JPH08329272A - Animation generator method - Google Patents

Abstract

PURPOSE: To provide the computer animation generation method for easy work. CONSTITUTION: Human fundamental motions are analyzed to input forces or torques acting on joints to a data base, and dynamical equations are applied to calculate respective parts of the body in distinction from the other parts, and restrictive conductions including mutual coupling relations of respective parts of the body and ranges of motions of respective joints are checked, and inverse dynamics are applied to calculate the motions and forces checked by these restrictive conditions, and dynamical equations are applied to calculate respective parts of the body in distinction from the other parts for the purpose of generating the animation, and restrictive conditions including mutual coupling relations of respective parts of the body and ranges of motions of respective joints are checked, and inverse dynamics are applied to calculate the motions and forces checked by these restrictive conditions, and the result of these motions and forces or the like is displayed.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an animation creating method for reproducing a human motion on a screen by using a computer.

[Prior art]

Computer animation has become more and more realistic and beautiful in recent years as Williams, Isaacs, Forsey and Selby have taken into account the physical properties of the object and the physical laws governing the object. ing. The method used by Nigel and Calvert to represent the movement of objects in animation is called kinematics. This method ignores forces and torques and expresses movements using only position, velocity, and acceleration. However, in animations to which this kinematics is applied, the animator must clearly design the movements. I won't.

However, since the motion of an object is determined by its own nature and its interaction with the environment, it is not necessary to measure the phenomena required to generate a realistic motion in animation in which kinematics is applied. It's not easy.

On the other hand, there is a method known as dynamics for discussing the motion of an object in relation to force, and if this method is applied to animation creation, complicated motions can be created with few operations. can do. In addition, according to the animation creation method to which the dynamics are applied, another big advantage, that is, the limitation of the motion design in the animation creation to which the kinematics is applied can be avoided. In the following description, the term inverse dynamics is used, but in general, the dynamics give a force to move (for example,
(Acceleration) is calculated, whereas the inverse calculation, that is, the force is calculated from the motion, and the term is different but the same method.

However, in the animation creating method to which the dynamics are applied, dynamic parameter data such as the moment of inertia which is difficult to measure, the center of gravity, the friction of the joint, the elasticity of the muscle / ligament and the like are required, and these data are required. Without it, the movement becomes unrealistic as in the case of the animation creation method applying kinematics. Moreover, it is necessary to solve a relatively complicated dynamic equation, and it has joints and two degrees of freedom.
For a human body with as many as 00, as many as 600 differential equations must be solved simultaneously.

[0006] Therefore, it cannot be said that the animation creation methods to which dynamics have been proposed so far are suitable for reproducing actual movements, and dynamics for reproducing complete movements of a human body or an animal body cannot be said. There was no animation method that applied it and considered the power of movement.

Until now, research based on artificial intelligence (AI) and expert systems has been conducted by Zelzer et al. In order to acquire the knowledge and skills of animators. In addition, Badler used constraints-based animation for his dynamics, and Drewery and Zotzos studied frame-based methods for goal-oriented animation.

Knowledge in these studies includes size,
Basic knowledge of the movement of the human or animal body, such as the moment of inertia, and constraints that define the range of movement of each joint are included, but this knowledge is obtained through the intuition of the animator.

A conventional animation creating method to which dynamics are applied is as follows: Create a human body model 2. Input actual human movements 3. Analyze the input motion 4. Create a new action 5. It consists of five stages of displaying movements on the screen.

In the first step, "creating a human body model", the human body is decomposed into parts that are the minimum units of motion, and the constraints peculiar to each of these parts, mutual relationships, joint motion ranges, etc. A human body model is created based on the data, and is input to a computer database as data created in an appropriate computer language.

In the second step "input actual human motion", the motion to be calculated is input to the computer database in the same way as in the first step in units of video image frames or frame units of film images. To do. In this case, if the images taken at the same time from a plurality of directions are used, the calculation in the next step can be performed more specifically.

In the third step, "Analysis of the input motion", the center of gravity of each part and the center of gravity of the whole of the motion input in the second stage are calculated, and the force acting on each joint is calculated by using the inverse dynamics. And torque, and the force and torque acting on the overall center of gravity are analyzed.

In order to calculate the motion of each part, in this motion analysis, Newton's equation is used to obtain the linear acceleration of the center of gravity, and Euler's equation is used to obtain the angular acceleration of the center of gravity. When is obtained, these are integrated to obtain the velocity, and further integrated to obtain the position.

In the fourth step of creating this new motion, the method proposed by Wilhelms for precisely solving the dynamic equation using Gibbs' equation is excellent. However, the Wilhelms method has a complicated concept, and the calculation amount O (f (n)) is n ⁴ when n is the number of parts constituting the human body which is the minimum unit of motion in animation.
Since it is a function O (n ⁴ ) of and has a large amount of calculation, it is not used in an actual animation because the calculation by the computer is expensive.

On the other hand, Armstrong calculates O (f
We have proposed an animation creation method using a calculation where (n)) is a function O (n) of n. The Armstrong method is based on the idea that there is a constant relationship between the linear and angular accelerations of a joint, but this relationship only holds when rotation about an axis can be ignored. Therefore, if only the line drawing that represents the parts that make up the human body with lines can be displayed on the screen and the rotation around the axis cannot be ignored, the method of Armstrong is used. Cannot be used. Therefore, according to the method of Armstrong, the human body cannot be three-dimensionally fleshed and displayed realistically.

As described above, in the actual animation work, although the method of responding in real time by the easy-to-use interactive form is required, the animation work up to now is performed by trial and error and responds in real time by the interactive form. It wasn't something.

In the fifth step of "displaying motions on the screen", a new motion during or after the design is displayed on the screen, but the inverse dynamics is applied to the relationship between the motion and force of the human body displayed at that time. Is calculated using the Lagrangian equation, and the position of the center of gravity and the direction of the force generated at each joint of the body are combined and displayed on the human body.

[0018]

And that require complex calculations of the in animation creating method according to the kinetics [0005], relates to human motion between the fact that there is no easy-to-use method Interactive There is a common problem of lack of knowledge. Therefore, in the present invention, a new animation creating method applying dynamics is proposed, and a new human body motion is completely created by using the knowledge about the human motion obtained by analyzing the actual human motion. To do.

That is, the present invention is an animation creating method capable of creating a three-dimensional motion of a human body or an animal body which is not a line drawing by using a computer interactively without trial and error or intuition of an animator. The challenge is to provide.

[0020]

In order to solve the above-mentioned problems, in the animation creating method to which the dynamics of the present invention is applied, first, the basic movements of humans and animals are analyzed, and the human and animal bodies are respectively analyzed. Data of dynamic parameters including forces and torques generated in the joints of the above are input to the database as knowledge about basic movements.

Next, animator access the database, but to process the obtained data, the computer in the form of a motion that is constrained in the constraint, the inverse dynamics calculation binding <br/> results the power of Design the movement interactively by feeding back to the animator in real time in the form and repeating this process until a satisfactory result is obtained.

The calculation amount of the animation creating method to which the dynamics are applied is a function O (n) of n, which solves the problem that the calculation by the computer, which is present in the Wilhelms method, is expensive. Also, it is possible to provide an animation creating method that can realize a three-dimensional and realistic movement of a human body or an animal body that is not a line drawing but is smoothly fleshed out interactively without trial and error or the intuition of an animator. it can.

[0023]

BEST MODE FOR CARRYING OUT THE INVENTION Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 shows a flowchart of the present invention. Create a human body model 2. Input actual human movements 3. Analyze the input motion 4. Design of movement 5. Dynamics 6. Apply constraints 7. Apply inverse dynamics 8. It consists of each stage of displaying the results on the screen.

In the first step "creating a human body model", the human body is decomposed into parts that are the minimum units of motion (that is, parts that are separated by joints), and the properties peculiar to these parts are
A human body model is created based on mutual relations and constraint conditions such as the range of joint movement, and is input to a computer database as data described in an appropriate computer language.

In the second step "input actual human motion", the motion to be calculated is input to the database of the computer in the frame unit of the video image or the frame unit of the film image as in the first stage. . In this case, if the images taken at the same time from a plurality of directions are used, the calculation in the next step can be performed more specifically.

In the third stage, "Analysis of input motion", calculation is performed using inverse dynamics from the motion input in the second stage, and the center of gravity of each part, force and torque acting on each joint, and total The center of gravity, the force acting on the center of gravity, and the torque are analyzed.

When only the input motion is analyzed,
The center of gravity of each part obtained in the third step, the force and torque acting on each joint, the overall movement and center of gravity, the force and torque acting on the entire center of gravity are overlaid on the actual motion input by arrows and the like and displayed on the screen. indicate. In this way, the operation is analyzed.

Next, a case of designing a new motion using the above analysis result will be described. In order to design a new motion, the human body model data of the first stage, the actual human motion data obtained in the second stage, and the analysis result data obtained in the third stage are used in the same manner as above. Fill in the database.

In the fourth stage motion design, the animator first selects a basic motion from the database. Figure 2 shows a control graph created based on a database of left elbow movements in Shorinji Kempo, which is a type of martial arts. In this control graph, the horizontal axis represents time.
The vertical axis represents the force generated at each joint of the body with respect to the three axes of X, Y, and Z. As a matter of course, the two forces generated in the same joint have the same magnitude and opposite directions.

Further, complicated operation is represented by several control graphs. For example, a control graph representing a motion of a person standing up from a chair and walking is configured by combining continuous motions. The graphs for the other parts are designed in the same way as for the left elbow shown in the example.

Next, for all the parts of the human body where the force is generated, the physical variables including the enlargement / reduction of the abscissa and ordinate of the control graph are collectively changed and a part of the body is changed. Partial changes are made to change physical variables such as force generated.

In the fifth stage, "Applying Dynamics",
The motion of each part is calculated based on the force specified by the animator and the dynamic equation governing the motion of each part. In that case, as shown in FIG. 3 (a), each part of the body originally has a mutual connection relationship, but in order to reduce the calculation amount, each part of the body is separated from other parts as shown in (b). It is released, and the constraint condition regarding the mutual connection relation of each part of the human body and the range of motion of the joint is temporarily ignored (see (b) of FIG. 3).

In order to calculate the motion of each part, in the motion analysis of the present invention, Newton's equation is used to obtain the linear acceleration of the center of gravity, and Euler's equation is used to obtain the angular acceleration of the center of gravity. When the angular acceleration is obtained, these are integrated to obtain the velocity and further integrated to obtain the position.

In the sixth step "application of constraint conditions", two physical constraint conditions, that is, the mutual connection relation of each part of the human body and the range of joint movement are checked for the calculation result of the motion of each part. This process starts from the basic part, and the position and arrangement direction of each of the lower parts (for example, the part away from the center of the human body) are sequentially checked. Here, two checks are made as to whether or not the lower part is always connected to the upper part (for example, a part close to the center of the human body) and whether or not the motion of each joint exceeds a predetermined range. To do.

As a result, when the lower part is not connected to the upper part as shown in FIG. 4A, the lower part is moved in parallel so that the lower part is connected to the upper part. If the movement of each joint exceeds a certain range, the joint is rotated so that the movement of the joint is within the range, and corrected to a natural posture as shown in FIG. 4 (b). To do.

In the seventh step, "application of inverse dynamics", the Lagrange equation expressing the relationship between force and motion is used to calculate the force generated at each joint of the body from the motion data.

When designing a new motion, if the satisfactory result is not obtained, the steps 5 to 6 are repeated to design the new motion interactively with the computer screen display. To do.

In the eighth step, "display motion on screen", a new movement during or after the design is displayed on the screen. The position of the center of gravity of the human body displayed at that time and the direction of the generated force can be combined and displayed on the human body. By doing so, the displayed action can be displayed more concretely. .

In the present invention, since a simple linear regression algorithm is used as shown in FIG. 1, the amount of calculation for calculating the inverse dynamics is O (n) which is a function of n. This makes it possible to adapt to all motions of the human body model having n joints, and by giving the values of X, Y, and Z at the time t to each part of the three-dimensional joint model, it is physically the same as the real world. The movement can be reproduced inside the computer.

By such so-called inverse dynamics, a rational and complete force combination can be obtained. Also, it is impossible for animators to find a perfect force design without inverse dynamics. In the present invention, the individual orientations of the body parts change beyond the limits of their joints, and the individual positions of each body part are adapted to the physical constraints of the body.

The motion of the human body thus obtained is natural because the lower part of the human body is always connected to the upper part, and the motion of each joint does not exceed the predetermined range. It is possible to add meat as it is and display it three-dimensionally and realistically.

Further, according to the present invention, the movement of the human body can be designed interactively using a computer without trial and error or the intuition of an animator.

In the embodiments described above, only the case where the human body is targeted has been described, but this method can be applied to the animal body in the same manner.

[0044]

As described above, the dynamics applying method of the present invention comprises two steps of analyzing the basic movement of the actual human and creating a new movement. Then, the basic human motion analysis proceeds in three stages: human body model creation, actual motion input, and input motion analysis, and new motion creation involves dynamics, constraints, and
Go through the three stages of inverse dynamics. At the stage of dynamics, the human body is divided into 50 independent parts separated by joints, and the motion of each part is calculated by using Newton's equation and Euler's equation and separated from the motion of other parts. . In the constraint stage, the mutual connection relation of body parts and the range of joint movement are checked. In the inverse dynamics stage, the force that produces the new motion modified by the constraints is calculated. In that case, the total calculation amount is O (n), and the calculation amount is significantly reduced as compared with the conventional O (n ⁴ ).
Therefore, according to the animation creating method of the present invention,
It solves the problem of calculation in animation by dynamics so that it can be applied to actual animation work well, and real-time feedback can be done. In order to calculate the movement of each part of the human body, the linear acceleration of the center of gravity is calculated by using Newton's equation and the angular acceleration of the center of gravity is calculated by using Euler's equation. Also, the force applied to those centers of gravity can be obtained and displayed, and similarly, the position of the entire center of gravity and the force exerted on the center of gravity can be obtained and displayed. Further, it is possible to realize the creation of a three-dimensional and realistic movement of the human body which is not a line drawing and is smoothly fleshed out. Furthermore, the animator can see the human body of the model from various viewpoints on the display screen, and can move or rotate the body parts in an interactive manner with the computer screen. Therefore, the animator can accurately grasp the correspondence between the picture and the human body of the model. The knowledge in the conventional animation creation method includes the basic data of the movement of the human body and the constraint conditions that define the range of the movement of each joint, but this knowledge is obtained by the intuition of the animator.
The knowledge in the animation of the present invention refers to actual dynamic parameters obtained by analyzing actual human movements. Therefore, the movements created from this knowledge can be scientific, reliable and realistic. Recently, object-oriented models have been used in various fields, and according to the object-oriented method, the user interface becomes a direct operation model. In this direct operation model, the image displayed on the screen reacts with the object, so that the object can be directly handled or operated in the object space by using the animation creating method of the present invention. In addition to the animation creating method, the motion analyzing method and the motion designing method shown in the present invention are applied to applications such as programming of industrial equipment such as robots or numerical control devices, learning of operations in sports or entertainment, training of animals, etc. It will be very useful.

[Brief description of drawings]

FIG. 1 is a flowchart of an animation creating method.

FIG. 2 is a control graph of an example of a force acting on a joint.

FIG. 3 is a schematic diagram of calculation by dynamics.

FIG. 4 is a schematic diagram of calculation by inverse dynamics.

Claims (1)

[Claims] 1. A method for creating an animation for reproducing the motion of a human body model on a computer screen, wherein the minimum unit of the human body model is composed of parts created by breaking down the human body into minimum units of motion divided by joints. Inputting to the database data describing the inherent properties of each other, the mutual relations, and the constraint conditions of the range of joint movements, and the actual movement of the human body in either frame units of video images or frame units of images. Based on the data input in the operation input step, the step of inputting the data described based on the data, the center of gravity of each part corresponding to the minimum unit of the actual human body and the human body model, the force acting on each joint, The step of analyzing and inputting the torque into the database, and the step of the human body corresponding to the basic movement of the human body model from the database Select the analysis data of the actual motion from the database, separate the new motion from other parts based on the center of gravity of each part corresponding to the smallest unit of the actual human body model, the force acting on each joint, and the torque, and the interrelationship of each part And ignoring the physical constraints of the range of motion of the joint, and applying the physical constraints of the range of the range of motion of the joint and the mutual relation of the respective parts to the calculated new motion. An animation creation method characterized by comprising the steps of repeating the process of modifying a new action, and.