JP2001052202A - Human body action visualizing device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To reduce the mount load of an action detecting sensor and to improve the visibility and recognizability of a concerned spot of three-dimensional computer graphic display by assembling a human body part model within a prescribed virtual space and forming a human body model corresponding to the acting attitude of a testee. SOLUTION: An action detecting part 2 generates human body action data expressing the position coordinate and direction of the human body part of the testee wearing a plurality of action detecting sensors 2b. A human body part model generation part 4 generates a joint model expressing joint positions where the sensors 2b are mounted, and a human body part model defined between these joints. A human body model forming part 7 synthesizes the human body by linking the human body part models and the joint models and arranges it within the prescribed virtual space. Then, the respective human body part models are moved by moving and rotating the respective joint models on the basis of human body action data and an action following the action of the testee is executed.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a human body movement visualizing apparatus for displaying a three-dimensional computer graphic of a subject's movement using human body movement data detected by a movement detection sensor attached to a predetermined part of the subject.

[0002]

2. Description of the Related Art In a conventional human body motion visualization apparatus, a human body model is animated using human body motion data detected by a motion detection sensor attached to a predetermined part of a subject, that is, a human body is displayed by three-dimensional computer graphic display. When operating the model, it is necessary to attach a motion detection sensor to the subject so as to correspond one-to-one with the joint points set in the human body model, and observe only the movement of some human body parts such as legs and arms If you do,
The motion detection sensors have to be mounted at the positions corresponding to all the joint points.

In the conventional human body motion visualization device, as described above, the subject must wear motion detection sensors at locations corresponding to all joint points. In the case of measuring a biological load and the like at the same time as detecting the movement, an unnecessary load due to the attachment of the movement detection sensor is applied to the subject, which hinders normal measurement of the biological load.

Further, when visualizing the human body motion data via a human body model, a human body part other than the point of interest is displayed, which hinders the observation of the movement of the point of interest. That is, a human body part other than the point of interest may be displayed in front of the point of interest, and there is a disadvantage that only the point of interest cannot be observed from various angles.

Further, most of the conventional human body motion visualization apparatuses employ a real human body model specialized for animation generation, and lack visibility of where the measurement location is on the human body model. Was something. In addition, there is a problem that the torso portion is formed of one human body portion, and the torso portion cannot be bent or twisted.

[0006]

SUMMARY OF THE INVENTION The present invention has been made in view of the above-described problems, and has as its object to use human body motion data detected by a motion detection sensor attached to a predetermined position of a subject. And a human body movement visualization device capable of reducing the mounting load of the motion detection sensor and improving the visibility and visibility of a point of interest at the time of displaying the three-dimensional computer graphic when displaying the motion of the subject in three-dimensional computer graphics. The point is to provide.

[0007]

A first feature of the human body movement visualizing device according to the present invention for achieving the above object is as described in claim 1 of the claims. Using the human body motion data detected by the motion detection sensor attached to the location,
A human body movement visualization device that displays a three-dimensional computer graphic, comprising: a human body part model generation unit that specifies a human body part to be observed based on mounting position information of the motion detection sensor and generates the human body part model; A human body model forming unit that assembles the human body part model generated by the part model generating unit in a predetermined virtual space and forms a human body model corresponding to the motion posture of the subject based on the human body motion data; and And a human body model display unit for displaying all or a part of the human body model formed by the forming unit on a predetermined display device.

[0008] In the second feature configuration, as described in claim 2 of the claims, in addition to the first feature configuration, the mounting position of the sensor in the human body movement data is determined by the subject. A data correction unit for correcting the human body motion data so as to match the corresponding position of the human body model having the same posture as that of the human body model.

The third feature configuration is, in addition to the first or second feature configuration, described in claim 3 in the claims section, in addition to the human body model formed by the human body model forming section. The present invention is characterized in that a display part selection unit for selecting a part to be displayed on the human body model display unit from the human body parts is provided.

[0010] In the fourth feature configuration, in addition to the first, second or third feature configuration, a joint position is different from that of the human body model, as described in claim 4 of the claims. The point is that it is highlighted in a special shape.

[0011] The fifth feature configuration is, in addition to the first, second, third or fourth feature configuration as described in claim 5 of the claims section, a torso of the human body model. The point is that the part consists of at least two human body part models connected vertically.

The operation and effect will be described below. According to the first characteristic configuration of the human body motion visualization device according to the present invention,
When observing the movement of a subject by focusing on a specific body part of the subject, the load on the subject to be fitted with the motion detection sensor can be reduced, and only the movement of the subject at the point of interest is selectively displayed in three-dimensional computer graphics. This improves visibility when observing a point of interest from various angles.

A specific description will be given of a case where the movement of both hands and arms of a subject is observed. First, the motion detection sensors are attached to the left and right shoulder joints, elbow joints, and wrist joints at both ends of the subject's upper arm and forearm. The human body part model generation unit generates each human body part model of the left and right upper arm and forearm connected to each joint at the mounting position of the motion detection sensor. The left and right joints of the shoulder, elbow, and wrist may be generated as independent part models, or may be treated as part models attached to the upper arm or the forearm.

For example, the human body model forming section connects the upper arm and the forearm to each other with left and right elbow joints and wrist joints to generate a human body model of both hands and arms, and detects the left and right shoulders detected by the motion detection sensor. Based on the human body motion data representing the changes in the positions and orientations of the joints, elbow joints, and wrist joints, the human body model with both hands and arms is operated by changing the position and orientation of each human body part model in a predetermined virtual space Let it. The human body model display unit displays the operation of the human body model of both hands and arms in a three-dimensional computer graphic display.

As a result, since only the movement of both hands and arms, which is the point of interest, is displayed in three-dimensional computer graphics, it is possible to observe the point of interest from various angles without obstructing the view of other human body parts. For example, when observing the motion of both hands and arms in front of the chest from the back side, the torso portion is not displayed, so that such motion can be visualized. In addition, the subject only needs to attach the motion detection sensors to the left and right shoulder joints, elbow joints, and wrist joints, and does not need to attach the motion detection sensors to other parts, so that the load on the motion detection sensors can be reduced.

According to the second characteristic configuration, when the human body model is operated based on the human body operation data, a more accurate and smooth operation display according to the operation of the subject can be performed. For example, when one human body model is prepared and the motions of many subjects are individually displayed, individual differences such as the size and ratio of the human body part of each subject, displacement of the mounting position of the motion detection sensor, and measurement conditions Even if the position of attachment of the motion detection sensor and the joint position on the human body model cannot be uniquely associated due to the variation of In order to correct the human body movement data so that the correspondence with the position can be reliably performed, even for subjects with different physiques,
Alternatively, accurate and smooth operation display can be performed even when the measurement environment changes. As a result, when a human body model at a point of interest is generated from one whole body model, the same human body model can be commonly used for a plurality of subjects.

According to the third characteristic configuration, the point of interest can be further selected and displayed, so that the visibility of the specific portion is further improved. For example, only the right hand and the right arm can be displayed with respect to the human body model of the two hands and the two arms, so that observation focusing on the right hand right arm is possible, and observation focusing on the left hand left arm is also possible.

According to the fourth characteristic configuration, since the measurement location of the subject can be easily visually recognized, the motion display of the subject can be displayed when, for example, the motion of the subject and the biological load are simultaneously measured. The observation can be easily associated with the measurement result of the biological load while observing.

According to the fifth characteristic configuration, by attaching the motion detection sensor to the back of the subject or the like, it is possible to display a three-dimensional computer graphic display of the bending and twisting of the body part.

[0020]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the human body movement visualizing device according to the present invention will be described below with reference to the drawings.

As shown in FIG. 1, a human body movement visualizing device 1
Is a motion detecting unit 2, a mounting position specifying unit 3, a human body part model generating unit 4, a human body motion data input unit 5, a data correcting unit 6, a human body model forming unit 7, a display part selecting unit 8,
And a human body model display unit 9.

The operation detecting section 2 includes one or more magnetic sensors, and one magnetic sensor includes:
One magnetic signal generating unit 2a and a magnetic signal generated by the magnetic signal generating unit 2a attached to a predetermined human body part of a subject at a predetermined sampling interval (for example, 50 millisecond intervals)
A plurality of motion detection sensors 2b that detect motion by detecting the position coordinates and orientation of the mounting position in time series
And controlling the operation of the magnetic signal generating unit 2a and the plurality of motion detection sensors 2b, receiving the detection results of the plurality of motion detection sensors 2b, and measuring the subject wearing the plurality of motion detection sensors 2b. A signal processing unit 2c for generating human body motion data representing the position coordinates and orientation of the human body part.

The mounting position specifying unit 3 specifies the mounting position of the motion detection sensor 2b mounted at the joint position at both ends corresponding to the human body part to be observed by an external input operation, and the specified mounting position. The information is output to the human body part model generation unit 4.

The human body part model generation section 4 generates a joint model representing the joint position where the motion detection sensor 2b is mounted and a human body part model defined between the joints from the mounting position information. Each human body part model is a model that can be connected to each other via a specific one of the joint models, and has a structure that can be expanded and contracted in accordance with the position of the joint model. Also, a direction vector representing the direction is assigned to the joint model. Here, the direction vector indicates the mounting direction of each of the motion detection sensors 2b, and is defined as a vector from the mounting surface toward the joint center. The human body part model and the joint model of each part of the human body are selected and used from those prepared in advance as a whole body model.

For example, as shown in FIG. 2, when one of the right and left arms is to be observed, a human body part model 1
As 0, an upper arm model 10a and a forearm model 10b are generated, and as a joint model 11, a shoulder joint 11a, an elbow joint 11b, and a wrist joint 11c are generated. The upper arm model 10a is a human body part model 10 to be arranged between the shoulder joint 11a and the elbow joint 11b, and the forearm model 10b is a human body model to be arranged between the elbow joint 11b and the wrist joint 11c. This is a part model 10. The shoulder joint 11a
The direction vectors 12a, 12b, and 12c are assigned to the elbow joint 11b and the wrist joint 11c, respectively. The direction vector 12a is the mounting direction of the motion detection sensor 2b mounted on the shoulder joint 11a, the direction vector 12b is the mounting direction of the motion detection sensor 2b mounted on the elbow joint 11b, and the direction vector 12c is This is the mounting direction of the motion detection sensors 2b mounted on the wrist joint 11c, and is detected by each motion detection sensor 2b. In the present embodiment, the forearm model 10b is a simplified model and integrally includes a hand. FIG.
, The motion detection sensor 2b is connected to the direction vectors 12a,
12b and 12c, the motion detection sensor 2b is actually mounted on each corresponding joint of the subject, and the motion detection sensor 2b and the direction vector 1
The indications 2a, 12b, and 12c are added for convenience of explanation. Therefore, in the three-dimensional computer graphic display, the motion detection sensor 2b and the direction vector 12
a, 12b, and 12c are not necessarily displayed. However,
Even if these are displayed, it does not matter.

As shown in FIG. 3, for example, as a human body part model 10, a head model 20, a chest model 21, an abdominal model 22, a waist model 23, left and right Upper arm model 10a,
A left and right forearm model 10b, a left and right thigh model 24, and a left and right lower leg model 25 are provided. As the joint model 11, a neck joint 26, a back joint 27, a waist joint 28, a left and right shoulder joint 11a, a left and right elbow joint 11b , Left and right wrist joints 11
c, left and right hip joints 29, left and right knee joints 30, and left and right ankle joints 31.

The human body motion data input unit 5 receives the human body motion data from the signal processing unit 2c of the motion detection unit 2. The received human body motion data is subjected to a correction process described later in the data correction unit 6.

The human body model forming section 7 combines the human body part model 10 and the joint model 11 generated by the human body part model generating section 4 to synthesize a human body model 13,
It is arranged in a predetermined virtual space. Further, based on the human body motion data that has been subjected to the correction processing by the data correction unit 6 with respect to the synthesized human body model 13, the position coordinates and orientation of each joint position of the human body motion data are converted into the respective joints. By moving and rotating each of the joint models 11 as the position coordinates and the direction vector of the model 11, each of the human body part models 10 connected thereto moves, and as a result, the human body model 13 responds to the human body motion data. By taking a posture and sequentially updating the human body motion data, the motion following the motion of the subject is executed.

The display part selecting section 8 selects a part to be displayed on the human body model display section 9 from the human body parts of the human body model 13 formed by the human body model forming section 7 based on an external operation input. It should be noted that the display site selection unit 8
In a default state where there is no external operation input, the entire human body model 13 formed by the human body model forming section 7 is selected.

The human body model display section 9 displays the operation of the human body part selected by the display part selecting section 8 in the human body model 13 formed by the human body model forming section 7 on a predetermined display device 9a. The data processing for displaying the operation of the human body model 13 formed by the human body model forming unit 7 in a three-dimensional computer graphic manner uses an existing image processing technology.

The above-described human body movement visualizing device 1 is specifically composed of a computer system such as a personal computer. The external operation input is performed from an input terminal such as a keyboard attached to the computer.
a uses a CRT monitor or the like attached to the computer.

Next, the correction processing of the human body motion data by the data correction section 6 will be described. For the sake of simplicity, the motion detection sensors 2b are attached to the left and right shoulders, elbows and wrists.
A description will be given of an example of the correction processing of the human body motion data when the camera is worn.

In this correction process, the position coordinates and the orientation of the mounting position detected by the motion detection sensor 2b in the reference posture such as "watch out" by the subject 14 are determined by the corresponding joints of the human body model 13 in the same reference posture state. Data conversion processing is performed so that the position coordinates of the position joint model coincide with the direction vector. The data conversion process is classified into a batch conversion process for collectively performing all the human body motion data and an individual conversion process for the mounting position of each motion detection sensor 2b.

First, the batch conversion process will be described. This batch conversion process includes a center-of-gravity shift process, a rotation process, and a scaling process.

As shown in FIG. 4, the center-of-gravity shift processing is performed by a center of gravity (hereinafter referred to as a data center of gravity) A determined from all the mounting positions (marked by ◎ in the figure) of the motion detection sensor 2b in the human body motion data, as shown in FIG. Is the center of gravity (hereinafter, referred to as a model center of gravity) B of the corresponding joint position of the human body model 13 (indicated by ● in the figure).
To translate. Specifically, data gravity center coordinates are calculated from the position coordinates of each of the motion detection sensors 2b in the reference posture state of the human body motion data received by the human body motion data input unit 5, and the model gravity center is determined as the origin of the predetermined virtual space. Is the translation vector M from the data center of gravity to the model center of gravity.
Is calculated. Here, since the center of gravity of the model is located at the origin, the translation vector M is equivalent to the one obtained by inverting the polarity of the three-dimensional coordinates of the center of gravity of the data. Then, by adding the parallel movement vector M to each coordinate data of the human body motion data, the center of gravity movement processing is executed. Although FIG. 4 shows the movement in a two-dimensional plane for simplification, three-dimensional processing is performed in the actual calculation. Therefore, the subject 14 and the human body model 13 based on the human body motion data are not necessarily oriented in the same direction. FIG. 5 shows the human body motion data after the center of gravity shifting process.

As shown in FIG. 5, after the process of moving the center of gravity, the mounting direction of each of the motion detection sensors 2b does not match the direction vector of the corresponding joint model. Here, instead of individually correcting the direction of each of the motion detection sensors 2b in the human body motion data, the entire human body motion data is collectively rotated around the data center of gravity after the center of gravity movement processing. This rotation process is performed on the x-axis, y-axis,
The rotation processing about the z-axis is performed separately. Specifically, the direction of each of the motion detection sensors 2b of the human body motion data is synthesized as a unit vector to generate a data synthesized vector. In addition, the direction vector of each joint model is synthesized as a unit vector to generate a model synthesized vector. The rotation processing is performed so that the data composite vector matches the model composite vector. FIG. 6 shows the human body motion data after the rotation processing.

As shown in FIG. 6, even after the rotation processing, due to the physique difference between the subject 14 and the human body model 13,
Since the position of the joint model corresponding to each of the motion detection sensors 2b does not completely match, the height ratio of the human body model 13 of the whole body and the height of the subject 14 is calculated, and the ratio is used as the human body motion data after the rotation process. A scaling process is performed by multiplying the position coordinates of each motion detection sensor 2b.

The human body movement data received by the human body movement data input unit 5 is subjected to the above-described series of batch conversion processing, so that the mounting position and direction of each of the movement detection sensors 2b can be substantially changed. The position and direction vector of the corresponding joint model of the model 13 match. However,
Due to the ratio of each human body part in the physique difference between the subject 14 and the human body model 13 and the displacement of the mounting position and the orientation of the motion detection sensor 2b, they do not completely match. Is individually corrected. Specifically, movement processing of the coordinate point of each mounting position and rotation processing of the direction vector are performed.

As described above, the human body model forming section 7
The position coordinates and the direction vector of each joint model 11 are determined based on the human body motion data after the correction processing by the data correction unit 6, and the human body part model 10 corresponding between the joint models 11 is determined. The posture of the human body model 13 is determined by arranging it. However, since the correction processing is performed in the reference posture, in the posture during the movement of the subject, for example, the displacement of the mounting position and the orientation of the movement detection sensor 2b due to the expansion and contraction of the skin of the subject, and the movement detection Due to the detection error of the part 2, a slight error may occur in the position and orientation of the joint model 11 and the connection with the human body part model 10 may not be possible.
Has a structure that can expand and contract in the longitudinal direction. As a result, the human body model 13 can perform a smooth motion following the motion of the subject.

Next, an example of a three-dimensional computer graphic display by the human body model display unit 9 will be described.

FIG. 7 shows a human body model 13 when all the human body parts of the whole body model shown in FIG.
2 shows a display example in which the walking motion of the user is looked down from the upper right front. In this case, FIG. 8 shows a display example when the upper body is selected by the display part selection unit 8, and FIG. 9 shows a display example when both arms and legs are selected. In this way, by selecting and displaying a point of interest in the display site selection unit 8, individual movements can be observed in more detail.

Further, as shown in FIG. 3, in the present embodiment, the torso portion is connected to the abdomen model 22 and the waist model 2
3 and the back joint 27 connecting these two human body part models. Here, the motion detection sensor 2b corresponding to the back joint 27 is mounted at the center of the back. With this configuration, as shown in FIG. 10, the front and rear bends (a), left and right bends (b), and torsion (c) of the body portion can be expressed more finely.

Also, as shown in FIG. 3, by highlighting the joint model 11 with a sphere, the measurement site by the motion detection sensor 2b can be clearly identified.

Another embodiment will be described below. <1> In each of the above embodiments, the configuration and the shape of the human body model 13 are shown in FIG. 3, but are not necessarily limited to the configuration and the shape of the present embodiment.

<2> Further, the correction processing of the human body motion data in the data correction section 6 is not limited to the above-described procedure.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a system configuration of a human body motion visualization device according to the present invention.

FIG. 2 is an explanatory diagram for explaining a configuration of a human body part model and a human body model;

FIG. 3 is a perspective view showing the whole body of a human body model.

FIG. 4 is an explanatory diagram for explaining a center-of-gravity shift process in a correction process;

FIG. 5 is an explanatory diagram for explaining rotation processing in correction processing;

FIG. 6 is an explanatory diagram for explaining a scaling process in a correction process;

FIG. 7 is a perspective view showing a whole body model during a walking operation.

FIG. 8 is a perspective view showing an example of selective display of a part of a human body model;

FIG. 9 is a perspective view showing another selected display example of a part of the human body model;

FIG. 10: Bending before and after the torso part of the human body model (A)
Perspective view showing left and right bending (b) and torsion (c)

[Explanation of symbols]

 Reference Signs List 1 human body motion visualization device 2 motion detection unit 2a magnetic signal generation unit 2b motion detection sensor 2c signal processing unit 3 mounting position designation unit 4 human body part model generation unit 5 human body motion data input unit 6 data correction unit 7 human body model formation unit 8 display Part selection unit 9 Human body model display unit 9a Display device 10 Human body part model 10a Upper arm model 10b Forearm model 11 Joint model 11a Shoulder joint 11b Elbow joint 11c Wrist joint 12a, 12b, 12c Direction vector 13 Human body model 14 Subject

 ────────────────────────────────────────────────── ─── Continued on the front page (72) Hirokazu Onoe Inventor Hirokazu Onoe 17 Shinjuku-ji Minamimachi, Shimogyo-ku, Kyoto, Kyoto F-term in Kansai Research Institute of Technology (reference) 4C038 VA04 VB01 VC20 5B050 BA07 BA09 BA12 EA05 EA06 EA12 EA13 EA28 FA02 FA06

Claims (5)

[Claims] 1. A human body motion visualization device for displaying a motion of the subject in a three-dimensional computer graphic using human body motion data detected by a motion detection sensor attached to a predetermined portion of the subject, Specifying a human body part to be observed based on the mounting position information, a human body part model generation unit that generates the human body part model, and the human body part model generated by the human body part model generation unit in a predetermined virtual space. Assembling, a human body model forming unit that forms a human body model corresponding to the motion posture of the subject based on the human body motion data, and all or a part of the human body model formed by the human body model forming unit is displayed on a predetermined display device. A human body motion visualization device comprising a human body model display unit for displaying. 2. A data correction unit for correcting the human body motion data so that a mounting position of the sensor in the human body motion data coincides with a corresponding position of the human body model having the same posture as the subject. 2. The human body movement visualization device according to 1. 3. The human body motion visualization according to claim 1, further comprising a display part selection unit that selects a part to be displayed on the human body model display unit from the human body parts of the human body model formed by the human body model formation unit. apparatus. 4. The human body motion visualization device according to claim 1, wherein the joint position is highlighted with a special shape different from the human body model. 5. The apparatus according to claim 1, wherein the body part of the human body model comprises at least two human body part models connected vertically.