JP2003337956A - Apparatus and method for computer graphics animation - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a computer graphics animation apparatus which can realize a variety of animations with a small proper amount of data and easily add and delete shapes and hierarchical structures. <P>SOLUTION: An animation control part 1 specifies shape data, hierarchical structure data, a group table, and state information. A character state calculation part 3 obtains the specified shape data from a shape data storing unit 5, the hierarchical structure data from a hierarchical structure storing unit 4, the group table from a table storing unit 2, motion data shown in the specified state information from a motion data storing unit 6, and specifies, from the obtained motion data, motion data identified by each group number. In accordance with the obtained hierarchical structure data, the character state calculating unit 3 corrects the shape data by using the specified motion data. A three-dimensional rendering unit 7 renders the corrected shape data to generate an image, and a display unit 9 displays the generated image. <P>COPYRIGHT: (C)2004,JPO

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to the field of three-dimensional computer graphics animation, and more particularly, to a complex object having a hierarchical structure such as a character.
The present invention relates to a three-dimensional computer graphics animation device.

[0002]

2. Description of the Related Art Conventionally, characters such as animals and humans are
When an animation moving image is generated by three-dimensional computer graphics (hereinafter, 3D CG), a key frame method of joint angles based on forward kinematics and a skeleton method is used as a key frame method in a broad sense (for example, see Non-Patent Document 1). .). These methods will be briefly described with reference to FIGS.

A skeleton structure as shown in FIG. 3 is defined for a character shape as shown in FIG. 2, and this is represented by a link having a hierarchical structure as shown in FIG. The motion of the character generates a motion of the whole body by the direction and movement of the root of the skeletal structure of FIG. 3, and the local motion of the body is
It is generated by transformation in the local coordinate system in the joint part of the skeletal structure. By moving the surface shape in accordance with the movement of the skeleton structure, the character moves.

As the transformation of the local coordinate system at the joint position,
Generally, a 4 × 4 matrix is used to represent rotation and parallel movement at a joint, but normally, a joint has only a rotation angle that does not have a degree of freedom for parallel movement. This rotation angle is called a joint angle. Perform these conversions in sequence from the root according to the link of the skeletal structure. Determine the state of each part of the body and express it as a movement in the coordinate system of the root (generally in the world coordinate system) The method of determining the position and direction of the whole body according to this is called forward kinematics. By repeating these at each time (for each display frame rate), the animation is expressed as a temporal movement. Therefore, in Non-Patent Document 1 described above, the joint angle data is required in accordance with the frame rate, but the joint angle key frame method has only the joint angle data (key data) at the sample time as data. The joint angles at other times are calculated by key data interpolation calculation.

The following is a summary of what has been described about the conventional motion data generation method and animation control (see, for example, Non-Patent Document 2) using such a method. Become.

(1) Design a CG character, define a skeleton structure (hierarchical structure), and create shape data. (2) The physical state (pose) of the CG character at the key time is generated (the facial expression is also attached). (3) The shape of the CG character at each time is determined by the key frame method, rendered under appropriate lighting and camera data, and continuously displayed to generate a 3D CG animation.

Therefore, as shown in FIG. 5, the movement of the whole body including the facial expressions is generated as a series of 3D CG animations as at the time of data generation. The same document also describes a method using morphing as a method for generating facial expressions including lip movements (called facial animation). As another method of expressing facial expressions and lip movements, there is a method of creating facial expressions by creating a number of face textures and exchanging them for texture mapping.

[0008]

[Non-patent Document 1] "3D CG", "Introduction to Advanced Technology, Series of Television Society", edited by Masayuki Nakajima, Ohmsha Publishing Co., Ltd., 1994, p143-161.

[0009]

[Non-Patent Document 2] George Maestri, Translated by Koichi Matsuda, et al. "Digital Character Animation" Published by Prentice Hall, 1999.

[0010]

Generally, in the case of 3D CG character animation, motion transitions are repeated based on the state transition diagram shown in FIG. 10A in order to handle interactive applications such as games. , A method of generating a series of actions of a CG character is often used. When based on the above-described conventional method, the basic state and the operations A, ..., And E of FIG. 10A are created by the 3D CG animation generation method as described in the second conventional example. Therefore, FIG.
As described above, the six types of operation states including the basic state are changed based on the state transition. In order to generate behaviors involving more various movements, motion data for that amount is required, and the amount of data increases.

In addition, for example, when there is a "waving hand" motion and a "sitting on a chair" motion, the "waving hand and sitting on a chair" motion is newly regarded as one motion. Unless created and incorporated into the state transition, it will not appear in the action of the CG character. This is more serious when the facial expression is included, and if the facial expression is different even if the physical movement is the same, it is necessary to create the same physical movement for each facial expression. Of course, the labor for creating the motion data can be saved because the physical motion can be reused, but the amount of data cannot be reduced. For example, in the case of a device with a rare memory environment, such an increase in the amount of data is fatal. Therefore, in order to reduce the amount of data, there is no choice but to reduce the types of motions, and as a result, the CG character's various actions are varied. There is no sense of sex, and the user will soon get bored.

Next, in actual human behavior, it often happens that objects are held, especially tools. However, in such a real world, if an attempt is made to use a conventional CG character animation method to avoid the often-caused bluntness, the following problems occur.

As a first problem, the state of holding an object is
Since the skeletal structure is different from the state without it, the shape data without the state and the shape data with the state are required, and the CG character data is required for each kind of object. The second problem is that since the skeletal structures are different, there are two cases, one with a physical object and one without, even if the body movement data are the same.
Some kind of action is required. Therefore, even with the same behavior, the shape data and the motion data have almost twice the data amount. In such a situation, if it is attempted to realize the variety of motions including facial motions as described above, the amount of data will explosively increase.

The present invention has been made in view of the above problems, and eliminates the redundancy of data that causes the above problems as much as possible, and realizes the animation of various CG characters with an appropriate data amount smaller than the conventional one. The object is to provide a computer graphics / animation device capable of performing the above.

Another object of the present invention is to realize a computer graphics / animation device which can easily add or delete a shape such as having an accessory or a thing or a hierarchical structure.

[0016]

In order to solve the above problems, a computer graphics / animation apparatus of the present invention is a computer graphics / animation apparatus for displaying an image of a character, wherein A hierarchical structure storing unit that stores hierarchical structure data that is data indicating the hierarchical structure of the part of
For a plurality of parts in which a group of the plurality of parts is associated as one part, a part table storage unit that stores a part table indicating a correspondence relationship between the parts and the parts, and a plurality of part tables registered in the part table. Motion data storage means for storing motion data in which a motion is associated with each part, and shape data indicating a temporal change in the shape of the character based on the motion data, the part table, and the hierarchical structure data. And character state calculation means for calculating.

The computer graphics animation apparatus according to the present invention further comprises animation control means for inputting event information relating to the movement of the three-dimensional character, and the animation control means is adapted to input the event information. The motion data, the part table, and the hierarchical structure data are specified according to the type of the character data, and the character state calculation unit receives shape data indicating a temporal change in the shape of the character according to the input from the animation control unit. It is characterized by being calculated.

The present invention can be realized not only as a computer graphics / animation apparatus as described above, but also as a computer graphics / animation method having means included in the animation apparatus as steps.

Further, it goes without saying that the animation method can be realized as a program realized by a computer or the like, or the program can be distributed via a recording medium such as a CD-ROM or a transmission medium such as a communication network.

[0020]

BEST MODE FOR CARRYING OUT THE INVENTION A computer graphics animation apparatus according to a first embodiment of the present invention will be described below with reference to the drawings.

The computer graphics animation apparatus of the present invention divides a three-dimensional CG character into parts and controls the operation of the entire CG character by controlling the operation of each part.

FIG. 1 shows the configuration of a computer graphics / animation apparatus according to the first embodiment of the present invention, in which event information is input and various data for controlling animation according to the event information is stored. An animation control unit 1 for notifying the character state calculation unit 3, a part table storage unit 2 for managing and storing a part table, and a character state for changing the shape data of a character by acquiring data from the shape data storage unit 5 and the like. A calculation unit 3, a hierarchical structure storage unit 4 in which hierarchical structure data defining the skeletal structure of a CG character is managed and stored as an address, and a shape data storage unit 5 in which shape data of CG characters and parts are managed and stored by an address. And motion data for animation of CG characters and parts. C but performs the operation data storage section 6 that is managed stored at addresses and operation ID, and image generation processing
3D drawing unit 7 for drawing G animation and CG
A texture data storage unit 8 that manages and stores the texture data necessary for drawing by an address or an ID, and a display unit 9 that displays a CG animation.

The computer graphics / animation apparatus according to the embodiment of the present invention configured as described above will be described in detail below. First, the description of the stored data will be given.

In the hierarchical structure storage unit 4, hierarchical structure data defining a skeletal structure of a CG character (CG may be omitted in the following description and may be simply referred to as a character) to be subjected to CG character animation is managed by an address. It has been saved. For example, a skeletal structure like a human has a skeletal structure as shown in FIG.
It can be defined as hierarchical structure data having a link structure as shown in. This hierarchical structure data may be called a scene graph in the field of CG. When a character having a skeleton structure as shown in FIG. 3 has a stick in his hand as shown in FIG. 7, it is necessary to define the stick as a part of the skeleton structure.

The hierarchical structure data corresponding to FIG. 7 is shown in FIG. 8. Since the structure other than the bar is exactly the same as the skeleton structure shown in FIG. 3, the bar can be defined as a part of the skeleton structure of FIG. When there is a part to be added in this way, the hierarchical structure of the part (bar) has a connection destination (right hand in the case of the bar shown in FIG. 7) in the parent hierarchical structure of the root of the hierarchical structure data. It can be defined as data. Therefore, the hierarchical structure data stored here includes (1) parent hierarchical structure data, (2) component hierarchical structure data, (3) data defining parent-child relationship, parent hierarchical structure data address, and component hierarchy. There are three types of hierarchical structure data consisting of structure data addresses.

Then, the hierarchical structure storage unit 4 saves the identifier indicating the type of the hierarchical structure data and the hierarchical structure data. Therefore, it is not necessary to separately have the hierarchical structure data of FIGS. 4 and 8. When the hierarchical structure data is accessed, the type is first discriminated by the identifier, and in the case of the above type (1) or (2), the hierarchical structure data is directly obtained, but in the case of the above type (3) Can obtain an address indicating the parent-child relationship and the whereabouts of the entity, then obtain the hierarchical structure data of the entity, and construct the entire hierarchical structure data based on the description of the parent-child relationship.

The shape data storage unit 5 manages and stores the shape data of CG characters and parts by address.
The shape data of the CG character corresponds to the above-mentioned skeleton structure, and is data corresponding to human skin or clothes covering the skeleton as shown in FIG. 2, such as the head, upper limbs, trunk, and lower limbs. Of the shape parts, and they are hierarchized corresponding to the above hierarchical structure data.

Therefore, the shape data includes an address of the corresponding hierarchical structure data in the hierarchical structure storage unit 4 and an identification tag of which hierarchical data each shape part corresponds to. The same applies to the shape data of parts.

Each shape part is usually represented by a surface model, which is called a surface model, that approximates the surface of an object by polygons, and the vertex coordinates in the three-dimensional space and the normal vector component (light source) at the same vertex are represented. (Required for brightness calculation), Texture coordinates (Required for texture mapping)
Of the indexed point sequence data and the phase data (for example, the vertex index is 1,
If written in the order of 2 and 3, it represents a triangle having points 1, 2, and 3 as vertices), and the reflectance (diffuse reflectance, specular reflectance) of each surface, ambient light intensity, and object color. Attribute data such as.

When the clothing or face of the CG character is represented by texture mapping, the address of the texture to be used in the texture data storage unit 8 is explicitly indicated in the corresponding part of the shape data of the CG character. Has been done.

In the motion data storage unit 6, motion data for animation of CG characters and parts is managed and stored by address and motion ID. In the present embodiment, the motion data of an object having a hierarchical structure will be described by using a CG character as an example, but the same can be applied to objects having other hierarchical structures.

In the case of a CG character, motions can be divided into physical motions and facial expression motions. Therefore, in this embodiment, there are two types of motion data: physical motion data and facial expression motion data.

The body movement data is position data and direction data (shown in FIG. 2) at the root of hierarchical structure data (P at the representative point of the body shown in FIG. 2), as is done in normal CG character animation. The posture vector) represents the position and posture of the entire body in each three-dimensional space at each time. Although the position data of the route (P) can be directly expressed as position data, it is generally expressed by a position vector at an initial time and a movement amount vector indicating a movement amount at each time. The direction is specified by the rotation transformation of the local coordinate system defined in the route (P), and specifically, this is the rotation angle amount around the three coordinate axes of the three-dimensional space or the component of the vector representing the central axis of rotation. And the angle of rotation around the vector. The motion data of this route is the motion part identifier m0 of the hierarchical structure data of FIGS.
Identified by.

The movement of each part of the body (upper and lower limbs, trunk, etc.) is expressed by time series data of the amount of rotation angle around the coordinate axis of the local coordinate system defined by each joint in the skeletal structure. Since the joint of the skeletal structure corresponds to the link part of the hierarchical structure data, the motion part identifier (motion ID) m0, m1, m2, ..., M shown in FIG.
27 is given. Therefore, in the conventional animation technique, a set of these multivariate time series data of m0, m1, ..., M27 becomes the body motion data. .., m27 are converted by the conversion system in the local coordinate system at these root positions and joints, and the CG character shape data is converted to the position and orientation of the CG character at each time and the CG character's position. A CG animation can be realized by generating a pose of the body, performing a three-dimensional drawing process, and continuously performing the process over time.

As shown in FIG. 5, when the key frame animation technique is used, time series data (f0, f5, f10 .. ), The time (f1
.. to f4, f6 to f9 ..,) are performed by interpolation calculation, the body motion data is the time series data of the parallel movement amount and the angle amount which are discontinuous in time.

The facial expression can be decomposed into movements of the eyes, cheeks, mouth, etc., as can be seen by an actual person. For example, in FIG.
In the case of the hierarchical structure of
It can be broken down into cheek movements, lip movements, and eyebrow movements. These are actions that depend on the face model.

A facial animation technique that is usually used is used to generate facial expressions.
There are a method of transforming the shape of the face and a method of pasting the texture of the face.

When transforming the shape of this face,
Of the face shape data, the time series data of the movement amount of the vertex coordinates corresponding to the end points such as the eyebrows, the eyes, and the mouth that generate the facial expression is the facial expression motion data. These movement amounts can also be calculated by simulation calculation based on the facial muscle model. If the vertices to be transformed span multiple transformation systems, give weights to each transformation to the vertices,
An envelope method is also used in which a plurality of vertices whose vertices are once transformed are calculated in each transformation system, and the vertices are transformed into averaged coordinates in consideration of weighting.

FIG. 6 is an explanatory diagram for the case where the facial textures are replaced to generate facial expressions. When the facial expressions are generated by replacing the textures, a plurality of laughing and crying facial expressions and intermediate textures are archived and IDs are internally managed. When using such a texture, select from the archived textures necessary for facial expression generation at each time by ID,
A facial expression is generated by performing texture mapping on the face. Therefore, the time-series parameters (frame number and time) in the case of physical motion data correspond to the management ID. These archived texture data are facial expression motion data for generating facial expressions.

Since there is motion data corresponding to the hierarchical structure in both the above-described method of deforming the shape of the face and the method of re-pasting the texture of the face,
It is identified by a motion identifier as in the case of physical motion. For example, in the case of FIG. 4, the identifiers of m28, ..., M34 are the motion identifiers of facial expressions.

In the present embodiment, the case where a facial expression is generated by the above-described archived texture mapping will be described. However, even when the facial animation is performed by the shape deformation or the envelope method, the physical movement is not performed. This can be dealt with by applying a similar method.

Further, in the present embodiment, the hierarchical structure data is provided separately from the shape data and the motion data described above, but the shape data description method and the motion data description method are devised. By doing so, it can be included in the shape data or the motion data. That is, since the shape data consists of parts, an identifier of the part and an identifier indicating the parent-child relationship using the identifier are inserted into the data, and the motion data similarly uses the operation identifier and the parent-child relationship using the identifier. It suffices to insert an identifier indicating For example, when the shape component identifiers 1 and 2 are in a parent-child relationship, the parent identifier 1 may be entered in the header part or attribute description part of the component data of the identifier 2.

In the part table storage unit 2, a part table is managed and stored by a part table number or address. As shown in FIG. 9, this part table is a table that holds a part ID and a motion ID set as data. Part ID in the present invention
Is defined independently of the above-described hierarchical link structure, and defines a set of a plurality of motion IDs (motion ID set) for one part ID.

For example, the part table 0 shown in FIG.
Part 0 to Part 4 are defined for the hierarchical structure of the CG character shown in FIG. Part 0 is the root (m0) and the base of the spine (m1), Part 1 is the left crotch (m2), right ankle (m7), etc., Part 2 is the left collar joint (m
8), upper cervical vertebrae (m27), etc.
8) and left and right cheeks (m29, m30), and Part 4 includes left and right eyes (m31, m32) and left and right eyebrows (m33, m34).

Further, in the part table 1 shown in FIG. 9, parts 0 to 5 are defined for the hierarchical structure shown in FIG. 8, and part 0 is a root (m0) and a spine root (m).
1), Part 1 is the upper and lower cervical vertebrae (m26, m27), Part 2 is the left crotch (m2) and left collar joint (m8) on the left side of the body.
Etc., Part 3 is the right crotch (m5) and right collar joint (m17) of the right part of the body, Part 4 is the lip of the face (m28) and left and right cheeks (m29, m30), and Part 5 is the left and right eyes (m31, m3
2) and the left and right eyebrows (m33, m34) are included.

Further, in the part table 2 shown in FIG. 9, parts 0 to 5 are added to the hierarchical structure shown in FIG.
Part 0 defines the root (m0) and the root of the spine (m
1), part 1 is left crotch (m2), right ankle (m7), etc., part 2 is left collar joint (m8), upper cervical vertebra (m27), etc., part 3 is lip (m28) and left and right cheeks (m29, m30) ), Part 4 includes left and right eyes (m31, m32) and left and right eyebrows (m33,
m34), part 5 is shown to include a rod (m35).

Further, in FIGS. 7 and 8, the division of each part of the CG character in the part table 2 of FIG. 9 is shown by a dotted line. Part 0 includes the root (m0) and the root of the spine (m1), and Part 1 includes the left crotch (m2), left knee (m3), left ankle (m4), right crotch (m5), right knee (m).
6) and right ankle (m7) are included, and part 2 is left collar joint (m8), left shoulder (m9), left elbow (m10), left wrist (m11), left hand first (m12 to m16), right. Collar joint (m17), right shoulder (m18), right elbow (m19), right wrist (m20), first right hand (m21 to m25), lower cervical spine (m
26) and the upper cervical spine (m27) are included, and part 3 includes the lips (m28) and the left and right cheeks (m29, m30),
Part 4 includes left and right eyes (m31, m32) and left and right eyebrows (m
33, m34) and part 5 contains a bar (m35).

As described above, the computer graphics / animation apparatus according to the present invention divides two or more groups of a plurality of parts constituting a CG character into parts, and outputs motion data and the like to each of these parts. By giving it, the motion control of the entire CG character is performed. Therefore, unlike the conventional key frame animation control, it is not necessary to add data to all parts in order to express one motion of the CG animation, and the character state calculation unit 3 operates in a form independent of each part. By adding the data, it becomes possible to calculate the shape data indicating the temporal change of the shape of the CG character.

For example, in the case of displaying a CG character "Walk while laughing", in the conventional animation device, it is necessary to add data "Walk while laughing" to all the parts constituting the CG character at once. However, in the present invention, by setting the CG character as a set of parts including a plurality of parts, it is possible to add “crying” motion data and “walking” motion data to each part, and the shape of each part can be changed. It is possible to change. Therefore, it is not necessary to add new data to all the parts, the redundancy of the data is eliminated as much as possible, and various CG character animations can be realized with a small and appropriate data amount.

Furthermore, in CG animation,
Although various motions can be expressed by assigning data to all parts of the character, in the animation device according to the present invention, the motion is simply complicated by supplying data to all parts of the CG character. Instead, by adding data to each part, which is a set of constituent parts, a meaningful motion can be expressed as the CG character as a whole.

Further, by using the part table as described above, the motion data can be selected from the motion data different for each part ID. Incidentally, when there is only one part ID, it becomes equivalent to the conventional key frame animation control. The maximum number of part IDs is equal to the number of motion IDs.

The texture data storage unit 8 manages and stores the texture data required for CG drawing by address or ID. The texture data to be stored include the above-mentioned archived texture data for generating the facial expression motion, data used for clothes attached to the shape data, and data used as a background. Usually RG
A color image of a format such as B or YUV is stored in a compressed or uncompressed state. However, if it is compressed and stored, it needs to be decompressed before performing texture mapping.

Next, the animation image generation process in this embodiment will be described. In the animation control unit 1, the address of the shape data of the CG character, the address of the shape data of the accessory parts, the address of the hierarchical structure data of the CG character and the accessory parts, the address of the part table to be used, for the CG animation image generation in advance. It is assumed that body movement pattern data and facial expression pattern data, which will be described later, are notified or held. When there are a plurality of CG characters, the pattern data is held for each character. However, even in the case of a plurality of CG characters, if the transition is commonly caused, the pattern data may be the same for each of the physical movement and the facial expression.

Although the part table may be fixed, in the present embodiment, for example, event information (key input information, voice analysis result, contact or collision with another shape, etc., the content depends on the system). It is determined, but the present embodiment does not matter what it is) and a case of transitioning to a different part table will also be described. However, the transition of the part table is assumed to be held in the form of a scenario in advance. For example, when the information of the event P is input, the process for the event P is determined with reference to the scenario,
The transition of the part table occurs when the determined content is the transition of the part table. In addition, this scenario also includes decision information regarding transitions of physical movements (physical movement pattern data), transitions of facial expression movements (facial expression pattern data), and movements of accessory parts (pattern data of accessory parts), which will be described later. There is.

The body movement pattern data is shown in FIG.
In the finite state graph data as shown in, the relationship between actions that can be transferred from a certain action and actual action information (action ID,
Data type, address and frame number of each physical action,
The transition probability of each transition). For example, in FIG. 10A, it can be seen that it is possible to shift from the basic motion data representing the standard state to the motion A, the motion B, the motion C, the motion D, and the motion E. When the CG character is in the standard state, some predetermined event occurs or 1
When one action is completed, the action is selected from action A, action B, action C, action D, and action E by the selection process (random number process) based on the transition probability (also called transition probability) described in the actual action information. Then, the substance of the operation is acquired by the address. It should be noted that this transition is performed for each part ID related to the physical movement in the part table.

The facial expression pattern data is transition graph data of these facial expressions, and like the transition graph data of body movements data, a finite state graph and an actual facial expression that enable transition from one facial expression movement data to another facial expression movement data. It is information (facial expression ID, data type, address and frame number of each actual facial expression motion data, transition probability of each transition). For example, as shown in FIG. 10B, in this example, it is shown that facial expression 0 can be transferred to another facial expression 1, facial expression 2, and facial expression 3 as the basic facial expression. It is performed based on the information transfer probability. Since it is possible to specify an applicable shape such as a facial expression motion or a texture based on the data type of the real facial expression information, the facial expression motion in the case of the above-described shape deformation can also be applied. For example, the first digit of the data type is used for classification of facial expression or texture, and 2
Numbers above the first digit are used as shape identification numbers. As a result, both facial expression motion generation by texture and facial expression generation by shape deformation can be supported depending on the data type.

In the facial expression movement, especially lip movement (lip,
In the present embodiment, the motion ID set m28, m29, m of the part 3 shown in the part table 0 is used for cheek motion or the like.
(Corresponding to 30) and performing a synchronization operation (lip sync) with voice can be realized by the following other control method instead of the above expression pattern data. As the event information, it is assumed that the voice analysis result for the voice data is input, and as the voice analysis result, the sound intensity analysis result and the phoneme analysis result (explained as vowel analysis is performed, but consonant analysis is also performed. The same can be done if
The case where is obtained will be described.

First, as lip movement data and cheek movement data, texture data corresponding to "A", "I", "U", "E", "O", "N" are prepared in advance, and each "A" ...
Regarding "o", it is assumed that there are small to large open and close types. Since the vowel and sound intensity can be known from the current event information, the opening and closing amount texture corresponding to the sound intensity is specified from the textures corresponding to that vowel, and this texture is selected as lip movement data and cheek movement data. By doing so, lip sync can be realized.

Further, when lip-sync is performed by shape modification, similarly, the shape modification amount corresponding to the vowel analysis result and the sound intensity analysis result may be selected. In this case, a delay of at least one frame occurs after the event information is input. For example, at a frame rate of 30 frames / sec for drawing, about 30 milliseconds, so this delay is a problem for human visual ability. Don't

In the present embodiment, since a rod, which is an accessory component of a CG character, is handled, the rod motion pattern data representing the motion transition of the rod is also described, but in the case of other accessory components. Can also be implemented in the same manner, and can also be implemented in the case of a plurality of accessory parts. For example, FIG. 10C is a transition graph for controlling the movement of a rod (attached component), and the rod movement is changed from the rod basic state to the rod movement 0, 1, 2.
The transition to is defined as in the case of the transition graph of body movements.

Hereinafter, the case where the operation transition is performed by the stochastic fluctuation will be described, but the lip sync and the selection processing as described above can be fixedly performed. When the selection process is to be performed fixedly, a transition scenario may be prepared in advance and the operation may be selected according to the scenario.

FIGS. 11 (a) and 11 (b) are diagrams showing transition results in which operation transitions and part table transitions are performed based on a predetermined scenario, and these are used for the animation in the present embodiment. The image generation processing operation will be specifically described.

The animation control unit 1 performs the following control. That is, (1) management of a drawing frame of an animation image, (2) part I required for generation of the next frame image.
Management of address of motion data and frame number in motion data for each D, (3) management of event information and processing corresponding to the event, (4) control of motion transition for each part ID of CG character, (5) The address of the CG character shape and the accessory part shape, and the notification of the address of the hierarchical structure corresponding to the shape data to the character state calculation unit 3 (this is done only once before the CG animation image is generated),
(6) Notification of the part table number or address currently used to the character state calculation unit 3 (notification at the beginning of drawing, when the part table is changed), (7) operation data for each part ID required for the next frame image generation. To the character state calculation unit 3 of the address and the frame number in the motion data.

First, the address of the shape data of the CG character for generating the CG animation image, the address of the hierarchical structure data corresponding to the shape data, the shape data of the accessory part if there is an accessory part, and the shape data. The character state calculation unit 3 is notified of the address of the hierarchical structure data. If there are multiple CG characters or accessory parts, multiple minutes are notified. Next, based on the scenario of FIG. 11A, since the first selected part table is part table 0, in the initial frame, the basic state of physical movement is selected for part numbers 0 to 2,
In part numbers 3 and 4, facial expression 0 of facial expression motion is selected.
The selection result is notified to the character state calculation unit 3 as the state information in the following enumeration.

That is, (state information) part number 0, body motion (currently basic state) address, frame number 0
And part number 1, body motion (now basic state) address, frame number 0, part number 2, body motion (now basic state) address, frame number 0, part number 3, facial expression motion (now Is the address of facial expression motion 0), frame number 0, part number 4, facial expression motion (currently facial expression motion 0)
And the frame number 0.

When there are a plurality of characters, this state information is notified for each character. The character state calculation unit 3 acquires the hierarchical structure data from the hierarchical structure storage unit 4 by the address of the hierarchical structure data of the CG character notified in advance from the animation control unit 1 as described above, and the shape data by the address of the shape data. The shape data is acquired from the storage unit 5. Similarly, the part table storage unit 2 acquires the relevant part table (part table 0 in the present description) by the notified part table number or address. Based on the status information notified as described above and the acquired part table, the corresponding operation data is acquired from the operation data storage unit 6.

The operation data is obtained as follows.
That is, as shown in FIG. 9, it is understood from the information in the part table 0 that the motion ID set of the part number 0 is the motion identifiers m0 and m1. Next, the body movement data stored in the movement data storage unit 6 is obtained from the address of the body movement data of the state information (currently the basic state). From the acquired physical motion data, by the motion identifier (now m0, m1) and the frame number of the state information,
The physical motion data (one frame) of the corresponding frame number of the corresponding motion identifier is acquired. At that time, as described above, in the case of key frame data, the initial frame is always held as an immediate value, but it may be between key frames depending on the frame number.

When there is an inter-key frame, the maximum key frame including the frame number below the frame number below the frame number and the minimum key frame out of the frame numbers above the frame number are acquired and acquired. The two data thus obtained are linearly interpolated to generate the body motion data of the frame number. Higher-order polynomials such as spline interpolation (second or higher)
In the case of generating with, the same number of keyframe data as the order (in the above description of the linear interpolation, there are two, “maximum below” and “minimum above”) “maximum below and second largest , 3rd largest, ... ”or“ minimum and 2nd smallest, 3rd smallest, ... ”above, the required number of keyframe data can be obtained.

Similarly for the part numbers 1 and 2, the physical motion data is acquired from the motion data storage unit 6, and from the acquired physical motion data, the motion identifier corresponding to the part number in the part table 0 and the corresponding frame of the status information. Get physical activity data for a number. With the acquired body motion data (joint angle data) and hierarchical structure data, the acquired shape data undergoes matrix transformation in the local coordinate system defined for each part of the hierarchical structure (forward kinematics) to generate shape data, Further, based on the movement amount and the direction of the body movement data of the route (m0) of the specified part 0, the whole character shape data is subjected to the rotation conversion and the parallel movement conversion, and the shape of the entire character is changed to the specified direction and position. Generate data. It should be noted that a smooth shape with no discontinuity can be generated by using the envelope method during the conversion processing of the body condition.

Next, in the case where the part numbers 3 and 4 generate the facial expression motion by the deformation of the shape or the envelope method,
The same shape data as in the case of the above-mentioned body movement is generated to generate shape data. If the facial expression movement is archived texture data, the address of the facial expression movement is
This is the address of the archived texture data required for expression generation in the texture data storage unit 8, and the frame number corresponds to the management ID number of that archive. This address and the management ID number are used as the texture data of the parts (lips, left and right cheeks, left and right eyes, left and right eyebrows) of the newly generated character shape data (data in which the position, orientation, and physical condition have been changed). Write in the specified location.

The shape data of the character changed in the character state calculation unit 3 is used as the three-dimensional drawing unit 7
Sent to. In addition, when performing the collision determination of the character (contact or collision with another CG character, background shape, etc.),
This is performed by the character state calculation unit 3. As a method, when the above-mentioned change is made to the character shape data, it is sufficient to determine and calculate whether or not this shape intersects with other shapes. However, since the character shape data is generally complicated, in order to simplify the character shape data and the background shape data, the boundary box (the smallest rectangular parallelepiped including the character shape) or the boundary sphere (the smallest sphere pair including the character shape or A pair of ellipsoidal spheres) may be used as an alternative shape to perform a determination calculation as to whether or not they intersect.

In the three-dimensional drawing unit 7, a normal three-dimensional CG process is performed using the shape data of the character and the texture data specified in the shape data, the light source data and the camera data input from the outside. Then, the generated image is sent to the display unit 9 and displayed. For example, in the three-dimensional drawing unit 7, when the image is generated by the Guro-shading and the Z-buffer process for the three-dimensional CG processing and displayed, the brightness value of each vertex of the shape data is calculated from the light source data and converted into the shape data. The visual field conversion and the perspective conversion are performed (clipping processing may be performed to reduce the calculation amount), and the specified texture data is read from the texture data storage unit 8 and texture mapping is performed.

At this time, the shape color is determined by multiplying the color specified by the shape data by the calculated brightness value to calculate the vertex color, and the color between the vertices is determined by linear interpolation. In the case of mapping a texture, the position of the texture data is calculated according to the mapping coordinates designated in advance in the shape data, and the position is determined by multiplying it by the brightness value. By these processes, the color (RGB or YC
A format such as bCr is normally used) is determined, an image is generated, and the generated image is sent to the display unit 9 and displayed. Through the above processing, one frame image is generated. Such processing is repeated until the operation data of each part is completed.

Next, the processing when the operation data of a part is completed will be described. For example, in FIG. 11A, the body motion data of parts 0, 1, and 2 are
It ends before 4. This end is detected when the frame number required to generate the next frame image is larger than the end frame number of the motion data in (1) animation frame drawing frame management of the animation control unit 1. When the end is detected, the animation control unit 1 makes a motion transition for the part. For example, in the case of parts 0, 1, and 2, it is performed based on the body movement pattern data of FIG. 10A, but this process is performed as follows for each part.

For example, in the case of completely random motion transitions with equal probabilities (that is, the transition probability to each state is 1
/ 6), there are 6 states that can be shifted from the basic state, including the basic state. Therefore, by using the remainder system of 6, if the remainder is 0, the basic state, if 1 the state A, if 2 the state B, Three
If it is state C, if it is 4, state D if it is 5, and if it is 5, state E is determined in advance. Under this condition, the transition calculation may be performed in the procedure of (1) generating a random number, (2) dividing the random number value by 6, (3) examining the remainder, and (4) transitioning to a state that matches the remainder. The same can be done when the probability is not equal, and when the common denominator of the denominator of the transition probability (if it is a small number, it is converted to the fraction display) is changed to the common multiple display, the common multiple is a line segment of a number line and the line segment area equal to the numerator of each transition probability. It is sufficient to make a transition to a region in which the remainder when the random number is divided by the common multiple is entered.

The above is the operation transition in the case of being completely random. However, if the next operation of the specified part number is predetermined when an event occurs in the scenario, the next transition is specified. To make it work. For example, in the case of lip movement, if the next movement is "yes" while the mouth movement of "a" is being performed, it is designated as an event and the next movement is forced to be "yes". Transition to. It is also possible to combine such forced transitions with the above-mentioned random transitions, and when the event occurs, the transition probability of the next transitional action specified in advance is made higher than usual, and other transitions are possible. The transition probability of the motion is set lower than usual, and the above-described random number processing is performed on the modified transition probability.
In this case, although the transition to the pre-designated operation is high, the transition is not necessarily made as in the case of forced transition. Therefore, the regularity to the event is weakened, and it is possible to create an effect such as the whimsicalness that can be seen by an actual person.

As a result of the above transition processing, FIG.
In (a), part 0 is for action A and part 1 is for action B
Part 2 shows that the operation has transitioned to the operation C. In addition, regarding facial expression motion, after the motion of facial expression 0 is completed, part 3 changes from facial expression 0 to facial expression 2 by the above transition processing.
In Part 4, the facial expression changed from 0 to 3. Such processing is repeatedly performed to generate the CG animation image of the character.

Next, the transition of the part table will be described. In the present embodiment, when the event P occurs, the transition from the part table 0 to the part table 1 occurs (see FIG. 11A), and when the event Q occurs, the transition from the part table 0 to the part table 2 occurs. The description will be made assuming that it occurs (see FIG. 11B) and that it is designated in advance by the scenario. Although the event P and the event Q have been described above, basically any matters may be used.
For example, the content depends on the system, such as a key input or a voice analysis result when the CG character has reached a predetermined position, a contact or a collision with another shape, but the content itself does not matter in the present embodiment.

In addition to the scenario described above, it is also possible to reflect the result of emotion estimation as an event in the scenario on the motion of the CG character. For example, by using the emotion estimation method that uses the frequency signal of voice data such as prosody, amplitude, and stress using the voice of the speaker of a mobile phone, the emotion such as “anger”, “sadness”, and “joy” can be followed. By specifying the motion data etc.
It is conceivable that the character state calculation unit 3 changes the shape data according to the emotion estimation and reflects it on the CG character. At this time, since the CG character displayed on the screen is estimated by expressing the emotion of the speaker, it becomes possible to enhance the entertainment property of the computer graphics / animation device according to the present invention.

Explaining with reference to FIG. 11A, first, the occurrence of the event P is notified to the animation control unit 1 from the outside or the character state calculation unit 3 at the illustrated time. When the animation control unit 1 receives the notification, the animation control unit 1 proceeds to the process of transitioning from the part table 0 to the part table 1 based on the scenario. The most important thing in this process is the synchronization process. For this reason, part 0 is forcibly transitioned to the basic state after the current operation D is completed. Similarly, Part 1 is a motion E to a basic state, Part 2 is a motion A to a basic state, and Part 3 is a facial expression 1.
To facial expression 0, and for part 4, transition from facial expression 2 to facial expression 0 after the end of each of the actions currently being performed. When all these transitions are completed, part table 0 to part table 1
Is performed (transition time of event P in FIG. 11A).

At this time, although the number of parts and the motion ID set in each part are different between the part table 0 and the part table 1, it is the same as when the part table 1 generates the CG animation image in the initial part table. Therefore, it is possible to generate a CG animation image with continuous contents. By the above process, after the part table has changed from 0 to 1, the CG animation in which the motion transition based on the transition probability is repeated is continued in the same manner as above.

In the case of FIG. 11B, the processing from part 0 to part 4 is the same as in the case of the part table 0 of FIG. 11A, but in the part table 2, the hierarchical structure itself is changed. Since the addition of a rod shape (Part 5) occurs, the first rod movement after the transition is performed from the rod basic state for Part 5, and thereafter, a CG animation image is generated based on the above-mentioned movement transition control processing. To do. This is the same as processing is performed using the part table 2 as the initial part table, and therefore continuous CG animation images can be generated. By the above process,
After the transition, in the same manner as above, the CG animation in which the motion transition based on the transition probability is repeated is continued.

As described above, the CG character is divided into parts including a plurality of parts, and the motion data storage unit 6 holds a plurality of motion data corresponding to each part.
Then, the character state calculation unit 3 adds the motion data to each part in time series according to the scenario, and calculates the shape data indicating the temporal change of the shape of the entire CG character. Then, the three-dimensional drawing unit 7 adds texture data to the shape data to create a three-dimensional image of the CG character, and the display unit 9 displays the created three-dimensional image on the screen.

In order to prevent the stick shape from appearing suddenly, the stick shape may be displayed in an animation image generated in advance. Further, in order to prevent the situation where the stick suddenly flies, it is sufficient that the event Q involves contact with the stick. However, these things depend on the scenario, for example, in a magical scene,
Conversely, sticks may suddenly appear or fly away. The transition from the part table 2 to the part table 0 can be performed by performing similar processing. The transition from the part table 0 to the part table 2 corresponds to the addition of the shape and the hierarchy, and the transition from the part table 2 to the part table 0 corresponds to the deletion of the shape and the hierarchy. Moreover, although the shape of a single layer such as a rod has been described, the shape having another multi-layer structure can be handled by performing the same processing.

FIG. 12 is a reference diagram in the case of changing the motion of the CG character 1201 using a plurality of parts.
Here, the CG character 1201 is the face part 120.
2, upper body part 1203 and lower body part 120
It is divided into 4 parts. Also, each part 1202,
Part 1203 and part 1204 have three types of operation data. Although the three parts are described as having three operations here, it goes without saying that the present invention is not limited to this.

The face part 1202 has three types of facial expression motion data: facial expression 0 (normal face), facial expression 1 (crying face), and facial expression 2 (laughing face), and upper body part 1203 has basic state A (shake arms). ), Motion B (dancing) and motion C (bowing), three types of body motion data, lower body part 1
The reference numeral 204 has three types of body motion data: basic state D (standing), motion E (running), and motion F (sitting). Therefore, each of the three parts 1202, 1203 and 1
Since 204 has three types of operation data, C
The G character can perform a total of 27 movements.

For this reason, the animation device according to the present invention defines a plurality of parts including parts constituting a CG character, and adds motion data associated with each part, thereby reducing the amount of data as described above. CG by quantity
Not only the motion of the entire character can be controlled, but also complicated motion expression is possible for the entire CG character, and the problem that the user of the animation device gets bored in a short period of time because the motion expression form of the CG character is small. Can be prevented.

FIG. 13 is a diagram showing a mobile phone 1301 having the functions of the computer graphics animation device according to the present invention. Mobile phone 1301
A character 1303 and a character 1304 are displayed on the screen 1302. Animation control unit 1
Specifies motion data and the like according to the part division of the character 1303 and the character 1304, and the character state calculation unit 3 calculates shape data with time change for each part of the character 1303 and the character 1304 to calculate the CG character. Control the overall behavior. Therefore, the character 1303 and the character 130
In No. 4, the operation control independent of each other is performed and displayed on the screen. Further, when a plurality of CG characters are displayed in this way, it may be possible to share motion data or the like between the CG characters.

It is needless to say that the present invention is not limited to the above-described first embodiment and can be carried out within its usable range.

[0090]

With the above structure, the present invention introduces a new concept of a part independent of the skeleton structure, and uses a part table to diversify the number of part divisions in combination with respect to motion data. It is possible to generate an action of a different CG character. Moreover, since the transition of the part table is possible, not only the versatility thereof is further increased, but also addition and deletion of the shape and the hierarchical structure itself can be dealt with.

Furthermore, the animation device according to the present invention controls the operation of the entire CG character by assigning data to each part with a set of a plurality of parts constituting the CG character as a part. For this reason, it is not necessary to add data to all the parts constituting the CG character at once, and it is possible to change the motion of the entire CG character with a small amount of data.

Therefore, the present invention solves the conventional problems, and can eliminate the redundancy of the data, which was a major cause of the problems, to the utmost, and with a proper data amount smaller than the conventional one, various data can be obtained. A CG character animation can be realized. Moreover, although it is not liable to have a daily operation such as holding an object, it is easy to generate a CG animation image, which was difficult to realize by the conventional technique, by easily adding or deleting a shape or a hierarchical structure. You can do it. From the above, it is concluded that the effects of the present invention are clear and very important in the field of CG animation technology.

[Brief description of drawings]

FIG. 1 is a block diagram showing a configuration of a computer graphics animation device according to an embodiment of the present invention.

FIG. 2 is an explanatory diagram of a character shape and a posture vector.

FIG. 3 is an explanatory diagram of a skeleton structure (without a bar)

FIG. 4 is an explanatory diagram of the hierarchical structure (without bars) of FIG.

FIG. 5 is an explanatory diagram of key frame animation.

FIG. 6 is an explanatory diagram of a facial expression operation using texture.

FIG. 7 is an explanatory diagram of a skeleton structure (with a bar)

8 is an explanatory diagram of the hierarchical structure (with bars) of FIG. 7.

FIG. 9 is an explanatory diagram of a part table of the present invention.

10A is an explanatory diagram of body movement pattern data of the present invention, FIG. 10B is an explanatory diagram of facial expression pattern data of the present invention, and FIG. 10C is an explanatory diagram of stick movement pattern data of the present invention.

FIG. 11A is an explanatory diagram of the operation transition of the present invention and the transition of the part table. FIG. 11B is an explanatory diagram of the operation transition of the present invention and the transition of the part table.

FIG. 12 is a reference diagram when a motion of a CG character is changed using a plurality of parts.

FIG. 13 shows computer graphics according to the present invention.
Figure showing a mobile phone with the functionality of an animation device

[Explanation of symbols]

1 Animation control section 2 Part table storage 3 Character state calculator 4 Hierarchical structure storage 5 Shape data storage 6 Motion data storage 7 3D drawing part 8 Texture data storage 9 Display

   ─────────────────────────────────────────────────── ─── Continued front page    (72) Inventor Toshinori Hijiri             1006 Kadoma, Kadoma-shi, Osaka Matsushita Electric             Sangyo Co., Ltd. (72) Inventor Ryo Uezaki             1006 Kadoma, Kadoma-shi, Osaka Matsushita Electric             Sangyo Co., Ltd. (72) Inventor Shigeo Asahara             1006 Kadoma, Kadoma-shi, Osaka Matsushita Electric             Sangyo Co., Ltd. F-term (reference) 2C001 BA01 BB04 BC00 BC05 BC08                       CB03                 5B050 BA08 BA12 EA19 EA24

Claims (20)

[Claims] 1. A computer graphics animation device for displaying an image of a character, comprising hierarchical structure saving means for saving hierarchical structure data which is data indicating a hierarchical structure of a plurality of parts constituting the character, For a plurality of parts in which a set of a plurality of parts are associated as one part, a part table storage unit that saves a part table indicating a correspondence relationship between the part and the parts, and a plurality of parts registered in the part table. For each of them, a motion data storage means for storing motion data associated with a motion, an animation control means for designating the motion data, the part table and the hierarchical structure data, and an input from the animation control means. According to the operation data, the part table and the floor Based on the structure data, computer graphics animation apparatus, characterized in that it comprises a character state calculating means for calculating the shape data indicating the temporal change of the shape of the character. 2. The computer graphics according to claim 1, wherein the part table storage means stores a plurality of part tables, and correspondence relationships between the parts and the parts are different for each part table. Animation device. 3. The computer graphics animation apparatus according to claim 1, wherein the part included in one part is not included in another part in the part table. 4. The part table storage means further stores a part table including a correspondence relationship between the part and the accessory part for a part in which an accessory part other than the character is associated as a part. The computer graphics animation device according to claim 1. 5. The animation control means, according to the motion information according to a predetermined scenario that is input or the motion information according to an event specified when an event occurs, the motion data, the part table, and the hierarchical structure. The data is specified, and the data is specified.
Computer graphics animation device described. 6. The computer graphics animation apparatus according to claim 1, wherein the animation control means specifies the motion data for each part according to a transition probability or randomly. 7. The animation device further has an emotion estimation function, the animation control means specifies the motion data for each part according to the estimation of the emotion estimation function, and the character state calculation means The shape data indicating the temporal change of the shape of the character is calculated based on the motion data, the part table and the hierarchical structure data according to the input from the animation control means. Computer graphics animation device. 8. The computer graphics animation apparatus according to claim 1, wherein the animation control means makes a transition between part tables in which the collection of the parts included in the part is different. 9. The animation control means performs transition between part tables having different numbers of the parts included in the part table.
Computer graphics animation device described. 10. The motion data storage means has the motion data in which a plurality of motions are associated with each of the parts, and one of the motion data is a motion data in a basic state, and the animation control is performed. The means, when performing transition between the part tables, synchronizes all parts by forcibly transitioning to the operation data of the basic state for each part, and then transitions to another part table. The computer graphics animation apparatus according to claim 1, wherein 11. The motion data storage means has the motion data in which a plurality of motions are associated with each of the parts, and one of the motion data is motion data in a basic state, and the animation control is performed. The computer graphics animation apparatus according to claim 1, wherein the means forcibly specifies the motion data of the basic state for each part when the transition between the part tables is performed. 12. The animation control means specifies state information including the part, an identifier of the location of the action data, and a frame number indicating a time-series parameter of the action data, and the character state The calculation means calculates shape data indicating a temporal change of the shape of the character based on the motion data, the part table and the hierarchical structure data according to the input from the animation control means. The computer graphics animation device according to claim 1. 13. The motion data storage means has a plurality of motion data for each of the parts, and one of the motion data is the motion data of a basic state, and the animation control means includes an initial motion. 13. The computer graphics animation device according to claim 12, wherein the state information from the motion data of the basic state is forcibly designated as data. 14. The animation control means specifies, for each of the parts, time series data of joint angles, shape deformation data, or archived texture data as the motion data of the state information, and the character state calculation means. 13. The computer graphics animation apparatus according to claim 12, wherein said shape data calculates the shape data indicating the temporal change of the shape of said character in accordance with the designated motion data of said state information. 15. The animation control means specifies the motion data, the part table, and the hierarchical structure data of each of the plurality of characters, and the character state calculation means, in accordance with an input from the animation control means, 15. The computer graphics animation apparatus according to claim 1, wherein shape data indicating temporal changes in shapes of the plurality of characters is calculated. 16. The motion data includes facial motion data indicating a facial expression of the character, physical motion data indicating a physical motion, and part motion data indicating a motion of the accessory part. 1 to claim 1
5. The computer graphics animation device according to any one of 5 above. 17. The computer graphics animation apparatus further includes a three-dimensional drawing means for performing a three-dimensional drawing process on the shape data changed by the character state calculation means to generate a character, and the three-dimensional drawing. 17. A computer graphics animation device according to any one of claims 1 to 16, further comprising display means for displaying the character generated by the means. 18. A computer graphics animation method for displaying an image of a character, comprising a hierarchical structure saving step of saving hierarchical structure data which is data indicating a hierarchical structure of a plurality of parts constituting the character. For a plurality of parts in which a set of a plurality of parts is associated as one part, a part table saving step of saving a part table indicating a correspondence relationship between the part and the parts, and a plurality of parts registered in the part table. An operation data saving step of saving operation data corresponding to each operation, an animation control step of specifying the operation data, the part table and the hierarchical structure data, and an input from the animation control step. According to the operation data, Based on the bets table and the hierarchical structure data, computer graphics animation method characterized by including the character state calculating step of calculating shape data indicating the temporal change of the shape of the character. 19. A computer-readable data recording medium on which data for a computer graphics animation device for displaying an image of a character is recorded, wherein the data is a hierarchy of a plurality of parts constituting the character. Hierarchical structure data that is data indicating a structure, and a part table that stores a part table indicating a correspondence relationship between the parts and the parts, for a plurality of parts in which a set of the plurality of parts is associated as one part, A computer-readable data recording medium, comprising: motion data storing motion data in which a motion is associated with each of a plurality of parts registered in the part table. 20. A program for causing a computer to function as a means included in a computer graphics animation device for displaying an image of a character, the hierarchical structure data being data indicating a hierarchical structure of a plurality of parts constituting the character. A hierarchical structure saving step of saving; a part table saving step of saving a part table showing a correspondence relationship between the parts and the parts for a plurality of parts in which a group of the plurality of parts is associated as one part; A motion data storage step of storing motion data in which a motion is associated with each of a plurality of parts registered in the part table; and an animation control step of designating the motion data, the part table and the hierarchical structure data. , The animation A character state calculation step of calculating shape data indicating a temporal change of the shape of the character based on the motion data, the part table and the hierarchical structure data according to the input from the control step. Program to do.