JP2009053760A - Image processing apparatus, image processing method, and program - Google Patents

Abstract

An image processing apparatus that easily generates image data indicating the motion of a character in accordance with mechanics is provided.
An action generation unit that generates action information indicating an action to be performed by a character in a virtual space, a force torque calculation unit that calculates a force for operating the character, and a force torque calculation unit. The key of the character associated with the simulator 13 that applies a force to the rigid model imitating the character to calculate the state variable, the action information generated by the action generation unit 11 and the arbitrary state variable calculated by the simulator 13 A database 14 in which frame data is stored, and an interpolation processing unit that interpolates frame data according to the behavior information generated by the behavior generation unit 11 and the state variable calculated by the simulator 13 with reference to the key frame data 15 and a renderer that outputs image data corresponding to the frame data interpolated by the interpolation processing unit 15 And a 16.
[Selection] Figure 2

Description

  The present invention relates to an image processing apparatus that performs image processing for generating and outputting image data representing a character, an image processing method, and a program for causing a computer to execute the image processing.

  With recent advances in human interface and computer graphics, the environment surrounding the virtual world used for games, entertainment, media art, etc. has changed significantly. For this reason, human interface such as haptic interface and motion sensor, and video presentation technology such as computer graphics, 3D video, and high-resolution video dramatically increase the amount of information exchanged between the user and the virtual world. I let you. It is also possible to interact directly with the virtual world using the body as in the real world.

  In this regard, it can be said that it has become easier for the experienced person to feel reality and empathize with the virtual space. As a result, it is becoming more important to enhance the attractiveness of the virtual space itself in order to provide an attractive experience to the experienced person. In the virtual space, a character that artificially represents an object, an animal, a person, or the like is a constituent element. Among these components, the character is particularly attractive because it attracts the attention of the experiencer, becomes the center of interaction between the experiencer and the virtual space, and is also the subject of emotional transfer and emotion of the experiencer. This is important in creating creative content. For this reason, it is important to create a character that is more realistic and easy to empathize as an object of interaction between the experiencer and the virtual space in order to provide an attractive experience to the experiencer. .

  With such advancement of human interface, physical interaction with virtual space has come to take place. For this reason, it is necessary for the character to react to the physical interaction with the body according to the dynamics in order to improve the reality of the character. This is because the objects in the real world behave according to the dynamics, so the experiencer can make use of the real-world empirical rules in the virtual space, and easily understand the behavior in the virtual space, making it easy to operate and predict It is. For example, when the character presses the object, if there is a difference in the pressing method and the amount of movement according to the weight of the object, the experience person can estimate the weight of the object from the movement of the character.

  On the other hand, in order for the experience person to transfer the emotion to the character, it is necessary to convey the emotion and intention of the character to the experience person. For example, in a game, emotions may be expressed by speech balloons or symbols, but by expressing them with actions like theater or movie actors, more diverse emotions and intentions can be expressed, and emotions can be transferred to the experience. Can help. Character motion designers are required to create actions according to the personality of the character and the situation in which the character is currently placed in order to create a highly realistic character by such an experienced person. Moreover, in order to advance a scenario or a story, it may be requested | required that a specific attitude | position and operation | movement are carried out.

  For example, in Patent Document 1, individual animations such as “walking” and “jumping” are held as action objects, and a transition from one action object to another is performed by an event, and the action of each action object is completed. It describes how to generate an animation that accepts an event and moves it to another action object even if it is not, smoothly connecting individual animations, or automatically generating an animation that takes into account the change in animation for the event. Yes.

JP 2000-1199 A

  In order to realize an attractive character that has a high sense of reality and is easy to transfer emotions as a target of interaction, there is a problem that the designer has to create an enormous amount of motion for each character. For example, even when the generation method described in Patent Document 1 described above is used, since the designer creates many types of motion objects, the production cost becomes very high. For this reason, it is required to increase the productivity of content by creating various motion objects by combining still images of characters produced by a designer. Such processing requires a production environment that can be performed as easily as possible in accordance with the skills of a conventional designer.

  The present invention has been proposed in view of such a situation, and an image processing apparatus, an image processing method, and an image processing apparatus that easily generate image data indicating a character's movement according to dynamics in a virtual space. An object of the present invention is to provide a program for causing a computer to execute.

  As a means for solving the above-described problem, an image processing apparatus according to the present invention is an image processing apparatus that generates image data indicating a motion of a character in a virtual space, and an action indicating an action to be performed by the character. Generating means for generating information; calculating means for calculating a force for operating the character in the virtual space in accordance with action information generated by the generating means; and calculating the force calculated by the calculating means for the character Simulation means for calculating a state variable indicating the dynamic behavior of the rigid model, each behavior information generated by the generation means, and any state variable calculated by the simulation means A database in which the posture data relating to the associated posture of the character is stored, and the database Interpolating means for interpolating the character's posture data in accordance with the action information generated by the generating means and the state variable calculated by the simulation means with reference to the stored posture data, and the interpolation means Output means for outputting image data indicating the motion of the character according to the interpolated posture data.

  Further, an image processing method according to the present invention is an image processing method for generating image data indicating a motion of a character in a virtual space, the generating step generating action information indicating an action to be performed by the character, In accordance with the action information generated by the generation step, a calculation step for calculating a force for moving the character in the virtual space, and a force calculated by the calculation step are given to a rigid model imitating the character, A posture for the posture of the character associated with a simulation step for calculating a state variable indicating the mechanical behavior of the rigid model, each action information generated by the generation step, and an arbitrary state variable calculated by the simulation means Referring to the database storing the data, the behavior information generated by the generation step and the An interpolation step for interpolating the posture data of the character according to the state variable calculated by the simulation step, and an output step for outputting image data indicating the motion of the character according to the posture data interpolated by the interpolation step; Consists of.

  The program according to the present invention is a program for causing a computer to execute image processing for generating image data indicating the motion of a character in a virtual space, and generating action information indicating an action to be performed by the character. A generating step, a calculating step for calculating a force for operating the character in the virtual space according to the action information generated by the generating step, and a rigid body imitating the character calculated by the calculating step A simulation process for calculating a state variable indicating the mechanical behavior of the rigid model, and each action information generated by the generation process and an arbitrary state variable calculated by the simulation process Referring to the database that stores the posture data related to the character's posture, An interpolation process for interpolating the posture data of the character according to the action information generated by the process and the state variable calculated by the simulation process, and an action of the character according to the posture data interpolated by the interpolation process. And an output process for outputting the image data shown.

  The present invention refers to a database in which posture data relating to the posture of a character associated with each behavior information and an arbitrary state variable is stored, and the above-described information according to the behavior information generated and the calculated state variable. Since the character's posture data is interpolated and image data indicating the character's motion according to the interpolated posture data is output, image data indicating the character's motion according to the mechanics in the virtual space can be easily generated. it can.

  An image processing apparatus to which the present invention is applied performs image processing that generates and outputs image data indicating the motion of a character in a virtual space. As an example of such an image processing apparatus, a mode for carrying out the present invention will be described below using an image processing apparatus 1 as shown in FIG.

  The image processing apparatus 1 is a computer including a general-purpose arithmetic processing processor, and includes a CPU 2 that performs arithmetic processing, a hard disk 3 that stores image data and a program for executing an application, and data that the CPU 2 performs arithmetic processing. Are temporarily stored, an interface 5 for inputting an operation input from the user from the outside, and a display 6 for displaying a result of arithmetic processing by the CPU 2.

  The CPU 2 reads out a program for performing image processing for generating image data indicating the motion of the character in the virtual space from the hard disk 3 to the RAM 4 and executes it.

  The hard disk 3 stores image data generated by arithmetic processing by the CPU 2 and a program for executing the above-described image processing.

  The RAM 4 temporarily stores data on which the CPU 2 performs arithmetic processing.

  The interface 5 receives an operation input from the user and supplies operation information related to the operation input to the CPU 2.

  The display 6 displays image data generated by arithmetic processing by the CPU 2.

  In the image processing apparatus 1 configured as described above, the CPU 2 reads out from the hard disk 3 the RAM 4 a program for performing image processing for generating image data indicating the movement of the character in the virtual space. Such an image processing unit 10 is realized.

  That is, as shown in FIG. 2, the image processing unit 10 includes a behavior generation unit 11 that generates behavior information indicating the behavior of the character, a force torque calculation unit 12 that calculates information about the force and torque that causes the character to operate, A simulator 13 that calculates a state variable indicating the mechanical behavior of a rigid model imitating a character, and a database 14 that stores in advance posture data relating to the posture of the character associated with predetermined action information and an arbitrary state variable; An interpolation processing unit 15 that interpolates the posture data, and a renderer 16 that generates image data indicating the action of the character according to the posture data interpolated by the interpolation processing unit 15.

  In the image processing unit 10 configured as described above, first, a simulator 13 that calculates a state variable indicating the mechanical behavior of a rigid model imitating a character will be described.

  The simulator 13 gives a force Fp and a torque Np calculated by a later-described force torque calculation unit 12 to a character rigid body model 131 that imitates a character as shown in FIG. Then, the simulator 13 simulates mechanical behavior such as collision and friction generated between the character rigid body model 131, the object rigid body model 132 imitating an object such as a wall, and the floor rigid body model 133 imitating the floor. Then, a state variable indicating the mechanical behavior of each rigid body model is calculated. Here, the state variables are, for example, the position r of the rigid model with respect to the preset origin on the three-dimensional plane, the speed v obtained by differentiating the position r with time, the rotation angle θ with respect to the center of gravity of the rigid model, and the rotation angle θ with time. This is information indicating the differentiated angular velocity ω and the like. The simulator 13 calculates state variables relating to all rigid body models of the character rigid body model 131, the object rigid body model 132, and the floor rigid body model 133 as environment information and supplies them to the action generation unit 11.

  In the simulator 13, the character is expressed using a character rigid model 131 having no joints. This is because the multi-joint model, which is a dynamic model for calculating the position and rotation angle of each part of the character, has a large number of state variables to handle and requires complicated calculations, and it is difficult to adjust to obtain the desired action. It is. In view of this, the image processing unit 10 simulates the mechanical behavior of the character using the rigid body model imitating the character in the simulator 13, and is generated by the interpolation processing unit 15 described later based on the state variables obtained by the simulation. The posture data expresses the joint angle of each joint of the character, and the renderer 16 outputs image data indicating the action of the character.

  Moreover, in the simulator 13, since the shape of each rigid body model is used for the collision determination between rigid body models, the thing close | similar to the external shape of the character displayed as an image | video is used. In this way, the simulator 13 can prevent the phenomenon that the objects are separated from each other or squeezed in the video when the character obtained from the simulation result contacts another object.

  In the simulator 13, the designer can easily create a desired motion by setting the characteristics of the rigid model as follows.

  First, the simulator 13 designs a character rigid body model 131 having a plane portion 131a as shown in FIG. 4A, so that the character rigid body model 131 and the object rigid body as shown in FIG. It is possible to make the character rigid body model 131 easily face the plane portion 132 a of the object rigid body model 132 at the time of contact with the model 132. Second, the simulator 13 can easily adjust the rotation motion of the character by using the sides and vertices of the rigid model as fulcrums for rotation. Third, the simulator 13 sets the shape, mass, and center of gravity of the character rigid body model 131 according to the character, and adjusts the inertia tensor L of the following expression, which is the rotation characteristic of the character, according to the set information. Can do.

N = Lω
Here, N is a torque indicating a rotational force applied in the direction of the rotational angle θ, and is a value provided by a force torque calculation unit 12 described later. Usually, a unit matrix is used for L.

  The simulator 13 uses a character 200 as shown in FIG. 5 as a specific example. Here, the origin is defined as the bottom of the character, and the three-dimensional orthogonal coordinate axes are defined as follows. That is, the width direction of the character 200 is defined as the x axis, the vertical direction is defined as the y axis, and the front direction is defined as the z axis. Further, when an object collides with the front of the character 200, that is, when an external force F1 is applied in the −z-axis direction, a torque N1 that rotates with respect to the center of gravity G of the character 200 is generated more so as to make it easier to roll. Therefore, in the simulator 13, as shown in FIG. 6A, the character rigid body model when the rigid body 210 having such a shape that the cross-sectional area of the xz plane decreases in the −y-axis direction collides with the object rigid body model 132. Used as 131. Further, in order to make sure that the character is in a state of lying or lying on the floor when the character falls on the floor, as shown in FIG. 6B, the simulator 13 forms a rigid body 220 having a shape that easily falls in the ± z-axis direction. It is used as the character rigid body model 131 when it collides with the rigid body model 133. As described above, the simulator 13 can easily express the natural movement of the character 200 by using a character rigid body model 131 having a different shape for each colliding object.

  In this specific example, the simulator 13 calculates a state variable of the character 200 in a multidimensional space (hereinafter referred to as a key frame space K) defined as follows.

  In the simulator 13, a function F for converting a character from a rotation matrix R indicating the position r, velocity v, and direction of the rigid body model in the key frame space K to a reference point located on the key frame space K It is defined as equation (2).

  In accordance with the environment information calculated by the simulator 13 as described above, the behavior generation unit 11 determines the character's behavior, and the behavior information regarding the determined character's behavior and the state variable of the character are determined as the force torque calculation unit 12. To supply. Further, the behavior generation unit 11 supplies this behavior information to the interpolation processing unit 15. Here, for example, the behavior generation unit 11 determines the behavior of the character by a predetermined AI (Artificial Intelligence) process according to the environment information, but sequentially determines the behavior of the character by a user operation from the interface 5. Also good.

  The force torque calculation unit 12 calculates information on the force and torque for operating the character in accordance with the action information supplied from the action generation unit 11 and the environment information calculated by the simulator 13. For example, in order to move the character 200 as shown in FIG. 5 described above, the force Fp and torque Np applied to the character 200 are calculated as shown in FIG. Here, the force Fp is a force that the character 200 generates with the legs, that is, a force that acts in the z-axis direction at the bottom of the rigid model. Further, the torque Np is a force for rotating the x axis direction at the center of gravity G of the character 200 as the central axis.

  The force torque calculation unit 12 performs the force Fp and the force Fp according to two controls: a posture control for stabilizing the posture of the character 200 in an upright state, and a movement control for moving the character 200 while the posture is stable. Torque Np is calculated.

  As shown in FIG. 8A, the force torque calculator 12 calculates a force Fp and a torque Np according to the state transition as posture control. Further, θx is a state variable indicating the posture of the character. Specifically, as shown in FIG. 8B, the rotation angle of the character rigid body model 131 with respect to the x-axis is specifically shown, and the initial state is θx = 0 °. To do. Moreover, the force torque calculation part 12 changes control mode as follows according to the state variable supplied from the action production | generation part 11. FIG.

  First, in the normal state mode S11, the force torque calculation unit 12 calculates the force Fp and the torque Np as shown in the following equation (3).

  Thereafter, when | θx |> θ1 is satisfied, the force / torque calculation unit 12 shifts to the inverted pendulum control mode S12 in order to restore this state to the upright state because the posture of the character 200 is out of balance.

  In the inverted pendulum control mode S12, the force torque calculation unit 12 calculates the force Fp and the torque Np as shown in the following equation (4).

  Thereafter, the force torque calculation unit 12 transitions to the normal state mode S11 when | θx | ≈0 ° and assumes that the character 200 cannot be restored to the upright state when | θx |> θ2, and the character 200 falls down. In order to perform the behavior until calming down, the transition is made to the stabilization standby mode S13.

  In the stabilization standby mode S13, the force torque calculation unit 12 calculates the force Fp and the torque Np as shown in the following equation (5).

  After that, the force torque calculating unit 12 transitions to the normal state mode S11 when | θx | ≈0 and | dθx | ≈0, and transitions to the standing mode S15 when | dθx | ≈0 and the character 200 is in the supine posture. .

  In the standing mode S15, the force torque calculation unit 12 calculates the force Fp and the torque Np as shown in the following equation (6).

  Thereafter, the force torque calculator 12 transitions to the normal state mode S11 when | θx | ≈0 °.

  Further, since the force torque calculation unit 12 performs movement control to move the position r when the character 200 is in the normal state mode, the force Fp and the torque Np are respectively expressed by the following equations (7) and (8). Is calculated.

  Here, Kpf and Kpn are control parameters of the PD control system. In the force torque calculator 19, for example, as shown in Table 1, the above control parameters are set so that each value associated with each type of operation is calculated.

  The database 14 stores posture data (hereinafter referred to as key frame data) relating to the character posture associated with each piece of behavior information generated by the behavior generation unit 11 and an arbitrary state variable calculated by the simulator 13. Yes. Specifically, the database 14 stores joint angle data of each joint constituting the character as the key frame data.

  Further, for example, as illustrated in FIG. 9, the database 14 is associated with a key frame set 141 associated with action information indicating that the character “walks” and action information indicating that the character “runs”. Each key frame data is classified and stored for each key frame set 143 associated with the key frame set 142 and the action information indicating the action of the character “pressing”.

  The key frame data stored in the database 14 is arranged in a multidimensional space corresponding to the state variables calculated by the simulator 13 for each of the key frame sets 141, 142, and 143 described above.

  For example, in the key frame set 141 associated with the action information indicating that the character “walks”, the key frame data 2 corresponds to the state variable indicating the position r and the rotation angle θ of the character as shown in FIG. Arranged in dimensional space.

  That is, the key frame set 141 includes key frame data P11 to P14, P21 to P24 corresponding to motions of the character rotation angle from 0 ° to 45 °, as key frame data indicating “motion according to the character's intention”, In addition, key frame data P31 to P34 corresponding to the motion of the character from 45 ° to 90 ° as key frame data indicating “motion contrary to the intention of the character”. Here, the “movement according to the character's intention” is a movement of the character in response to the state variable calculated by the simulator 13. Further, “an action contrary to the character's intention” is an action in which the character is out of balance without being able to walk even if he or she tries to walk.

  As described above, the reason why the key frame set 141 is provided with the key frame data corresponding to the “motion contrary to the intention of the character” is as follows.

  That is, in image processing for generating image data indicating a character, for example, even if the action generation unit 11 generates action information that causes the action generation unit 11 to “walk” in a state where the character is falling, the action corresponding to the action information is generated. This is because there are states that cannot be performed.

  In addition, the key frame data is created in such a manner that only a specific dimension is changed first in the multi-dimensional space of each key frame set 141, 142, 143, and another dimension is changed from each created key frame data. Things are created. For example, in the key frame space K corresponding to the key frame set 141, as shown in FIG. 11, first, key frame data P111, P121, and P131 indicating the posture of the character according to the rotation angle are created (S21). . In the key frame space K, key frame data P112, P113, and P114 indicating the posture in which the character moves while maintaining the rotation angle indicated by the key frame data P111 are subsequently created (S22). Similarly, in this key frame space K, key frame data P122, P123, and P124 indicating the posture in which the character moves while maintaining the rotation angle indicated by the key frame data P121 are created (S23). Key frame data P132, P133, and P134 indicating the posture in which the character moves while maintaining the indicated rotation angle are created (S24).

Further, in order to prevent dissociation between the mechanical behavior of the rigid body model and the character displayed as an image, a three-dimensional coordinate system serving as a reference of joint angle information of each joint constituting each key frame data is shown in FIG. As described above, the character coordinate system x ₁ -y ₁ -z fixed to the rigid body model corresponding to the key frame data P201, P202, and P203 indicating the characters having different rotation angles through the points on the key frame space K. is used _{1, x 2 -y 2 -z 2} , x 3 -y 3 -z 3. In addition, each key frame data is created so that the character does not exceed the rigid body model of the corresponding rotation angle because the display portion of the character that protrudes from the rigid body model exceeds the range of the collision determination by the simulator 13. Use things.

  Specifically, the database 14 stores key frame sets as shown in Table 2 as key frame sets corresponding to the key frame space described above.

  The interpolation processing unit 15 refers to the key frame data stored in the database 14 according to the behavior information generated by the behavior generation unit 11 and the state variable calculated by the simulator 13, and refers to the key frame referred to. Data is interpolated to generate key frame data in the state variable.

  Specifically, in the interpolation processing unit 15, among the key frame data associated with the behavior information generated by the behavior generation unit 11, the character from the key frame data in which the state variables calculated by the simulator 13 are close to each other Are calculated, and frame data that is character posture data corresponding to the calculated joint angles is interpolated.

  That is, the interpolation processing unit 15 first searches the database 14 for a key frame set corresponding to the behavior information generated by the behavior generation unit 11. Subsequently, as illustrated in FIG. 13, the interpolation processing unit 15, the reference key frame data P <b> 1 and P <b> 2 that are close to each other in the two-dimensional space of the position r and the rotation angle θ that are state variables calculated by the simulator 13. , P3 and P4 are read from the database 14, and the following linear interpolation processing is performed to generate frame data Pi corresponding to the position r and the rotation angle θ. That is, the interpolation processing unit 15 generates frame data obtained by linearly interpolating the joint angles of the joints constituting the character in a state where the character coordinate systems in the reference key frame data P1, P2, P3, and P4 are matched. The frame data Pi to be interpolated is generated by inclining the state according to the rotation angle. Then, the interpolation processing unit 15 supplies the generated frame data Pi to the renderer 16.

  The renderer 16 performs rendering processing on the frame data Pi generated by the interpolation processing unit 15 to generate image data for display and supplies it to the display 6. Specifically, the renderer 16 synthesizes an image showing each joint constituting the character according to the frame data interpolated by the interpolation processing unit 15 to generate image data indicating the action of the character, and displays it for display. The image data is supplied to the display 6.

  In the image processing unit 10 including the processing unit as described above, interpolation is performed by referring to the database 14 in which key frame data is stored as posture data relating to the posture of the character associated with each action information and an arbitrary state variable. Since the processing unit 15 interpolates and outputs the character frame data corresponding to the behavior information generated by the behavior generation unit 11 and the state variable calculated by the simulator 13, the character of the character according to the dynamics in the virtual space is output. Image data indicating the operation can be easily generated.

  Therefore, in the image processing unit 10, in order for the experience person who experiences the virtual space to realize a physical interaction with the character in the virtual space, the character responds to the movement according to the mechanics, the collision with the wall or the like. It is possible to generate image data showing the dynamic behavior in real time.

  Further, since the image processing unit 10 generates image data indicating a character's motion using a key frame prepared in advance, there is no need to prepare other types of moving image data corresponding to a series of motions, and an animation image is generated. Production costs to produce can be reduced.

  In addition, this invention is not limited only to the above embodiment, Of course, a various change is possible in the range which does not deviate from the summary of this invention.

It is a block diagram which shows the whole structure of an image processing apparatus. It is a block diagram which shows the structure of each process part with which an image process part is provided. It is a figure which shows typically each rigid body model implement | achieved in virtual space. FIG. 4A is a diagram illustrating an example of a character rigid body model, and FIG. 4B is a diagram schematically illustrating a collision between a character rigid body model and an object rigid body model. It is a figure which shows an example of a character. It is a figure which shows an example of a character rigid body model. It is a figure which shows an example of a character. FIG. 8A is a state transition diagram provided for explaining the processing flow of attitude control, and FIG. 8B is a diagram provided for explaining the rotation angle to be controlled. It is a figure which shows typically the key frame set memorize | stored in the database. It is a figure which shows typically the key frame arrange | positioned on two-dimensional space. It is a figure provided in order to demonstrate the production process of each key frame which comprises a key frame set. It is a figure which shows the character coordinate system fixed to the character for every attitude | position. It is a figure which shows typically the frame data produced | generated by the interpolation process.

Explanation of symbols

  1 image processing device, 2 CPU, 3 hard disk, 4 RAM, 5 interface, 6 display, 10 image processing unit, 11 action generation unit, 12 force torque calculation unit, 13 simulator, 14 database, 15 interpolation processing unit, 16 renderer, 131 character rigid body model, 131a plane part, 132 object rigid body model, 132a plane part, 133 floor rigid body model, 141, 142, 143 key frame set, 200 character, 210, 220 rigid body

Claims (4)

In an image processing apparatus that generates image data indicating the motion of a character in a virtual space,
Generating means for generating action information indicating actions to be performed by the character;
Calculating means for calculating a force for moving the character in the virtual space according to the action information generated by the generating means;
Simulation means for applying a force calculated by the calculation means to a rigid model imitating the character and calculating a state variable indicating a mechanical behavior of the rigid model;
A database in which posture data relating to the posture of the character associated with each action information generated by the generating unit and an arbitrary state variable calculated by the simulation unit is stored;
Interpolating means for interpolating the posture data of the character according to the action information generated by the generating means and the state variable calculated by the simulation means with reference to the posture data stored in the database;
An image processing apparatus comprising: output means for outputting image data indicating the motion of the character according to the posture data interpolated by the interpolation means. The database stores joint angle data of each joint constituting the character as the posture data,
The interpolation means includes the posture data associated with the behavior information generated by the generation means, and the posture data of the joints constituting the character from the posture data in which the state variables calculated by the simulation means are close to each other. The image processing apparatus according to claim 1, wherein a joint angle is calculated, and the posture data of the character corresponding to the calculated joint angle is interpolated. In an image processing method for generating image data indicating a motion of a character in a virtual space,
A generation step of generating behavior information indicating the behavior to be performed by the character;
A calculation step of calculating a force for moving the character in the virtual space according to the action information generated by the generation step;
A simulation step of applying a force calculated by the calculation step to a rigid body model imitating the character and calculating a state variable indicating a mechanical behavior of the rigid body model;
Generated by the generation step with reference to a database storing posture data relating to the posture of the character associated with each action information generated by the generation step and an arbitrary state variable calculated by the simulation means An interpolation step of interpolating the posture data of the character according to the action information to be performed and the state variable calculated by the simulation step;
An image processing method comprising: an output step of outputting image data indicating the motion of the character according to the posture data interpolated by the interpolation step. In a program for causing a computer to execute image processing for generating image data indicating a motion of a character in a virtual space,
A generation step of generating behavior information indicating the behavior to be performed by the character;
A calculation step of calculating a force for moving the character in the virtual space according to the action information generated by the generation step;
A simulation step of applying a force calculated by the calculation step to a rigid body model imitating the character and calculating a state variable indicating a mechanical behavior of the rigid body model;
Generated by the generation step with reference to a database storing posture data relating to the posture of the character associated with each action information generated by the generation step and an arbitrary state variable calculated by the simulation step An interpolation process for interpolating the posture data of the character according to the action information and the state variable calculated by the simulation process;
A program for causing a computer to perform image processing including an output step of outputting image data indicating the motion of the character according to posture data interpolated by the interpolation step.

Images