Animation generation system, animation generation method, and program

Abstract

To allow for easily creating an animation of natural movement of a person who uses a tool to take an action on an object.SOLUTION: An animation creation system comprises: a position information acquisition unit which acquires position information of a prescribed tool at the time of taking a prescribed action on an object with the prescribed tool; a blend processing unit which calculates position information of a hand of a person based on the position information of the prescribed tool or attitude information of a trunk of the person on the basis of the position information of the prescribed tool and, on the basis of the attitude information of the trunk, uses blend processing to create an animation of the trunk; and an inverse kinematics calculation processing unit which, on the basis of the position information of the hand and the attitude information of the trunk which are calculated by the blend processing unit, uses the inverse kinematics calculation processing to calculate attitude information of an arm so as to connect the trunk and the hand of the person by the arm and creates an animation of the arm on the basis of the calculated attitude information of the arm.SELECTED DRAWING: Figure 1

Description

Technical Field

[0004] , ,

[0006] , , , ,

[0005] , , , ,

[0001] The present disclosure relates to an animation generation system, an animation generation method, and a program for generating an animation of a person in motion.

Background Art

[0002] An animation generation system for generating an animation of a person in motion is known (see, for example, Patent Document 1).

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] For example, a person may perform a predetermined operation such as a processing operation on surrounding objects using a tool. When attempting to generate an animation of such a person, adjustment of blend conditions is performed to adjust the relative positional relationship between the person and the tool, but this adjustment may take a lot of time and effort.

[0005] In contrast, the above animation generation system can reduce the time and effort of adjusting the blend conditions by using inverse kinematics calculation processing to make the person's movement follow the movement of the tool. However, since the person's movement follows the movement of the tool, there is a risk that the person will be in a state of being pulled by the tool, the natural posture of the person will be disrupted, and the movement will become unnatural.

[0006] This disclosure was made to solve these problems, and its main purpose is to provide an animation generation system, animation generation method, and program that can easily generate animations of a person's natural movements when they use a tool to perform actions on an object. [Means for solving the problem]

[0007] One aspect of this disclosure for achieving the above objectives is: An animation generation system that generates an animation of a person performing a predetermined action on an object around them using a predetermined tool, A position information acquisition unit acquires position information of the predetermined tool when the predetermined tool performs the predetermined operation on the object, A blending processing unit calculates posture information of the person's torso based on the position information of a predetermined tool acquired by the position information acquisition unit, or based on the position information of the predetermined tool, and generates an animation of the torso using blending processing based on the calculated posture information of the torso. An inverse kinematics processing unit calculates the position information of the arm and the position information of the torso based on the position information of the hand calculated by the blend processing unit and the posture information of the torso, using inverse kinematics calculation processing to connect the torso and hand of the person with the arm, and generates an animation of the arm based on the calculated posture information of the arm. It is an animation generation system equipped with [features / equipment]. One aspect of this disclosure for achieving the above objectives is: An animation generation method for generating an animation of a person performing a predetermined action on an object around them using a predetermined tool, The steps include: acquiring positional information of the predetermined tool when the predetermined tool performs the predetermined operation on the object; The steps include: calculating posture information of the person's torso based on the position information of the person's hand based on the position information of the predetermined tool obtained, or based on the position information of the predetermined tool, and generating an animation of the torso using blend processing based on the calculated posture information of the torso; Based on the calculated hand position information and the torso posture information, the steps include: calculating the arm posture information using inverse kinematics processing so as to connect the torso and hand of the person with the arm; and generating an animation of the arm based on the calculated arm posture information; This is an animation generation method that includes [the following]. One aspect of this disclosure for achieving the above objectives is: A program that generates an animation of a person performing a predetermined action on an object around them using a predetermined tool, A process for acquiring positional information of the predetermined tool when the predetermined tool performs the predetermined action on the object, Based on the position information of the person's hand, or the position information of the predetermined tool, obtained, the process calculates the posture information of the person's torso, and based on the calculated posture information of the torso, generates an animation of the torso using a blending process. Based on the calculated hand position information and the torso posture information, the process involves calculating the arm posture information using inverse kinematics calculations so as to connect the torso and hand of the person with the arm, and generating an animation of the arm based on the calculated arm posture information. It is a program that causes a computer to execute something. [Effects of the Invention]

[0008] The primary objective of this disclosure is to provide an animation generation system, an animation generation method, and a program that can easily generate animations of a person's natural movements when they perform actions on an object using a tool. [Brief explanation of the drawing]

[0009] [Figure 1] This block diagram shows a schematic system configuration of the animation generation system according to this embodiment. [Figure 2] This figure shows an example of blending parameters and torso posture parameters. [Figure 3] This figure shows a method for generating animations of the upper and lower body of the torso. [Figure 4] This figure shows another example of blending parameters and trunk posture parameters. [Figure 5] This figure shows a method for generating animations of the upper and lower body of the torso. [Figure 6] This figure shows another example of blending parameters and trunk posture parameters. [Modes for carrying out the invention]

[0010] This embodiment will be described below with reference to the drawings. Figure 1 is a block diagram showing a schematic system configuration of the animation generation system according to this embodiment. The animation generation system 1 according to this embodiment generates an animation of a person when, for example, a person uses a predetermined tool to perform a predetermined action, such as a processing action, on a target object around them in a three-dimensional space. The predetermined tool is, for example, a hammer, drill, power tool, rag, bat, etc.

[0011] Incidentally, when attempting to generate animations of a person performing the actions described above, adjustments to the blending conditions are made to adjust the relative positional relationship between the person and the tool, but this adjustment can be time-consuming. In contrast, conventional animation generation systems can reduce the effort required to adjust the blending conditions by using inverse kinematics calculations to make the person's movements follow the tool's movements. However, because the person's movements follow the tool's movements, the person may appear to be pulled by the tool, disrupting the person's natural posture and potentially making the movements unnatural.

[0012] In contrast, the animation generation system 1 according to this embodiment performs inverse kinematics calculation only on a person's arm (forearm and lower arm), and performs blending processing according to the blending parameters described later on the other trunk parts. Thereby, it is possible to generate an animation of a natural movement of a person while minimizing the labor of adjusting the blending conditions for adjusting the relative positional relationship between the person and the tool. Further, in the blending process, the position information of the hand holding a predetermined tool or the position information of the predetermined tool is used as the blending parameter, and the value of the posture parameter indicating the posture of the trunk part is determined according to the value of the blending parameter. Thereby, with a simple process, it is possible to link the posture of the trunk part to the movement of the tool while maintaining the natural posture of the trunk part, and to realize a more natural movement of a person.

[0013] The animation generation system 1 according to this embodiment includes a position information acquisition unit 2 that acquires the position information of a predetermined tool, a blending processing unit 3 that generates an animation of a person's trunk part using blending processing, and an inverse kinematics calculation processing unit 4 that generates an animation of a person's arm part using inverse kinematics calculation processing.

[0014] Note that the animation generation system 1 has a hardware configuration of a normal computer including, for example, a processor such as a CPU (Central Processing Unit) or a GPU (Graphics Processing Unit), an internal memory such as a RAM (Random Access Memory) or a ROM (Read Only Memory), a storage device such as a HDD (Hard Disk Drive) or an SDD (Solid State Drive), an input / output I / F for connecting peripheral devices such as a display, and a communication I / F for communicating with devices outside the apparatus.

[0015] The position information acquisition unit 2 acquires the position information (3D position coordinates, trajectory, etc.) of a predetermined tool when the predetermined tool performs a predetermined operation on an object. The position information of the predetermined tool may be input to the position information acquisition unit 2 using, for example, an input device. The position information of the predetermined tool may be set based on, for example, the position of the processing target (processing target position). The predetermined tool may perform, for example, a swinging motion such as a motion of swinging a hammer or a broom based on a reference point, or a translational motion such as a motion of pushing a drill.

[0016] The blending processing unit 3 calculates the position information of the human hand holding the predetermined tool based on the position information of the predetermined tool acquired by the position information acquisition unit 2. For example, the relative positional relationship between the human hand and the predetermined tool held by the hand may be set in the blending processing unit 3 in advance. The blending processing unit 3 may calculate the position information of the human hand holding the predetermined tool based on the position information of the predetermined tool acquired by the position information acquisition unit 2 and the set relative positional relationship.

[0017] The blending processing unit 3 can generate animations of the predetermined tool and the hand using blending processing based on the position information of the predetermined tool acquired by the position information acquisition unit 2 and the calculated position information of the human hand. Further, the blending processing unit 3 generates an animation of the torso using blending processing. First, the blending processing unit 3 calculates the posture information of the human torso based on the calculated position information of the hand.

[0018] In this blending processing, for example, as described above, the position information of the human hand is used as a blending parameter, and the value of the posture parameter indicating the posture of the torso is determined according to the value of the blending parameter. In this way, with a simple process that simplifies the human posture, while maintaining the natural posture of the torso, the posture of the torso can be linked to the movement of the tool, and a more natural movement of the human can be realized.

[0019] Figure 2 shows an example of blending parameters and torso posture parameters. For example, if the tool is a drill and a person pushes the tool toward the processing target, the blending parameter Z may be the relative height between the midpoint S of the wrist centers of both hands and the midpoint of the contact points of the soles of both feet, as shown in the upper part of Figure 2. Alternatively, the forward flexion and backward flexion angles of the upper body of the torso may be used as posture parameters α, and the standing and squatting amounts of the lower body of the torso may be used as posture parameters β.

[0020] The relationship between the blend parameter Z and the torso posture parameters α and β may be pre-set as table information, for example, as shown in the lower part of Figure 2, so that the values ​​of the torso posture parameters α and β are determined according to the value of the blend parameter Z. Alternatively, the relationship between the blend parameter Z and the torso posture parameters α and β may be set as a predetermined function. Furthermore, the relationship between the blend parameter Z and the posture parameters α and β may be taught to a machine learning model such as a neural network, using the blend parameter Z as the input value and the posture parameters α and β as the output values. For example, the blend processing unit 3 calculates the values ​​of the posture parameters α and β, which are the posture information of the human torso, based on the value of the blend parameter Z, which is the position information of the hand, the table information, the predetermined function, or the machine learning model.

[0021] Next, the blending unit 3 generates an animation of the upper body of the torso using blending based on the value of the posture parameter α, which is the calculated posture information of the torso, as described later. Similarly, the blending unit 3 generates an animation of the lower body of the torso using blending based on the value of the posture parameter β, which is the calculated posture information of the torso, as described later. The blending unit 3 generates an animation of the torso by connecting the generated upper and lower body animations.

[0022] Here, we will explain a specific example of how to generate animation for the upper body of the torso. Figure 3 is a diagram illustrating how to generate animation for the upper and lower body of the torso. For example, as shown in the upper part of Figure 3, the blend processing unit 3 generates animation for the upper body of the torso using blend processing based on the posture parameters α of the upper body of the torso, which are the forward flexion and backward flexion angles.

[0023] The blending unit 3 has two base animations set: animation (a) for the upper body of the torso at its maximum forward flexion angle, and animation (b) for the upper body of the torso at its maximum backward flexion angle. The blending unit 3 generates animation (c) for the upper body of the torso so that the posture is an intermediate ratio between (a) and (b) above, depending on the forward flexion and backward flexion angles α of the upper body of the torso.

[0024] Next, we will explain a specific example of how to generate animation for the lower half of the torso. For example, as shown in the lower part of Figure 3, the blend processing unit 3 generates animation for the lower half of the torso using blend processing based on the standing and crouching amounts, which are the posture parameters β of the lower half of the torso.

[0025] The blending unit 3 has two animations set as base animations: (d) for the lower body of the torso at its maximum standing position, and (e) for the lower body of the torso at its maximum crouching position. The blending unit 3 generates an animation (f) for the lower body of the torso so that the posture is at an intermediate ratio between (d) and (e) above, depending on the standing and crouching position β of the lower body of the torso.

[0026] The blending parameters and torso posture parameters described above are examples only and are not limited thereto. Figure 4 shows another example of blending parameters and torso posture parameters. For example, if the predetermined tool is a bat or an axe, and a person performs a motion of swinging the predetermined tool toward a target point, the horizontal relative distance between the contact point P of the predetermined tool and the target point T when the predetermined tool contacts the target point T (such as a processing point) may be used as the blending parameter L, as shown in Figure 4. Alternatively, the amount of twisting of the upper body of the torso may be used as the posture parameter γ, and the amount of offset of the center of gravity in the anterior-posterior direction of the lower body of the torso (the direction in which the torso approaches or moves away from the target point T) may be used as the posture parameter δ.

[0027] The relationship between the blend parameter L and the torso posture parameters γ and δ may be pre-set as table information, a predetermined function, or a machine learning model, such that the values ​​of the torso posture parameters γ and δ are determined according to the value of the blend parameter L.

[0028] The value of the blending parameter L, which is the position information of the specified tool, may be a value input to the position information acquisition unit 2, or a value calculated based on the position information of the specified tool acquired by the position information acquisition unit 2, or it may be calculated based on the position information of a person's hand.

[0029] The blending processing unit 3 calculates the values ​​of posture parameters γ and δ, which are posture information of the human torso, based on the value of the blending parameter L, which is the position information of a predetermined tool, and a predetermined function, for example, as shown below. Note that L0 is the reference distance between the human torso and the movement target point T, and may be pre-set in the blending processing unit 3. γ = 2 × Atan(L) δ = (1 / 10) × (L / L0)

[0030] The blending processing unit 3 may also calculate the values ​​of posture parameters γ and δ, which are posture information of the human torso, based on the value of the blending parameter L, which is the position information of a predetermined tool, and table information or a machine learning model.

[0031] The blending processing unit 3 generates an animation of the upper body of the torso using blending processing based on the value of the posture parameter γ, which is the posture information of the torso calculated as described above. Figure 5 is a diagram showing the method for generating animations of the upper and lower body of the torso. For example, as shown in the upper part of Figure 5, the blending processing unit 3 generates an animation of the upper body of the torso using blending processing based on the amount of twist of the upper body, which is the posture parameter γ of the upper body of the torso.

[0032] The blending unit 3 has two base animations set: an animation (g) for when the upper body of the torso is twisted to the maximum right, and an animation (h) for when the upper body of the torso is twisted to the maximum left. The blending unit 3 generates an animation (i) for the upper body of the torso so that the posture is at an intermediate ratio between (g) and (h) above, depending on the amount of twist γ of the upper body of the torso.

[0033] Similarly, as shown in the lower part of Figure 5, for example, the blending processing unit 3 generates animation of the lower body of the torso using blending processing based on the offset amount of the center of gravity in the anterior-posterior direction of the torso, which is the posture parameter δ of the lower body of the torso. The blending processing unit 3 has two animations set as base animations: (j) for when the lower body of the torso is tilted backward to the maximum extent (when the offset amount of the center of gravity is at its maximum backward extent) and (k) for when the lower body of the torso is tilted forward to the maximum extent (when the offset amount of the center of gravity is at its maximum forward extent). The blending processing unit 3 generates animation (l) of the lower body of the torso so that the posture is at an intermediate ratio between (j) and (k) in accordance with the offset amount δ of the center of gravity in the anterior-posterior direction.

[0034] Furthermore, Figure 6 shows another example of blending parameters and torso posture parameters. For example, if the given tool is a rag, and a person moves the given tool on a wide plane in all directions (up, down, left, and right) while their footing is fixed, the horizontal position and height position of the contact point R of the given tool may be used as blending parameters X and Z, respectively, as shown in the upper part of Figure 6. Alternatively, the amount of cartwheel of the upper body of the torso may be used as the posture parameter ε, and the amount of standing and squatting of the lower body of the torso may be used as the posture parameter β.

[0035] The relationship between the blend parameters X and Z and the torso posture parameters ε and β may be pre-set as table information, a predetermined function, or a machine learning model, so that the values ​​of the torso posture parameters ε and β are determined according to the values ​​of the blend parameters X and Z.

[0036] The blending parameters X and Z, which are the position information of the specified tools mentioned above, may be values ​​input to the position information acquisition unit 2, or values ​​calculated based on the position information of the specified tools acquired by the position information acquisition unit 2, or they may be calculated based on the position information of a person's hand.

[0037] The blending processing unit 3 calculates the values ​​of posture parameters ε and β, which are posture information of the human torso, based on the values ​​of blending parameters X and Z, which are position information of a predetermined tool, and a predetermined function, for example, as shown below. In the predetermined function below, k, m, and n are coefficients that have been set in advance in the blending processing unit 3, etc. ε = k × X β = m × Zn × X

[0038] The blending processing unit 3 may also calculate the values ​​of posture parameters ε and β, which are posture information of the human torso, based on the values ​​of blending parameters X and Z, which are position information of a predetermined tool, and table information or a machine learning model.

[0039] The blending processing unit 3 generates an animation of the upper body of the torso using blending processing, based on the value of the posture parameter ε, which is the posture information of the torso calculated as described above. For example, as shown in the lower part of Figure 6, the blending processing unit 3 generates an animation of the upper body of the torso using blending processing, based on the amount of torso somersault, which is the posture parameter ε of the upper body of the torso.

[0040] The blending unit 3 has two base animations set: an animation (m) for when the upper body of the torso is turned to the maximum right, and an animation (n) for when the upper body of the torso is turned to the maximum left. The blending unit 3 generates an animation (o) for the upper body of the torso so that the posture is an intermediate ratio between (m) and (n) above, depending on the amount of torsional rotation ε of the upper body of the torso.

[0041] Similarly, as shown in the lower part of Figure 3, for example, the blend processing unit 3 generates animation of the lower body of the torso using blend processing based on the standing and squatting amounts β of the torso, which are posture parameters β of the lower body of the torso.

[0042] The blending unit 3 has two animations set as base animations: (d) for the lower body of the torso at its maximum standing position, and (e) for the lower body of the torso at its maximum crouching position. The blending unit 3 generates an animation (f) for the lower body of the torso so that the posture is at an intermediate ratio between (d) and (e) above, depending on the standing and crouching position β of the lower body of the torso.

[0043] As described above, the relationship between the blending parameters and the torso posture parameters differs depending on the predetermined tool used by the person. Therefore, the table information, function, or machine learning model that shows the relationship between the blending parameters and the torso posture parameters may be changed depending on the predetermined tool used by the person. For example, the blending processing unit 3 may have table information, function, or machine learning model set up in association with each predetermined tool. The blending processing unit 3 may calculate the values ​​of the posture parameters based on the values ​​of the blending parameters and the table information, function, or machine learning model corresponding to the predetermined tool used by the person. This makes it possible to calculate the natural posture of the torso with simple processing by switching the table information, function, or machine learning model according to various predetermined tools.

[0044] The inverse kinematics calculation unit 4 calculates the posture information of the arm using inverse kinematics calculation, based on the hand position information and torso posture information calculated by the blending unit 3, so as to connect the shoulder and hand parts of the human torso with the arm. This inverse kinematics calculation may be, for example, an inverse kinematic calculation based on the Jacobian matrix and the Singularity-Robust Inverse (SR-Inverse) inverse matrix. By performing this calculation, a more accurate animation of the arm can be generated.

[0045] The inverse kinematics calculation processing unit 4 generates an animation of the arm based on the arm posture information calculated as described above.

[0046] Next, an example of an animation generation method according to this embodiment will be described. The position information acquisition unit 2 acquires the position information of a predetermined tool when the predetermined tool performs a predetermined action on an object (step S101).

[0047] The blending processing unit 3 calculates the position information of the person's hand holding the predetermined tool based on the position information of the predetermined tool acquired by the position information acquisition unit 2 (step S102). Based on the position information of the predetermined tool and the position information of the person's hand, the blending processing unit 3 generates an animation of the predetermined tool and the person's hand using blending (step S103).

[0048] The blending processing unit 3 calculates posture information of the human torso based on the calculated hand position information (step S104). Based on the calculated posture information of the torso, the blending processing unit 3 generates an animation of the torso using blending (step S105).

[0049] The inverse kinematics calculation processing unit 4 calculates the posture information of the arm using inverse kinematics calculation processing, based on the hand position information and torso posture information calculated by the blend processing unit 3, so as to connect the torso and hand of the person with the arm (step S106). The inverse kinematics calculation processing unit 4 generates an animation of the arm based on the calculated arm posture information (step S107).

[0050] While several embodiments of this disclosure have been described, these embodiments are presented as examples only and are not intended to limit the scope of the invention. These novel embodiments can be carried out in a variety of other forms, and various omissions, substitutions, and modifications can be made without departing from the spirit of the invention. These embodiments and their variations are included in the scope and spirit of the invention, as well as in the claims and their equivalents.

[0051] This disclosure can also be implemented, for example, by having a processor execute a computer program to perform the processing of each part of the animation generation system 1 described above. The program can be stored and supplied to the computer using various types of non-transitory computer-readable medium. Non-transitory computer-readable mediums include various types of tangible storage mediums. Examples of non-transitory computer-readable mediums include magnetic recording media (e.g., flexible disks, magnetic tapes, hard disk drives), magneto-optical recording media (e.g., magneto-optical disks), CD-ROMs (Read Only Memory), CD-Rs, CD-R / Ws, and semiconductor memory (e.g., mask ROMs, PROMs (Programmable ROMs), EPROMs (Erasable PROMs), flash ROMs, RAMs (random access memory)).

[0052] Programs may be supplied to a computer by various types of transient computer-readable medium. Examples of transient computer-readable medium include electrical signals, optical signals, and electromagnetic waves. Transitory computer-readable medium can be supplied to a computer via wired communication channels such as electric wires and optical fibers, or via wireless communication channels. [Explanation of symbols]

[0053] 1 Animation generation system, 2 Position information acquisition unit, 3 Blending processing unit, 4 Inverse kinematics calculation processing unit

Claims

1. An animation generation system that generates an animation of a person performing a predetermined action on an object around them using a predetermined tool, A position information acquisition unit acquires position information of the predetermined tool when the predetermined tool performs the predetermined operation on the object, A blending processing unit calculates posture information of the person's torso based on the position information of a predetermined tool acquired by the position information acquisition unit, or based on the position information of the predetermined tool, and generates an animation of the torso using blending processing based on the calculated posture information of the torso. An inverse kinematics processing unit calculates the position information of the arm and the position information of the torso based on the position information of the hand calculated by the blend processing unit and the posture information of the torso, using inverse kinematics calculation processing to connect the torso and hand of the person with the arm, and generates an animation of the arm based on the calculated posture information of the arm. An animation generation system equipped with the following features.

2. An animation generation system according to claim 1, The position information of the person's hand or the position information of the predetermined tool is used as a blending parameter, and table information, a predetermined function, or a machine learning model is pre-set to show the relationship between the blending parameter and a posture parameter indicating the posture of the person's torso. The blending processing unit calculates the value of the posture parameter, which is posture information of the torso, based on the value of the blending parameter, the table information, a predetermined function, or a machine learning model, and generates an animation of the torso using the blending process based on the calculated value of the posture parameter, an animation generation system.

3. The animation generation system according to claim 2, The table information, predetermined function, or machine learning model is set up in correspondence with each of the aforementioned predetermined tools. The blending processing unit is an animation generation system that calculates the value of the posture parameter based on the value of the blending parameter, the table information corresponding to a predetermined tool used by the person, a predetermined function, or a machine learning model.

4. An animation generation method for generating an animation of a person performing a predetermined action on an object around them using a predetermined tool, The steps include: acquiring positional information of the predetermined tool when the predetermined tool performs the predetermined operation on the object; The steps include: calculating posture information of the person's torso based on the position information of the person's hand based on the position information of the predetermined tool obtained, or based on the position information of the predetermined tool, and generating an animation of the torso using blend processing based on the calculated posture information of the torso; Based on the calculated hand position information and the torso posture information, the steps include: calculating the arm posture information using inverse kinematics processing so as to connect the torso and hand of the person with the arm; and generating an animation of the arm based on the calculated arm posture information; An animation generation method that includes this.

5. A program that generates an animation of a person performing a predetermined action on an object around them using a predetermined tool, A process for acquiring positional information of the predetermined tool when the predetermined tool performs the predetermined action on the object, Based on the position information of the person's hand, or the position information of the predetermined tool, obtained, the process calculates the posture information of the person's torso, and based on the calculated posture information of the torso, generates an animation of the torso using a blending process. Based on the calculated hand position information and the torso posture information, the process involves calculating the arm posture information using inverse kinematics calculations so as to connect the torso and hand of the person with the arm, and generating an animation of the arm based on the calculated arm posture information. A program that causes a computer to execute something.

Images