KR20040036282A - A design and implementation method of markup language for integrating of motion capture data formats and apparatus thereof - Google Patents

Abstract

The present invention newly defines MCML (Motion Capture Markup Language) as a markup language based on XML (eXtensible Markup Language) as a standard format that can integrate motion capture data of different motion data formats according to the motion capture system. By converting a motion data file into a standard format defined above and then dividing it into a meaningful information unit instead of a file unit and storing it in a database, the motion capture data can be easily reused and exchanged. The present invention relates to a method and apparatus for designing, implementing, and reusing a motion capture data markup language for reuse.

Description

A DESIGN AND IMPLEMENTATION METHOD OF MARKUP LANGUAGE FOR INTEGRATING OF MOTION CAPTURE DATA FORMATS AND APPARATUS THEREOF}

The present invention relates to a method and apparatus for designing and implementing a motion capture data markup language for reuse of motion capture data, and more particularly, to integrate motion capture data of different motion data formats according to a motion capture system. Newly defined MCML (Motion Capture Markup Language) as a markup language based on XML (eXtensible Markup Language) in the standard format, and converting the captured motion data file into the standard format defined above and then a meaningful unit of information The present invention relates to a method and apparatus for designing and implementing a motion capturing data markup language that is divided into and stored in a database to facilitate reuse and interchange of motion capturing data stored in the database.

In general, the motion capture technique is one of motion generation techniques that capture the movement of a character in the real world with a special camera and apply it to a virtual character on a computer. Unlike other motion generation techniques, a motion capture technique can be reproduced as it is. There is a big advantage.

Recently, high quality motion data can be obtained by improving motion capture equipment.

However, the high quality motion data generated by the motion capturing technique is a solidified data adapted to a specific character and situation, and has a disadvantage in that it is not easy to edit or modify and reuse for other purposes.

In particular, the captured motion data uses different data formats according to the motion capturing system, and each capture system defines its own data format that can represent the captured content from simple format in the form of a segment to complex format in the form of a hierarchy. .

The problem is that developing motion software that understands all the different formats, or adding plug-ins to existing software, is both time-consuming and expensive.

In addition, the conventional system provides motion captured data in the form of a file unit library. You can, of course, store them in a database, but they are also stored file by file.

However, in order for the animator to reuse the captured motion data, it must first be able to easily search for the desired motion (eg, walk, run, dance, etc.) and sometimes select and retrieve only some frames of the entire motion. To do this, the motion capture file should be stored separately in meaningful units of information and searchable for each of these separated parts.

When animating a character or creating a game, many animators often work by dividing different motions of a character. In this case, motion data must be exchanged with each other. When using different formats, a plug-in supporting each format is required. Without it, exchange becomes difficult and it is impossible to create a new motion by synthesizing a part of motion data from several motion files.

The present invention was devised to solve the above problems, and an object of the present invention is to use MCML as an XML-based markup language in a standard format that can integrate motion capture data of different motion data formats according to a motion capture system. The newly defined and captured motion data file is converted into the standard format defined above, divided into meaningful information units instead of file units, and stored in a database, so that the motion capture data stored in the database can be easily reused and exchanged. The present invention provides a method and apparatus for designing and implementing a motion capture data markup language.

Another object of the present invention is to provide a method of designing and implementing a markup language in which motion capture data of different formats are first converted into an MCML document according to the MCML document structure and then stored in the XML database.

In addition, another object of the present invention is to provide a method of designing and implementing a markup language in which MCML documents stored in a database can be searched at any time and can be searched at an element level so that a specific motion can be easily searched. have.

In addition, another object of the present invention is to design and implement a markup language that allows an animator to search for a desired motion and then apply the motion to another character, or to synthesize several motions to perform a linked motion. To provide a method.

1 is a flowchart illustrating a method for designing and implementing a motion capture data markup language of the present invention;

2 is a flowchart illustrating a definition process for representing the human joint of FIG. 1 in detail;

3 is a table illustrating a mapping relationship between names of MCML defined according to a method of designing and implementing a motion capture data markup language of the present invention and names of ASF / AMC, BVH, and HTR file formats for representing human joints;

4 is a table showing a mapping relationship between names of BVH, ASF / AMC, and HTR files and names of MCML for indicating header information of motion capture data according to the present invention;

5 is a document showing header information expressed in MCML according to the present invention;

6 is a table showing a mapping relationship between names of ASF / AMC, BVH, and HTR files and names of MCML for indicating joint layer information of motion capture data according to the present invention;

7 is an MCML document showing an example of the structure of the articular body of FIG. 6;

8 is a table showing a mapping relationship between names of ASF / AMC, BVH and HTR files and names of MCML for indicating motion information of motion capture data according to the present invention;

9 is an MCML document showing an example of the character motion of FIG. 8;

10 is a block diagram of an apparatus for designing and implementing an apparatus for capturing motion capture data of the present invention;

11 is a flowchart of a process for converting a motion capture data file of a BVH format into a corresponding MCML document according to the present invention.

<Description of the symbols for the main parts of the drawings>

10: motion capture system 20: MCML converter

30: database 40: MCML reconversion unit

50: MCML editor 60: motion conversion unit

70: motion retargeting unit 80: motion viewer

In order to achieve the above object, the present invention provides an ASF / AMC, BVH, HTR file format name and XML-based markup language for expressing a human joint in a motion capture data file generated from various motion capture systems. Defining a mapping relationship between names of a motion capture markup language (MCML) newly defined as; Defining a mapping relationship between names of BVH, ASF / AMC, and HTR files for indicating header information of the motion capture data and names of the MCML; Defining a mapping relationship between names of ASF / AMC, BVH, and HTR files and names of the MCML to indicate joint body layer information of the motion capture data; Defining a mapping relationship between names of ASF / AMC, BVH and HTR files and names of the MCML for indicating motion information of the motion capture data; And a process for converting, searching, editing, reprocessing, and retargeting motion capture data based on the MCML.

Another feature of the present invention is an MCML conversion unit for converting motion capture data format files, such as ASF / AMC, BVH, HTR generated from the motion capture system to MCML format document; A DB for storing the edited document as well as the MCML format document converted through the MCML converter; An MCML reconversion unit converting the MCML format document stored in the DB back into a motion capture data format; An MCML editor that edits and outputs motion capture data or an MCML document input through the MCML reconversion unit using an XML editor; A motion converter converting the edited document (contents) output through the MCML editor into a motion capture data format file such as ASF / AMC, BVH, HTR, etc .; A motion retargeting unit for supporting new motions or retargeting motions by applying previously captured motion data to a newly created character or another existing captured character using data output through the motion converter; And a motion viewer connected to the MCML converter to easily display various formats. It is in including.

Hereinafter, with reference to the accompanying drawings will be described the present invention in more detail.

1 is a flowchart illustrating a method for designing and implementing a motion capture data markup language of the present invention.

As shown, the method of designing and implementing the motion capture data markup language of the present invention includes the names of ASF / AMC, BVH, and HTR file formats for representing human joints in motion capture data files generated from various motion capture systems. And a mapping relationship between names of MCML (Motion Capture Markup Language) newly defined in a standard format as an XML-based markup language (S10), and BVH, ASF / for indicating header information of the motion capture data. Defining a mapping relationship between names of AMC and HTR files and names of the MCML (S20), names of ASF / AMC, BVH, and HTR files for indicating joint body layer information of the motion capture data, and names of the MCML. Defining a mapping relationship between names (S30), and defining a mapping relationship between names of ASF / AMC, BVH, and HTR files for indicating motion information of the motion capture data and names of the MCML (S4). 0) and a process (S50) for converting, searching, editing, reprocessing, and retargeting motion capture data based on the MCML.

FIG. 2 is a flowchart illustrating a definition process S10 for expressing a human joint of FIG. 1 in detail. The method and the name of the method of expressing a joint of the human body are examined (S11) and the step S11 investigated. Among the common ones of the mapping (mapping) between the step (S12), after the mapping step (S12) adding the non-mapped portion between each other (S13), and as a result through the steps (S11 ~ S13) Defining a standard name of (S14).

FIG. 3 is a diagram illustrating a mapping relationship between names of MCML defined according to a method of designing and implementing a motion capture data markup language of the present invention and names of ASF / AMC, BVH, and HTR file formats for representing human joints.

As shown, among the data contained in the motion capture data file to which the present invention is applied, there is data representing a joint of a human body.

Representatively, looking at the three formats of ASF / AMC, BVH, and HTR, it can be seen that the method of expressing the joints of the human body and the names of the joints are different and the levels of expression are different.

To integrate them into one, we first examine how to express human joints and names in each data format and map them to common ones (steps S11 and S12), add the parts that are not mapped to each other, and then define one standard name. (Step S13, S14).

Therefore, FIG. 3 shows the MCML tags representing the mapping relationship between the three file formats and the human joint, and this is a table for mapping between different files and MCML in a system implemented for integrating motion capture data. Used as

In other words, in the ASF / AMC file format, a method for representing a human joint uses a name for a human bone, a BVH file format uses a name for a marker position, and an HTR for a human bone. Although it uses names, it's an early format, so you have only a few names.

To integrate these files, MCML will have extended expressiveness to include all three formats based on ASF / AMC.

4 is a diagram illustrating a mapping relationship between names of BVH, ASF / AMC and HTR files and names of MCML for indicating header information of motion capture data according to the present invention.

As shown, the motion data file according to the present invention has a header data area containing additional information such as file type, version, and annotation. There are also files that do not contain header information, such as BVH. According to the present invention, the MCML provides a tag set that can include all header information of these files.

FIG. 5 is a part of a document representing header information expressed in MCML according to the present invention. Among the header information according to the present invention, many of them are not directly related to a modeled character structure or motion, and may be used in a future version. Some names are defined only.

In the MCML system according to the present invention, a tag set for header information capable of maintaining all of this information is defined, and new header information may be further defined.

FIG. 6 is a diagram illustrating a mapping relationship between names of ASF / AMC, BVH, and HTR files and names of MCML for indicating joint layer information of motion capture data according to the present invention.

As shown, the motion capture data contains information about the character being modeled. That is, the character to apply motion is defined in the same file. Of course, Acclaim's ASF / AMC motion capture data file lets you manage your character and motion separately.

The ASF file contains the character's joints and hierarchy, and initial position information. The AMC file contains the motion information of the character's movement.

The advantage of separating the character from the motion is that the same motion can be applied to the character with similar size and skeleton and the character is reusable.

However, file formats such as BVH and HTR are less reusable because character information and motion information must be in one file.

In the MCML defined according to the present invention, human joint information, header information, joint layer information, motion information, and the like can be separated, and each part can be combined to create a single file, which is highly reusable and redundant. This can help animators a lot.

7 is a part of the MCML document showing an example of the structure of the articulation of FIG. 6, wherein each bone of the human body is represented by a bone element, and a hierarchy between bones is represented by a hierarchy between bone elements.

8 is a diagram illustrating a mapping relationship between names of ASF / AMC, BVH and HTR files and names of MCML for indicating motion information of motion capture data according to the present invention.

As shown, when the motion data according to the present invention is converted to MCML and stored, it is possible to search for each frame through an XML query word such as XPATH or XQL, and at this time, the motion corresponding to each frame is searched to search for a desired motion. By storing the informed motion name, it is possible to search using this motion name.

In addition, the animator can easily apply to a new character by creating a new character and searching for the desired motion by using the motion name among the saved motion data without changing a new motion capture to apply the motion to the character. can do.

In addition, it defines the movement of each part of the body such as legs, torso, head and shoulders. The important thing is that there should be nothing expressing impossible behavior.

FIG. 9 is a part of an MCML document showing the example of the character motion of FIG. 8, wherein a frame element contains one frame information, and the frame_bone element has values for position and rotation of each joint in the frame. have.

10 is a block diagram of an apparatus for designing and implementing the present invention motion capture data markup language.

As shown, the apparatus for designing and implementing the motion capture data markup language of the present invention converts motion capture data format files such as ASF / AMC, BVH, and HTR generated from various motion capture system 10 into MCML format documents. The MCML converter 20, a database (hereinafter referred to as 'DB') 30, which stores the edited document, as well as the MCML format document converted by the MCML converter 20, and the DB 20 MCML reconstruction unit 40 for converting the MCML format document stored in the back to the motion capture data format again, and outputs by editing the motion capture data or MCML document input through the MCML reconversion unit 40 using an XML editor The MCML editor 50 and the motion conversion unit 60 for converting the edited document (contents) outputted through the MCML editor 50 into motion capture data format files such as ASF / AMC, BVH, HTR, and the like. The data output through the motion conversion unit 60 is Motion retargeting unit 70 for supporting new motion or retargeting motion by applying previously captured motion data to a newly created character or another existing captured character, and connecting to the MCML conversion unit 20. And a motion viewer 80 to easily display various formats.

11 is a process flow diagram for converting a motion capture data file in BVH format into a corresponding MCML document in accordance with the present invention.

As shown, the MCML converter 20 which is in charge of the conversion between the motion capture data file and the MCML document is composed of a Motion Capture Data Parser and an MCML Wrapper.

Therefore, when a motion capture data file is input from the motion capture system 10 (e.g., ① *. Bvh), first the tokens of the file are analyzed by the Motion Capture Data Parser, and then they are transferred to the MCML Wrapper. (②).

The MCML Wrapper uses these tokens to create an MCML document. The bone naming mapping table is used at this time. This mapping table is a table having mapping information on joint names between the MCML shown in FIG. 3 and each motion capture data (③).

In this manner, the original data of the motion captured data can be reused to not only provide the motion captured data in the form of a library in file units, but also to store the data in the DB 30 and store the data in file units ( ④, ⑤).

In addition, in order to reuse the captured motion data, the animator can easily search for the motion he / she wants, and in order to make it possible to select and search only some frames of the entire motion, the motion capture file is a meaningful information unit. The ability to save them separately allows for easy retrieval of these separate parts.

As described above, the present invention enables the conversion and integration of motion capture data between different animators in a character animation or game production system using motion capture data information.

In addition, the production time is greatly reduced to combine the stored motion clips to create a new motion.

In addition, the present invention can efficiently retrieve the stored motion data, it is possible to combine the data to create a new motion or to perform the same motion by applying the stored motion to other characters, high reusability of the data You can make animation more efficiently.

In addition, by providing a motion capture viewer, various formats can be easily viewed through the viewer, and can be converted to other formats to maximize user convenience.

In addition, the present invention provides a motion capture data integration technology that can be effectively used in terms of error correction and complementation of motion capture data, and data integration as a base technology element of the human body motion control part. This has the effect of providing a technology foundation that can be flexibly applied to problems in dimensional animation.

Claims (9)

Motion Capture Markup Language (MCML), defined in ASF / AMC, BVH, HTR file format and XML-based markup language to express human joints in motion capture data files generated from various motion capture systems. Defining a mapping relationship between the names of (S10); Defining a mapping relationship between names of BVH, ASF / AMC, and HTR files for indicating header information of the motion capture data and names of the MCML (S20); Defining a mapping relationship between names of ASF / AMC, BVH, and HTR files and names of the MCML for indicating joint body layer information of the motion capture data (S30); Defining a mapping relationship between names of ASF / AMC, BVH and HTR files and names of the MCML for indicating motion information of the motion capture data (S40); And A process process (S50) for converting, searching, editing, reprocessing, and retargeting motion capture data based on the MCML; and designing and implementing a motion capture data markup language, comprising: . The method of claim 1, Defining process (S10) for expressing the human joint, step (S11) to investigate the method and name for representing the joint of the human body; Mapping common ones of the ones irradiated in the step S11 (S12); Adding portions (S13) which are not mapped to each other after the mapping step (S12); And Defining one standard name using the steps S11 to S13 (S14); Motion capture data markup language design and implementation method comprising the. The method of claim 1, The file type included in the motion data file is included in the motion data file in the MCML document (S20) of defining a mapping relationship between names of BVH, ASF / AMC, and HTR files and names of the MCML for indicating the header information of the motion capture data. Design and implementation method of the motion capture data markup language, characterized in that it provides a tag set including all of the header data area containing additional information, such as version, comment. The method of claim 1, The new header information may be additionally defined in the MCML in the step S20 of defining a mapping relationship between the names of BVH, ASF / AMC, and HTR files and the names of the MCML for indicating the header information of the motion capture data. A method of designing and implementing a motion capture data markup language. The method of claim 1, In the MCML, the human body joint information, header information, joint layer information, motion information, and the like can be separated, and each part can be combined to create a single file. A method of designing and implementing a motion capture data markup language, comprising constructing a relationship. The method of claim 1, In the MCML, each frame is separated separately, and joints of one frame are also separately separated so that it is easy to extract a specific frame or specific joint motions, and change, combine, or separate the motion capture data markup language. Design and implementation method. The method of claim 1, When the motion capture data is converted into MCML and stored, the motion capture data markup language designation method is characterized in that it can be searched for each frame through an XML query language such as XPATH or XQL. An MCML conversion unit 20 for converting motion capture data format files such as ASF / AMC, BVH, and HTR generated from the motion capture system 10 into MCML format documents; A DB 30 for storing the edited document as well as the MCML format document converted by the MCML converter 20; An MCML reconversion unit 40 for inversely converting an MCML format document stored in the DB 20 into a motion capture data format; An MCML editor 50 for editing and outputting motion capture data or an MCML document input through the MCML reconversion unit 40 using an XML editor; A motion converter 60 for converting the edited document (contents) output through the MCML editor 50 into a motion capture data format file such as ASF / AMC, BVH, HTR, etc .; The motion retargeting unit supports to create a new motion or retarget the motion by applying the motion data already captured using the data output through the motion converter 60 to a newly created character or another existing captured character. 70; And A motion viewer 80 connected to the MCML converter 20 to easily display various formats; Design and implementation of a motion capture data markup language comprising a. The method of claim 8, The MCML converter 20 is designed and implemented in the motion capture data markup language, characterized in that consisting of a Motion Capture Data Parser and MCML Wrapper.