisition unit 30 acquires position information of a predetermined part of the body of the person 12 in the sitting state. Specifically, the position acquisition unit 30 acquires position information estimated from shooting data obtained by shooting by the shooting device 6.

  The position acquisition unit 30 estimates the position of the body part of the seated person 12 by modeling a human skeleton in advance by machine learning or the like and applying the model to the captured image from the imaging device 6. And obtain location information. The position acquisition unit 30 only needs to acquire position information on a predetermined part of the body of the person 12 in the sitting state, and the method for estimating the position information is not limited. Further, the position acquisition unit 30 may acquire position information estimated by another device.

  In the present embodiment, the position acquisition unit 30 acquires position information of parts such as the head, shoulders, waist, and buttocks as position information of predetermined parts. The position acquisition unit 30 acquires, for example, a position estimated as a center part of the head as position information of the head, acquires a position estimated as, for example, a joint part of the shoulder as position information of the shoulder, and positions of the waist For example, the position estimated as the central part of the lumbar vertebra is acquired as information, and the position estimated as the central part of the sacrum is acquired as the buttocks.

  The inclination detection unit 32 detects the inclination of the body of the person 12 based on the position information acquired by the position acquisition unit 30. The inclination detection unit 32 outputs information about the detected inclination to the reference data storage unit 18 when obtaining data serving as a reference when determining the posture of the person 12. On the other hand, when the posture of the person 12 is determined, the tilt detection unit 32 outputs information about the detected tilt to the evaluation unit 44 described later. Details of tilt detection by the tilt detection unit 32 will be described later using a specific example.

  The pressure acquisition unit 34 has a function of acquiring pressure information applied to the seating surface on which the person 12 is seated, and acquires measurement data of the pressure measured by the pressure measurement device 8 in the present embodiment. In the present embodiment, a pressure distribution made up of pressure values at a total of 256 points of 16 × 16 is arranged on the seating surface area at equal intervals.

  The center-of-gravity position detection unit 36 detects the center-of-gravity position of the pressure applied to the seat surface by the seated person 12 from the measurement data acquired by the pressure acquisition unit 34. When the posture of the person 12 is determined, the center-of-gravity position detection unit 36 outputs information about the detected center-of-gravity position to the evaluation unit 44 described later.

The statistic calculation unit 38 calculates a predetermined statistic from the measurement data acquired by the pressure acquisition unit 34. In the present embodiment, the average value μ, variance σ ² , standard deviation σ, peak value (maximum value) Vp, and kurtosis Vk are calculated based on the pressure distribution obtained from the pressure acquisition unit 34. Note that the statistic calculator 38 may calculate not only these statistics but also other statistics. The statistic calculation unit 38 outputs information about the calculated statistic to the reference data storage unit 18 when obtaining data serving as a reference when determining the posture of the person 12. On the other hand, when determining the posture of the person 12, the statistic calculation unit 38 outputs information about the calculated statistic to the evaluation unit 44 described later.

  The symmetry determination unit 40 determines the symmetry of the pressure distribution acquired by the pressure acquisition unit 34. In this embodiment, the symmetry determination unit 40 estimates the posture of the person 12 from the symmetry of the pressure distribution between the divided seating surface regions. When determining the posture of the person 12, the symmetry determination unit 40 outputs information about the determined symmetry of the pressure distribution to the evaluation unit 44 described later. Details of the symmetry determination by the symmetry determination unit 40 will be described later using a specific example.

  The reference data storage unit 18 stores reference data serving as a reference when determining the posture of the person 12. Specifically, the person 12 takes a reference posture state, and the data obtained from the inclination detection unit 32 and the statistic calculation unit 38 at that time is stored as reference data.

  The evaluation unit 44 evaluates the posture of the person 12 in the sitting state based on the position information acquired by the position acquisition unit 30 and the pressure information acquired by the pressure acquisition unit 34. The evaluation unit 44 calculates a difference between the reference data stored in the reference data storage unit 18 and the data representing the position information and the pressure information acquired by the position acquisition unit 30 and the pressure acquisition unit 34 when the posture of the person 12 is determined. It is used to evaluate the posture of the person 12. A specific example of evaluation by the evaluation unit 44 will be described later.

  The display unit 46 displays the evaluation result by the evaluation unit 44 on the UI unit 22. A display example by the display unit 46 will be described later.

  Next, details of tilt detection by the tilt detection unit 32 will be described with reference to FIGS. FIG. 4 is a schematic diagram for explaining a portion where the inclination detecting unit 32 detects the inclination. In the diagram shown in FIG. 4, a situation from the side of the seated person 12 is shown, and the estimated positions of the respective parts are illustrated on the person 12. Specifically, in FIG. 4, the head A, shoulder B, waist C, and buttock D of the person 12 are shown as each part.

In the present embodiment, the inclination detection unit 32 includes a front-rear direction inclination θ ₁ of the person 12 in the head A, a front-rear direction inclination θ ₂ of the person 12 in the shoulder B (a center part between both shoulders), and a waist part. The inclination θ _{3 in} the front-rear direction of the person 12 in C is detected. In the example shown in FIG. 4, the inclination θ ₁ is a person formed by a vector connecting the shoulder B (the center between both shoulders) and the head A and a perpendicular passing through the shoulder B (the center between both shoulders). The inclination θ ₂ is an angle in the front-rear direction of the person 12 formed by a vector connecting the waist C and the shoulder B (the center between both shoulders) and a perpendicular passing through the waist C. The inclination θ ₃ is an angle in the front-rear direction of the person 12 formed by a vector connecting the waist C and the buttocks D and a perpendicular passing through the buttocks D.

  In the present embodiment, the configuration in which the inclination detection unit 32 obtains the inclination of each part in the front-rear direction of the person 12 will be described, but the inclination detection part 32 may obtain the inclination of each part in other directions.

  Next, an example of a tilt detection method by the tilt detection unit 32 will be described. First, an example of inclination detection when the person 12 is photographed from the front of the person 12 by the photographing device 6 will be described with reference to FIG.

  FIG. 5 (a) shows a situation where the seated person 12 is photographed from the front, and is a schematic diagram illustrating the estimated position of each part on the person 12. Here, a distance image of the person 12 is photographed by the photographing device 6, and the head A, left shoulder B1, right shoulder B2, waist C, buttocks D, left elbow E1, right elbow E2, left hand of the person 12 is captured. It is assumed that the positions of F1 and right hand part F2 are acquired by the position acquisition unit 30. Here, the position information of each part includes the x coordinate in the left-right direction (horizontal direction) when viewed from the front of the person 12, the y coordinate in the vertical direction (vertical direction) when viewed from the front of the person 12, It is assumed that the distance from the imaging device is represented by coordinates consisting of (x, y, z) with z coordinates. In the present embodiment, since the z coordinate is a distance to the last, the direction is not uniquely specified.

The inclination detecting unit 32 calculates the inclination θ ₁ as follows.
The inclination detection unit 32 first calculates the coordinates (x _p , y _p , z _p ) of the shoulder center P that is the center of the left shoulder B1 and the right shoulder B2. If the position acquisition unit 30 has already acquired the coordinates of the shoulder center P, the inclination detection unit 32 does not need to calculate the coordinates of the shoulder center P.

  Next, the inclination detector 32 obtains the distance in the front-rear direction and the distance in the vertical direction of the person 12 between the head A and the shoulder center P. The distance in the front-rear direction is approximated by a difference in distance from the imaging device 6, that is, a difference in z coordinates, as shown in FIG. Further, the distance in the vertical direction is calculated from the difference in y coordinates.

Next, as illustrated in FIG. 5C, the inclination detection unit 32 calculates an angle θ of a right triangle having a head A and a shoulder center P on the hypotenuse. Specifically, as represented by the equation shown in FIG. 5D, the angle θ is calculated using an inverse tangent function, and the inclination θ ₁ is obtained by subtracting θ from 90 °. The inclination detection unit 32 similarly obtains the inclination θ ₂ and the inclination θ ₃ .

  Next, an inclination detection method when the person 12 is photographed from the side of the person 12 by the photographing apparatus 6 will be described with reference to FIGS. 4 and 6.

  It is assumed that a distance image of the person 12 is taken by the photographing device 6 and the positions of the head A, shoulder B, waist C, and buttock D of the person 12 are acquired by the position acquisition unit 30 (see FIG. 4). Here, in the position information of each part, the horizontal direction (horizontal direction) when viewed from the side surface of the person 12 is the x coordinate, and the vertical direction (vertical direction) when viewed from the side surface of the person 12 is the y coordinate ( Let it be represented by coordinates consisting of x, y). Note that the photographing device 6 does not necessarily have to shoot a distance image. The predetermined portions of the person 12 are viewed from the left and right direction (horizontal direction) when viewed from the side of the person 12 and from the side of the person 12. In this case, when the coordinate position in the vertical direction (vertical direction) can be estimated, the imaging device 6 may be a device that captures a two-dimensional image instead of a distance image.

The inclination detecting unit 32 calculates the inclination θ ₁ as follows.
First, the inclination detection unit 32 calculates a vector s from the shoulder B to the head A and a vector to a point B ′ where the y coordinate on the perpendicular passing from the shoulder B to the shoulder B is equal to the y coordinate of the head A. With respect to an angle θ ₁ (see FIG. 6A) formed with t, a cosine is obtained by an inner product as shown in FIG. 6B. Here, the coordinates of the head _{A (x a, y a)} , the coordinates of the shoulder portion B (x _b, y _b) When the vector s is _{(x a -x b, y a} -y b) and represented, the vector t is expressed as _{(0, y a -y b)} , the cosine is calculated as shown in Figure 6 (b). Thereafter, the inclination detection unit 32 obtains the inclination θ ₁ by the inverse function of cosine. The inclination detection unit 32 similarly obtains the inclination θ ₂ and the inclination θ ₃ .

Next, a specific evaluation example by the evaluation unit 44 will be described.
The evaluation unit 44 calculates an evaluation value E evaluated for the appropriateness of the posture of the seated person 12 by the following equation.

E = 100− (E ₁ + E ₂ + E ₃ + E ₄ ) (1)

Expression (1) is obtained by scoring the posture by subtracting the disturbance from the reference posture, with 100 points when the posture of the person 12 at the posture determination is the same as the reference posture. In Equation (1), E _{1 to} E ₄ indicate deduction points. E ₁ represents the number of points deducted by the degree of inclination of each part determined in advance by the person 12, and E ₂ represents a decrease according to the deviation from the reference position with respect to the position of the center of gravity calculated from the pressure information acquired by the pressure acquisition unit 34. E ₃ represents a deduction point corresponding to a deviation from the reference value for the pressure distribution on the seating surface acquired by the pressure acquisition unit 34, and E ₄ represents the pressure distribution on the seating surface acquired by the pressure acquisition unit 34. Denotes the number of points deducted according to symmetry. Hereinafter, calculation examples of E _{1 to} E ₄ will be described.

First, the deduction number E ₁ will be described. The evaluation unit 44 calculates the deduction point E ₁ by, for example, the following equation based on the inclination at the posture determination detected by the inclination detection unit 32 and the inclination at the reference posture stored in the reference data storage unit 18.

E ₁ = k ₁ × Tθ ₁ + k ₂ × Tθ ₂ + k ₃ × Tθ ₃ (2)

In Expression (2), for the above-described inclinations θ _{1 to} θ ₃ , the weighted sum is calculated for the difference between the inclinations θ _{1 to} θ _{3 in} the reference posture and the inclinations θ _{1 to} θ ₃ in the posture determination. . Here, k _{1 to} k ₃ are deduction coefficients indicating the degree of contribution of the inclinations θ _{1 to} θ _{3 to} the deduction. Further, Tθ ₁ ~Tθ _3, for the difference [Delta] [theta] ₁ through? ₃ of inclination theta ₁ through? ₃ when the posture to be the inclination theta ₁ through? ₃ and the reference at the time of posture determination, for calculating the deduction number E ₁ It is the value converted into.

_{k 1 ~k 3, Tθ 1 ~Tθ} 3 , for example, is predetermined according to the difference [Delta] [theta] ₁ through? ₃ as shown in Table shown in Fig. Taking FIG. 7 as an example, the deduction number E ₁ when Δθ ₁ = 10 °, Δθ ₂ = 20 °, and Δθ ₃ = 10 ° is 1.03 × 5 + 1.10 × 25 + 1.02 × 5 = 37.75. It becomes. Incidentally, [Delta] [theta] _n = (slope at the posture determining theta _n) - a (tilt when in relation to the standard posture theta _n).

  Further, as can be seen from Equation 1, Equation 2, and FIG. 7, the evaluation unit 44 evaluates the inclination of the person 12 in the front direction lower than the inclination of the person 12 in the rear direction when calculating the evaluation value E. Do (many points deducted). Thereby, even if the inclination degree of the body 12 of the seated person 12 in the front-rear direction is the same, the leaning posture is evaluated relatively lower than the warping posture. For example, when considering the burden on the body, it seems that the forward bending posture is more burdensome to the body in the forward bending posture and the warping posture, and this point is reflected in the evaluation according to the above evaluation. It will be.

Further, in the example shown in FIG. 7, the deduction coefficient is set such that k ₁ is relatively larger than k ₃ and k ₂ is relatively larger than k ₁ . Accordingly, the degree of deviation from the reference in the inclination between the waist C and the buttock D, the degree of deviation from the reference in the inclination between the head A and the shoulder B, and the degree of deviation from the reference in the inclination between the shoulder B and the waist C In this order, the impact on deductions increases. For example, when the inclinations θ _{1 to} θ ₃ in the reference posture are 0 °, the inclination between the waist C and the buttocks D, the inclination between the head A and the shoulder B, and between the shoulder B and the waist C Where the burden on the body seems to increase in the order of inclination, this point will be reflected in the evaluation. In the example shown in FIG. 7, the deduction points k _{1 to} k ₃ are different values according to the differences Δθ _{1 to} θ ₃ , respectively, but may be constants regardless of the differences Δθ _{1 to} θ ₃ .

The method for calculating the deduction point E ₁ has been described above. In calculating the differences Δθ _{1 to} θ ₃ , the inclinations θ _{1 to} θ ₃ in the reference posture are the postures that the person 12 actually uses as the reference. It may not be the detection result of the inclination detection unit 32 when it is performed. A predetermined reference value, for example, 0 ° may be used as the inclinations θ _{1 to} θ ₃ in the reference posture.

Next, the deduction point E ₂ will be described. Based on the position of the center of gravity at the time of posture determination detected by the center-of-gravity position detection unit 36, the evaluation unit 44 calculates the deduction point E ₂ by the following equation, for example.

E ₂ = k ₄ × Gr_Level (3)

In Expression (3), Gr_Level represents the number of points deducted according to the degree of deviation from the reference position of the position of the center of gravity at the time of posture determination. K ₄ is a deduction coefficient indicating the degree of deduction contribution associated with the position of the center of gravity with respect to the overall deduction, and is determined experimentally, for example, in order to adjust the evaluation value E.

  FIG. 8 is a schematic diagram illustrating a method for determining the deduction point Gr_Level in the evaluation unit 44. The evaluation unit 44 determines a value for the deduction point Gr_Level depending on which region of the seating surface divided in the front-rear direction of the seating surface (the front-rear direction of the person) is located.

  In the example shown in FIG. 8, the seating surface is divided into five regions in the front-rear direction of the seating surface, and the deduction points Gr_Level are determined as 1, 0, 2, 3, 4 in order from the rear of the seating surface. It has been. In the example shown in FIG. 8, the position coordinates (8.2, 3.0) of the center of gravity are in the coordinate area of the deduction point Gr_Level = 3, and the value of the deduction point Gr_Level in this case is 3.

  In this way, in the example shown in FIG. 8, the deduction point Gr_Level is set so that the degree of appropriateness of the posture is evaluated with respect to the deviation of the position of the center of gravity toward the front of the seat surface rather than the deviation toward the rear of the seat surface. Where the seating position becomes shallower than the appropriate position (Gr_Level = 0) with respect to the seating surface of the chair, the burden on the body will increase, and this point will be reflected in the evaluation.

Next, the deduction point E ₃ will be described. Based on the statistic calculated by the statistic calculation unit 38, the evaluation unit 44 calculates the deduction point E ₃ by the following equation, for example.

E ₃ = K ((k ₅ × ΔVp) × (k ₆ × ΔVk) / (k ₇ × Δσ)) (4)

In the present embodiment, the difference ΔVp between the peak value Vp, kurtosis Vk and standard deviation σ based on the pressure distribution at the time of posture determination, and the peak value Vp, kurtosis Vk and standard deviation σ stored in the reference data storage unit 18, Using ΔVk and Δσ, the value of the deduction point E ₃ is determined based on the above equation (4). Here, k _{5 to} k ₇ are deduction coefficients indicating the degree of contribution of ΔVp, ΔVk and Δσ to the deduction, and K is a deduction coefficient indicating the degree of contribution of deduction accompanying the statistic with respect to the total deduction, k _{5 to} k ₇ and K are determined, for example, experimentally in order to adjust the evaluation value E.

In the present embodiment, the deduction points are determined based on the degree of pressure concentration at the position on the seat surface where the sciatic portion of the person 12 is placed in the sitting state. In addition, in the calculation formula shown by Formula (4), the peak value Vp, the kurtosis Vk, and the standard deviation σ () (in order to suppress erroneous determination in the case where the pressure concentration level is determined only by the peak value Vp or the kurtosis Vk. However, the deduction points may be determined based on either the peak value Vp or the kurtosis Vk and the standard deviation σ, or the deduction points may be determined without using the standard deviation σ. Further, the variance σ ² may be used as the degree of variation.

Here, with reference to FIG. 9 to FIG. 11, for the specific example of the deduction point E ₃ , a correct posture that is an example of a posture serving as a reference for posture determination (having fewer adverse effects on the body than other postures, physical health A case where a stooped posture is taken as compared to a desirable sitting posture will be described as an example.

  FIG. 9 is a schematic diagram showing a reference sitting posture, a sitting posture of the back of the posture determination target, and a shallow sitting posture in which the pelvis is collapsed, and FIG. 9A shows the reference sitting posture. 9 (b) shows the sitting posture of the stoop, and FIG. 9 (c) shows a shallow sitting posture with the pelvis collapsed. In addition, the curve drawn with the thick line in FIG. 9 is a model of the state of the spine in the sitting posture. In FIG. 9C, the body part within the imaging range by the imaging device 6 is represented by a solid line, and the body part outside the imaging range is represented by a broken line. Further, in the example shown in FIG. 9C, since the photographing range of the photographing device 6 is limited, the position information about the head A and the shoulder center P can be acquired by the position acquisition unit 30 from the shoulder. As for the lower part, it is assumed that position information cannot be acquired. Here, in the sitting posture shown in FIG. 9C, the posture of the body part within the range photographed by the photographing device 6 is the same as the standard sitting posture. In FIG. 9C, as an example, the distance (z coordinate) of the head A from the imaging device 6 and the distance (z coordinate) of the shoulder center P from the imaging device 6 are the reference sitting posture. Similarly, it is shown that they are approximately equal.

  When the person 12 is wearing clothes, the posture of the person 12 may not be correctly captured from the photographed image of the photographing device 6. As a result, for example, as shown in FIGS. 9A and 9B, the position acquisition unit based on the photographed image of the photographing apparatus 6 is used although the spine line shape is actually different and the posture is different. It is conceivable that the difference between the two cannot be determined by acquiring the position by 30. Further, for example, when the position information is acquired by the position acquisition unit 30, the position acquisition is performed even though the positions are actually different as shown in FIGS. 9A and 9C. It is conceivable that the difference between the two cannot be determined only by acquiring the position by the unit 30. On the other hand, it is expected to determine a posture that cannot be determined by the captured image of the image capturing device 6 by determining the posture based on statistics.

  FIG. 10 is a schematic diagram illustrating an example of pressure distribution in a correct posture, which is a posture serving as a reference for posture determination, and calculation of a statistic based on the pressure distribution example. As shown in FIG. 10, the pressure acquisition unit 34 acquires a pressure distribution composed of pressure values at a total of 256 points of 16 × 16. In FIG. 10, the numbers written in each square indicate the pressure value (mmHg) at each point. In addition, the pressure value of the square which does not describe the number is 0.

The statistic calculator 38 calculates the statistic based on the pressure distribution shown in FIG. 10 as follows, for example. First, in order to obtain the variance, an average value μ is calculated. Here, although the average value μ is calculated for a total of 198 points where the pressure value is not 0 among all 256 points, the average value μ may be calculated for all 256 points. Further, the statistic calculation unit 38 calculates the variance σ ² for the pressure distribution at 198 points using the calculated average value μ, and calculates the standard deviation σ from the calculated variance value. Further, the statistic calculation unit 38 also calculates the peak value Vp and the kurtosis Vk. Specifically, the statistic calculation unit 38 calculates the average value μ = 56.5 mmHg, variance σ ² = 1126, standard deviation σ = 33.6, peak value Vp = 198 mmHg, according to the calculation formula shown in FIG. The kurtosis Vk = 3.34 is calculated. The kurtosis takes a value of 0 in the case of a normalized distribution, takes a positive value if it is sharper than the normalized distribution, and takes a negative value if it is not sharper than the normalized distribution. Take.

FIG. 11 is a schematic diagram showing an example of pressure distribution in the sitting posture of the dorsum and the calculation of statistics based on this example of pressure distribution. Similar to the calculation described with reference to FIG. 10, the statistic calculation unit 38 has an average value μ = 57.8 mmHg, a variance σ ² = 536, a standard deviation σ = 23.2, a peak value Vp = 99 mmHg, a peak The degree Vk = −1.15 is calculated.

The evaluation unit 44 determines the deduction point E ₃ as follows based on the above calculation result.
First, the evaluation unit 44 sets ΔVp = abs | 198−99 | = 99, ΔVk = abs | 3.34 − (− 1.15) | = 4.49, Δσ = abs | 33 for ΔVp, ΔVk, and Δσ. .6-23.2 | = 10.4. Here, abs | x | is a function indicating the absolute value of x.

The evaluation unit 44 applies the calculated ΔVp, ΔVk, Δσ to the above equation (4) to determine the deduction point E ₃ . For example, for k _{5 to} k ₇ , when k ₅ = 0.1, k ₆ = 1.0, k ₇ = 1.0, and K = 2.0, the deduction points E ₃ = 8.5.

Next, the deduction points E ₄ will be described with reference to FIGS. The evaluation unit 44 determines the deduction point E ₄ based on the symmetry of the pressure distribution on the seating surface determined by the symmetry determination unit 40. The symmetry determination unit 40 estimates which of the predetermined classifications the symmetry of the pressure distribution corresponds to, and sets the deduction point assigned to the estimated classification as the value of E ₄ . Thereby, when the seating state is a predetermined specific posture, the evaluation according to the specific posture is reflected.

  FIG. 12 is a table showing an example of the classification specified by the symmetry determining unit 40. Here, the symmetry determination unit 40 predicts the sitting posture of the person 12 based on the objectivity between the regions when the seating surface is divided into four regions from the first quadrant to the fourth quadrant (see FIG. 13). Identify the posture to be done.

To explain based on the example shown in FIG. 12, for example, the symmetry determination unit 40 compares the pressure distribution in the first quadrant and the pressure distribution in the second quadrant, and the pressure distribution in both regions becomes a predetermined pattern. If applicable, the person 12 is estimated to be in a posture with one leg crossed. In this case, the evaluation unit 44 sets the value of the deduction point E ₄ to 5.

In addition, the symmetry determination unit 40 compares the pressure distribution in the third quadrant and the pressure distribution in the fourth quadrant, and if the pressure distributions in both regions correspond to a predetermined pattern, the person 12 balances the left and right. In this case, the evaluation unit 44 sets the value of the deduction point E ₄ to 7.

The pressure distribution in the first quadrant and the pressure distribution in the second quadrant correspond to a predetermined pattern that is assumed to be a posture in which one leg is crossed, and the pressure distribution in the third quadrant and the pressure distribution in the fourth quadrant In a case where a predetermined pattern estimated to be sitting out of balance is satisfied, the symmetry determination unit 40 sits with the person 12 crossing one leg and out of balance between the left and right sides. In this case, the evaluation unit 44 sets the value of the deduction point E ₄ to 9.

In addition, the symmetry determination unit 40 compares the pressure distribution in the first quadrant and the fourth quadrant with the pressure distribution in the second quadrant and the third quadrant, and when the pressure distribution in both regions corresponds to a predetermined pattern. Is estimated that the person 12 is in a posture with one elbow. In this case, the evaluation unit 44 sets the value of the deduction point E ₄ to 7.

When there are a plurality of postures (corresponding classifications) that are estimated to be applicable by the determination by the symmetry determination unit 40, the evaluation unit 44 determines the maximum value among the deduction points assigned to each posture (classification). Is the value of E ₄ .

  In the present embodiment, the center-of-gravity position detection unit 36 detects whether the person 12 is seated at any position in the front-rear direction of the seat surface (whether the seat position is shallower or deeper than the reference position of the seat surface). However, the symmetry determination unit 40 may determine whether the seating position is shallower or deeper than the reference position of the seating surface. For example, the pressure distribution in the first quadrant and the pressure distribution in the fourth quadrant are compared, the pressure distribution in both regions corresponds to a predetermined pattern, and the pressure distribution in the second quadrant and the pressure distribution in the third quadrant In comparison, when the pressure distributions in both regions correspond to a predetermined pattern, the person 12 may be estimated to be seated at a position shallower than the reference position.

  Next, a specific example of determination by the symmetry determination unit 40 will be described with reference to FIGS. 13 and 14. For example, the symmetry determination unit 40 determines whether or not the shape of each histogram in the x direction and the y direction of the pressure distribution in each region corresponds to a predetermined pattern.

  FIG. 13 is a schematic diagram illustrating an example of a pressure distribution when a region is divided from the first quadrant to the fourth quadrant as a seating surface division when determining symmetry. Here, the seat surface is divided into four regions, with the center in the left-right direction and the center in the front-rear direction as a boundary. FIG. 14 is a schematic diagram showing the histograms of the first quadrant and the second quadrant. FIG. 14A shows the histograms in the x and y directions in the first quadrant, and FIG. 14B shows the histogram in the second quadrant. Histograms in the x and y directions are shown.

  In FIG. 14, the x-direction histogram is the sum of the pressure values of all y-coordinates in the region in each x-coordinate (the sum of the pressure values in the y-direction for each x-coordinate). ). The y-direction histogram is the sum of the pressure values of all the x coordinates in the region in each y coordinate (the sum of the pressure values in the x direction for each y coordinate).

For example, the symmetry determining unit 40 determines that the combination of the shapes of the four histograms shown in FIG. 14 corresponds to a predetermined pattern, and sets one leg as the posture estimated from the pressure distribution shown in FIG. Estimated to be in a sitting position. On the other hand, the evaluation unit 44 determines a predetermined deduction point 5 when it is estimated that the posture is one leg crossed as the value of E ₄ .

  Although the estimation using the histogram has been described here, the method for determining the symmetry of the pressure distribution is not limited to the histogram. For example, the symmetry between each region of the pressure distribution may be determined by calculating a statistic for each region and comparing the calculated statistic for each region. For example, the variance may be calculated for each region, the degree of asymmetry may be determined according to the magnitude of the difference between the variance values between the regions, and a deduction point may be set according to the degree of asymmetry.

  Next, a display example by the display unit 46 will be described. FIG. 15 is a schematic diagram illustrating an example of a screen displayed on the UI unit 22 by the display unit 46. The display unit 46 displays the score of the evaluation value E calculated by Expression (1) as the evaluation result of the evaluation unit 44. Further, as shown in FIG. 15, the display unit 46 may illustrate the position information acquired by the position acquisition unit 30 or the pressure distribution acquired by the pressure acquisition unit 34.

Next, the operation of the posture determination system 2 will be described with reference to FIGS.
FIG. 16 is a flowchart illustrating the operation of the posture determination system 2. As shown in FIG. 16, the posture determination system 2 first acquires reference data at a posture serving as a reference when determining the posture of the person 12 (S <b> 10), and then determines the posture of the person 12. (S20).

  At this time, it is preferable that the determination target person 12 is seated on the seating surface in a reference posture and the reference data is acquired. However, the reference data is not necessarily the data of the sitting by the determination target person 12. . Moreover, it is preferable that the seating surface to be seated in step 10 is the same as the seating surface to be seated in the posture determination. Thereby, since the measurement conditions by the pressure measuring device 8 can be made uniform at the time of acquisition of the reference data and at the time of posture determination, the influence on the measurement due to the elasticity of the seating surface is suppressed. Details of step 10 will be described with reference to FIG. Details of step 20 will be described with reference to FIG.

  FIG. 17 is a flowchart illustrating an example of the reference data acquisition operation (S10) in the posture determination system 2.

  In step 100 (S100), the photographing device 6 photographs the person 12 seated in a reference posture.

  In step 102 (S102), the position acquisition unit 30 estimates the position of the body part from the captured image captured in step 100, and acquires position information.

  In step 104 (S104), the inclination detector 32 detects the inclination of the body of the person 12 based on the position information acquired in step 102.

  In step 106 (S106), the reference data storage unit 18 stores body tilt data during a reference posture.

  In step 108 (S108), the pressure measurement device 8 measures the pressure of the seating surface in the seating in the reference posture, and the pressure acquisition unit 34 acquires the pressure distribution.

  In step 110 (S110), the statistic calculation unit 38 calculates a predetermined statistic from the measurement data acquired in step 108.

  In step 112 (S112), the reference data storage unit 18 stores the statistical data for the posture at the reference position calculated in step 110.

  As described above, as an example of the flowchart of the reference data acquisition operation, the example of performing the pressure measurement and the processing based on the measurement data after the imaging processing and the processing based on the imaging data has been described. These operations may be performed in parallel.

  FIG. 18 is a flowchart illustrating an example of a posture determination operation (S20) in the posture determination system 2.

  In step 200 (S200), the photographing device 6 photographs the person 12 sitting in the determination target posture.

  In step 202 (S202), the position acquisition unit 30 estimates the position of the body part from the captured image captured in step 200, and acquires position information.

  In step 204 (S204), the inclination detection unit 32 detects the inclination of the body of the person 12 based on the position information acquired in step 202.

  In step 206 (S206), the pressure measuring device 8 measures the pressure of the seating surface in the seating in the determination target posture, and the pressure acquisition unit 34 acquires the pressure distribution.

  In step 208 (S208), the barycentric position detection unit 36 detects the barycentric position of the pressure applied to the seating surface by the seated person 12 from the measurement data acquired in step 206.

  In step 210 (S210), the statistic calculation unit 38 calculates a predetermined statistic from the measurement data acquired in step 206.

  In step 212 (S212), the symmetry determination unit 40 determines the symmetry of the pressure distribution acquired in step 206.

  In step 214 (S214), the evaluation unit 44 calculates a difference between the reference data stored in the reference data storage unit 18 in steps 106 and 112 described above and the data to be determined in steps 204 and 210.

  In step 216 (S216), the evaluation unit 44 calculates an evaluation value E based on the equation (1).

  In step 218 (S218), the display unit 46 displays the evaluation result by the evaluation unit 44 on the UI unit 22.

  As described above, as the flowchart of the posture determination operation, the example of performing the measurement based on the pressure measurement and the measurement data after the photographing process and the process based on the photographing data has been shown, but the order of these operations may be reversed, These operations may be performed in parallel.

  As mentioned above, although the example which performs attitude | position determination about the person 12 seated on the chair 10 was demonstrated as embodiment of this invention, the seating of the person 12 is not restricted to the seating on the chair 10, but if it is sitting on the seat surface It may be sitting on the floor or tatami mat. Further, the posture determination system 2 may evaluate not only the sitting method shown in FIG. 1 but also an arbitrary sitting method such as a straight seat and a crossed seat.

  Further, in the above-described embodiment, the case has been described in which the imaging device 6 captures images from the front, back, or side of the person 12, but the shooting direction may not be the front, back, or side of the person 12.

  Moreover, although the position acquisition part 30 demonstrated as what acquires the positional information estimated from the imaging | photography data obtained by imaging | photography with the imaging device 6, a marker is attached to the predetermined site | part of the body of the person 12, The position information may be acquired by measuring the position of this marker.

2 posture determination system 4 terminal device 6 imaging device 8 pressure measurement device 30 position acquisition unit 32 inclination detection unit 34 pressure acquisition unit 36 center of gravity position detection unit 38 statistic calculation unit 40 symmetry determination unit 42 reference data storage unit 44 evaluation unit 46 Display section

Claims (12)

Position acquisition means for acquiring position information of a predetermined part of the person's body, which is estimated from a captured image of a person in a seated state;
Pressure acquisition means for acquiring pressure information applied to a seating surface on which the person sits;
A posture determination apparatus comprising: an evaluation unit that evaluates the posture of the person in the sitting state based on the position information acquired by the position acquisition unit and the pressure information acquired by the pressure acquisition unit. Inclination detecting means for detecting the inclination of the body of the person based on the position information acquired by the position acquiring means;
The posture determination apparatus according to claim 1, wherein the evaluation unit evaluates the posture of the person in the sitting state based on the inclination detected by the inclination detection unit and the pressure information acquired by the pressure acquisition unit. The position acquisition means acquires information indicating a position on at least a two-dimensional coordinate of a predetermined part of the person's body, which is estimated from a captured image taken from a side direction of the person,
The detection means detects the degree of inclination of the person in the front-rear direction of a predetermined part based on the two-dimensional coordinate value of the predetermined part acquired by the position acquisition means,
The posture determination apparatus according to claim 2. The position acquisition means acquires information indicating a position on a three-dimensional coordinate of a predetermined part of the person's body, which is estimated from a captured image taken from the front or back direction of the person,
The detection means detects the degree of inclination of the person in the front-rear direction of a predetermined part based on the three-dimensional coordinate value of the predetermined part acquired by the position acquisition means,
The posture determination apparatus according to claim 2. The evaluation means evaluates the appropriate degree of posture lower than the inclination of the person's body in the forward direction with respect to the inclination of the person's body in the forward direction.
The posture determination apparatus according to any one of claims 2 to 4. The posture determination apparatus according to any one of claims 1 to 5, wherein the evaluation unit evaluates according to a deviation from a reference position with respect to a position of a center of gravity calculated from pressure information acquired by the pressure acquisition unit. The posture determination apparatus according to claim 6, wherein the evaluation unit evaluates the degree of appropriateness of the posture with respect to the shift of the position of the center of gravity to the front of the seat surface, lower than the shift to the rear of the seat surface. The posture determination apparatus according to any one of claims 1 to 7, wherein the evaluation unit evaluates according to a deviation from a reference value with respect to a pressure distribution on the seating surface acquired by the pressure acquisition unit. The posture determination apparatus according to any one of claims 1 to 8, wherein the evaluation unit performs evaluation according to a symmetry of a pressure distribution on a seating surface acquired by the pressure acquisition unit. The posture determination apparatus according to claim 1, further comprising display means for displaying an evaluation result by the evaluation means. A photographing device for photographing a person in a sitting state, a measuring device for measuring pressure applied to a seating surface on which the person is seated, and a posture determination device,
The posture determination device includes:
Position acquisition means for acquiring position information of a predetermined part of the body of the person, which is estimated from a captured image captured by the imaging apparatus;
Tilt detecting means for detecting the tilt of the person's body based on the position information acquired by the position acquiring means;
Pressure acquisition means for acquiring pressure information measured by the measuring device;
A posture determination system comprising: an evaluation unit that evaluates the posture of the person in the sitting state based on the tilt detected by the tilt detection unit and the pressure information acquired by the pressure acquisition unit. A position acquisition step of acquiring position information of a predetermined part of the body of the person, which is estimated from a captured image obtained by capturing a person in a sitting state;
A pressure acquisition step of acquiring pressure information applied to a seating surface on which the person sits;
A program for causing a computer to execute an evaluation step of evaluating the posture of the person in the sitting state based on the acquired position information and the acquired pressure information.

Images