[go: up one dir, main page]

WO2001011570A1 - Systeme d'animation personnalisable - Google Patents

Systeme d'animation personnalisable Download PDF

Info

Publication number
WO2001011570A1
WO2001011570A1 PCT/US2000/022232 US0022232W WO0111570A1 WO 2001011570 A1 WO2001011570 A1 WO 2001011570A1 US 0022232 W US0022232 W US 0022232W WO 0111570 A1 WO0111570 A1 WO 0111570A1
Authority
WO
WIPO (PCT)
Prior art keywords
motion
animation
server
blending
user
Prior art date
Application number
PCT/US2000/022232
Other languages
English (en)
Inventor
Oshri Even-Zohar
Original Assignee
Motek Motion Technology, Inc.
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Motek Motion Technology, Inc. filed Critical Motek Motion Technology, Inc.
Priority to AU67703/00A priority Critical patent/AU6770300A/en
Publication of WO2001011570A1 publication Critical patent/WO2001011570A1/fr

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T13/00—Animation
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2200/00—Indexing scheme for image data processing or generation, in general
    • G06T2200/16—Indexing scheme for image data processing or generation, in general involving adaptation to the client's capabilities

Definitions

  • the present invention most generally relates to computer animation, and more specifically to an animation system that blends motion capture data into a customizable work incorporating motion capture data that produces animation that is much closer to real life movements.
  • Animation generally refers to the combination of graphics with the dimension of time. When single images are sequenced together, it gives the appearance of movement. Prior to computers, animators drew hundreds of images to represent a single cartoon movement. Computers now automate the drawing of the images to some extent and aid in manipulating the images to portray certain movements.
  • Computer animation is well known in the art, and the animator uses the processing power of the computer to create images and motion sequences that represent the desired animation. While it is relatively simple to create jagged and unrealistic movements, generating interactive sequences of animation that follow natural human/animal motion is difficult to create. To produce quality computer animation, it is necessary to study the motions of the object being represented and form an animation program on the basis of the analysis. Animation that perform the motions of a human require observation by the animator. On the basis of the observations, a program for producing the same motions by the computer model is formed. But, observations are subjective and based on the understanding of the animator. Therefore, the animator's representation always incorporate some deviation from reality.
  • Keyframing has been used to produce animation that involves identifying the position of body parts and interpolating between the positions of the body parts in successive key frames to generate the intermediate positions of the body parts over the time interval of the animation.
  • Forward kinematic models have generally been discussed in relation to the interpolation techniques.
  • This concept of keyframing builds motion sequences frame by frame and is fairly costly and time consuming.
  • this method requires expertise to generate quality realistic motions.
  • IK Inverse kinematics
  • the IK methodology is based upon complex matrix manipulations that require significant amounts of processing to determine the joint angles associated with limb movements.
  • the amount of processing increases as a cubic of the number of joints involved, translating into lengthy processing times when numerous joints are involved.
  • a human being has more than 200 degrees-of-freedom, which makes it practically impossible to use matrix-based inverse kinematic methods to interactively animate any realistic human or animal-like character in real time.
  • matrix-based IK methods may not even work on certain joint configurations known as singular configurations.
  • the prior art notes that with multiple limb and joint movements, the end result of the computations will not appear natural.
  • prior IK methods must converge to a solution before the results of the computation can be used for the animation.
  • Partially computed results causes unstable, jerky and oscillatory motion of the limbs, with large positioning errors.
  • Animation based on keyframing and motion capture is also not highly successful with interactive applications, such as video games and virtual reality simulators.
  • the virtual world of interactive video gives a user total freedom within a three dimensional (3D) space.
  • the animator typically employs existing animation and works within the realm most familiar, producing animated characters that all move in like manner. Besides resulting in poor quality and non-original movement sequences, the existing technology is time consuming and laborious.
  • the frames are created for each movement, and there must be numerous frames with slight movements or the resulting animation appears to move in a jerky fashion.
  • animators usually limit the number of moving parts.
  • life-like movement is usually the result of multiple joint and limb movement and using a lesser degree of movement results in an unnatural animation.
  • Behavioral animation refers to movements generated in response to commands and interaction with the environment. This method of animation relies more heavily on computer processing to determine the movement sequence to accomplish the command. High-level goal-based behaviors are typically made possible by using a combination of forward kinematics and inverse kinematics.
  • One advantage of behavioral animation is that the movements can be created on-line and modified in real-time.
  • Dynamics provides another method of motion analysis.
  • Dynamics analyzes motions of objects based on the relation between movements and forces, as compared to kinematics, which provides motion analysis in terms of positions, velocities and accelerations.
  • Dynamics allows more complex behavioral or animated results.
  • computer analysis of motions utilizing dynamics requires data on parameters such as the moments of inertia, the centers of gravity, joint friction and muscle/ligament elasticity of the moving body being represented by the animation.
  • Dynamics motion analysis also requires complex dynamics equations of multiple and simultaneous differential equations.
  • PCT/US97/05513 '513 is an animation system involving online computational methods for coordinating limb movements of articulated characters by solving associated forward and inverse kinematics problems in real time subject to multiple goals and constraints.
  • the '513 invention determines how the joints of an interactive character should rotate as a function of time to produce goal-directed limb movements.
  • the '513 invention uses an on-line successive approximation method for solving inverse and forward kinematics problems concurrently, taking into account multiple goals and constraints.
  • the '513 invention is capable of producing fully interactive goal-directed behaviors in real-time for three-dimensional articulated characters with multiple limbs through the simultaneous satisfaction of position, alignment, posture, balance, obstacle avoidance, and joint limitation constraints.
  • the '513 invention performs separate calculations at each joint using goal-based motion primitives by a process which produces a set of joint angle movements that attempt to satisfy a higher-level goal, such as a positioning, alignment, posture, balance, obstacle avoidance, or joint limitation constraint.
  • a geometric successive approximation process simplifies the problem of limb coordination.
  • One object of the present invention is an animation system that utilizes custom blending of motion sequences in real-time.
  • the blending uses motion sequences from comparatives of a motion capture libraries as a starting point and customizes the sequences for the desired effect.
  • a further object is a system that uses inverse kinematics for linear computations of rotation (degrees) of all limbs. Forward dynamics are used to calculate a blend per limb that considers the temporal acceleration/velocity at any given frame of the source and any given frame of the destination. The blending results in a dynamic velocity based blending.
  • An object of the invention is a customizable animation system that provides network access to a database of recorded motion sequences and a means for customizing motion sequences over the network.
  • Network in this context is defined any Internet, Intranet, and wireless interconnection between computer systems.
  • An additional object is to provide access anywhere and anytime to the database and the tools required to customize motion sequences in real time.
  • a further object of the invention is a customizable animation system that permits a user to pay only for the sequence generated. Payment per frame or per motion sequence provides an inexpensive means of authoring new motion sequences at a fraction of the cost of other alternatives.
  • the concept of the blending is the capacity to predict the correct bio-mechanical behavior of the human body, as long as it is target driven. If for example you are sitting on a chair, and you decide to get up, it is possible to predict the vector of your perceived trajectory. There are an infinite number of ways to up from the sitting position, however, all the motion variances are created to facilitate the optimal force vector that gets your body to an upright standing position.
  • the process of going from a seated position to a standing position can be broken down as follows according to the mental processes: 1) Observing obstacles (floor space, nearby objects) and other objects to evaluate distance and assertive elements such as side arms. 2) Comparing an internally generated motion model of the perceived trajectory of getting up, relying on learned motion macro that is learned at an early age by trial and error. 3) Building an anticipatory force vector and begin activating muscle clusters leading to the buildup of perceived vector of optimum for the required standing motion.and 4) Evaluating the next macro (eg; what to do after standing) before committing to the perceived vector. Committing to the 'standing up macro' and continuously evaluating the performance of the body.
  • the underlying premise of the behavioral blender of the present invention is that every human macro can be computed.
  • the two main modules of the present invention are a motion capture library (as compared to the motion memories and 'learned' macros) and two sets of rules.
  • the two rule sets are internal and external. Internal rules refers to the kinematic skeletal bio-mechanical conditions of the bones, muscles, etc... .
  • the external rule set refers to the environmental conditions such as gravity, objects, and other dynamic forces.
  • the blender functions by first implementing the internal basic environment that allows blending between two or more sets of motion capture data, subject to the correct bio- mechanical behavior (eg: reaching the fulfillment of a set of conditions relating to marker placement based on skeleton segment length and other variables.
  • a predictive algorithm is implemented to handle 'perceived' optimal motion vector.
  • the predictive algorithm offers the user the "most natural" blend interval in time.
  • the environmental rule sets are implemented, facilitating realtime simulations of the velocity / accelerations variances in response to a changing conditions in the environment. Those Simulations make use of Lagrange equations in combinations with the (above discussed) velocity base blends.
  • the present invention has several components, including a motion database, blender software, custom user interface.
  • a motion database blending algorithm and template matching as well as a detailed description of motion capture in general is incorporated by reference from the published application PCT/US99/21246 to the same applicant.
  • the motion capture database comprises an entire full body setup. All the compensatory motions of the body are captured while isolating one body part for the eventual blends.
  • the database contains many general or standard motions such as walking, running, sitting, standing, falling, recoveries, gestures, etc... .
  • the blender software computes the blends between any matching elements of the motion databases in a manner that complies with the over-riding bio-mechanical laws. Blends that result in motion that are not natural or are otherwise impossible for the human body to execute are extrapolated to fit a correct behavior.
  • Another embodiment of the present invention is through the use of blender software using genetic algorithms. Instead of looking at blending motion capture databases according to bio-mechanical laws, parsing those motion sequences through a genetic algorithm and presenting multiple possibilities of variations of thew original motions. The operator can then choose the preferred 'mutation' and parse that sequence through same algorithm to create a new set of variations. Each new selection introduces a new set of mutations for operator selection.
  • the custom user interface is a GUI interface that allows full access of all subsets and blending methods, or as otherwise prescribed by agreement or license.
  • the present invention is an Internet product that is downloadable for implementation.
  • the motion capture libraries are also accessible via the Internet. A user can obtain any range of motion capture data and then use the software of the present invention to custom tailor the animation for the desired effect.
  • the blender can create any subset between those two data sequences. It is not as simple as an interpolation between data sets - the over-riding laws of bio-mechanics must be accounted for in creating the custom set. As a person walks, there are many forces at work and moving parts. Changing one element of the mix affects all the other elements.
  • the motion capture data sequences are obtained using sensors attached to various parts of the body.
  • the movements can be plotted a three dimensional space.
  • the blending between two sequences is not a simple interpolation between the plots because the actual trajectory and bio-mechanical laws dictate the outcome.
  • Custom motion capture also exists, but involves extraordinary time and expense to create a sequence for only a particular movement.
  • Clip art works for generic images or animation, such as famous landmarks, but does not permit custom tailoring. Cost advantages of a clip art library that allows you to customize any work create an original work. For example, a walking sequence can be extracted from the library. The user can then custom-tailor the swinging arm movement. The user may then introduce a swinging head motion - which in turn affects the swinging arm movement and the walk. Every tweak by the user creates a new and original work.
  • Another method of customizing animation is through genetic algorithms. It is a separate way to look at blending where mutations are created depending on the input selections. Mutations of the motion capture data get generated from the behavioral data and a set of different mutations are created for each behavioral alteration. Applying genetic algorithms to variances in motion capture data is another object of the present invention and will yield very different results from the blending.
  • the blender algorithm uses forward and inverse kinematic equations and Lagrange equations to calculate and determine the joint torque and joint forces at each degree of animation.
  • the human body is susceptible to over 200 degrees of freedom, in reality, the number is closer to 40 for realistic applications.
  • the present invention also describes the use of genetic algorithms the necessary equations and calculations that are modeled in the algorithms for the present invention.
  • the present invention accomplishes that which is described in the '471 application as "practically impossible” due to the of computationally intensive matrix inverse calculations.
  • the present invention allows inverse kinematic methods to interactively animate any realistic human or animal-like character in real time using both forward kinematics and inverse kinematics.
  • the '471 invention does not use matrix inverse calculations and has computational requirements that scale linearly with the number of joints. Scaling linearly with the number of joints enables the '471 invention to interactively animate moderately complex articulated characters in real time using commonly available personal computers.
  • the '471 invention is an iterative computational procedure that always produces somewhat meaningful joint motion (with no singularity conditions) after each iteration of the computation. It is not necessary to wait for the '471 process to converge to a minimum error condition.
  • the results of the calculation can be used 'anytime", i.e., within the constraints of a real time animation system.
  • the '471 methods tolerate more tracking error, i.e., greater position, velocity or acceleration error, than matrix inverse methods, but in return provide more natural motion, and faster calculations.
  • the present invention operates in real-time with less tracking error.
  • the '471 animation techniques are interactive and others are not.
  • the '471 techniques of keyframing and motion capture produce fixed, i.e., scripted and non- interactive motion data.
  • the present invention encompasses more interactive elements.
  • the present invention integrates interactive motion data with interactive control, which is not disclosed or mentioned in the '471
  • the demand for motion capture has been growing exponentially. Unfortunately it has been too expensive and too laborious for most applications.
  • the browser of the present invention allows any person to be an animator and blend motion sequences on the high speed servers of the Internet based product. When the sequence is completed, the user simply pays for the 3D data and downloads the data to the users workstation.
  • the software has been benchmarked and works at 15fps on a 56k modem.
  • a significant advantage of the present invention is the implementation and method of incorporating the system.
  • the present invention has a database of motion capture data that is accessible on-line over an intranet or the Internet.
  • the blending algorithms either inverse kinematics or genetic algorithms, can be resident on the user's computer system either from purchasing the software or downloaded over the Internet.
  • a further feature of the present invention is to allow on-line and real-time manipulation of the motion capture library on-line. An individual license can be purchased for a particular animation and a new work created by a user on-line, making this a cost-effective and speedy transaction.
  • the present invention allows the user to choose and "blend" motion sequences including individual body parts, until the correct effect is created.
  • the end result is infinitely more flexible than "canned” motion capture sequences, and extremely cost-effective as compared to studio production.
  • the present invention brings big production motion capture to projects lacking big production budgets, and makes the motion capture animation project faster, easier, and less costly to complete.
  • This system provides a breakthrough in price/performance for 3D animators. Lifelike sequences are built in a fraction of the normal time, at approximately one-tenth the cost of traditional studio production
  • the present invention allows for each human body part motion to be altered independently.
  • the user manipulates a huge library of captured data to produce unique sequences.
  • the blending tools also marries individual sequences to produce smooth, lifelike movements, employing inverse kinematics.
  • the present invention is accessed through the Internet and is available anywhere and anytime.
  • the data sets created by the present invention can be directly exported into a wide range of professional 3D packages including products from SOFTIMAGETM, DiscreetTM, AliasIWavefrontTM and Side-EffectsTM.
  • One of the unique features of the present invention is that is allows on-the-fly data sets viewed through the "blender" window. Selected sequences and movements are reviewed in real-time before you commit to purchase and download.
  • the algorithms continually enforces bio-mechanical rules so that sequences maintain the ultimate in natural physical motion. For even more creativity of movements for non-human characters or to deliberately portray unreal human motion, the rules can be bypassed.
  • the motion capture data bank is a growing depository of motion capture data. As more data is captured the database will continue to expand.
  • Another object of the invention is a simple-to-understand interface, allowing the user to quickly select and manipulate motion sequences, then fine tune each body part using slider bars.
  • the graphical user interface permits an untrained user to begin from basic building blocks and customize these animations according to individual preferences.
  • the user must access the blender download site and download the client browser.
  • the user selects one or more motion sequences from the data library. Using slider bars, the user modifies and fine tunes individual body parts and manipulates the motion sequences until the desired result is achieved.
  • the user selects a format for exporting the data.
  • the user selects from the various payment methods and chooses the format for downloading the data.
  • the user imports the data to the individual animation tool or environment.
  • FIG. 1 System overview showing elemental components
  • FIG. 2 Simplified Flow Chart
  • FIG. 3 Client Browser
  • FIG. 4 Viewing mode of a motion sequence (source or blend) which shows the original marker placement on the body.
  • FIG. 5 Viewing mode of a motion sequence (source or blend) which shows the correct skeletal structure under the markers, including the right offsets and joint rotation values.
  • FIG. 6 Viewing mode of a motion sequence (source or blend) which shows how a 3D deformation model behaves in motion.
  • FIG. 7 Constant decay distribution for leaking.
  • FIG. 8 Linear decay distribution for leaking.
  • FIG. 9 Exponential decay distribution for leaking.
  • FIG. 10 Cutoff decay distribution for leaking. DESCRIPTION OF TFIE PREFERRED EMBODIMENT
  • FIG. 1 shows a top-level block diagram of the elements of one embodiment of the invention.
  • a main server 10 that acts as the administrator of the system.
  • the various other components are interconnected to the main server 10 through a network, o wherein a network is defined herein as any means of communicating between devices, including Internet, Intranet, and wireless means.
  • Motion capture library databases 20 are connected to the main server 10 and house the various captured data.
  • the captured data represents the various motion sequences and 5 actions recorded using the optical and magnetic markers as is known in the art as well as a growing volume of motion sequences that may have been blended.
  • a human template 30 is accessible to the main server 10 and is used in conjunction with a template matching algorithm as is described in the prior art and published PCT o application PCT/US99/21246.
  • the blend engine 40 that encompasses the UNICA algorithm 50.
  • the algorithm 50 uses forward and inverse kinematics, Lagrange equations and even genetic algorithms to blend the data and produce 5 the realistic motion sequences.
  • the blender functions by implementing the internal basic environment that allows blending between two or more sets of motion capture data, subject to the correct bio-mechanical behavior.
  • the algorithm 50 is implemented to handle 'perceived' optimal motion vector.
  • the source's velocities and accelerations are extrapolated, the reverse happens with the destination's acceleration and velocities.
  • the predictive algorithm offers the user the "most natural" blend interval in time.
  • the environmental rule sets are implemented, facilitating realtime simulations of the velocity / accelerations variances in response to a changing conditions in the environment. Those simulations make use of Lagrange equations in combinations with the velocity base blends discussed herein.
  • a security layer 60 tries to protect the main server 10 and the other components from various cybercrimes.
  • a plurality of proxy servers 70 provide the communication access between the outside world and the system of the present invention.
  • the proxy units are basically gateways for connecting users to the main server 10, and the more proxy servers 70 the more accessibility.
  • the proxy servers use login names and passwords to provide secure access once the user 120 has completed an application and downloads the client browser 1 10.
  • this invention is accessible anywhere and anytime via the Internet or other network means.
  • a registered user 120 accesses the client browser 1 10 and gains access to the main server 10, the libraries 20 and the blending system 40, 50.
  • the user can work on a motion sequence and even save a sequence for later use, but the actual motion capture sequence is not available until the user 120 goes through the payment layer 90 and purchases the sequence.
  • the payment layer is handled by a 3 rd party online transaction service.
  • the compressed blended data 130 can be stored by the user 120 in a usable form.
  • a production server 80 is a specific server dedicated to the registration of new users. It provides a secure interface for unregistered users to access the system in a limited environment. A new user accesses the website 160 and is given limited access through a filter 150 block that requires registration for normal access to the system. In the preferred embodiment the client computer must have sufficient resources to successfully access and use the client browser and system. The minimum system requirements are shown in Table A.
  • FIG. 2 shows a flow chart representing the basic steps in the system.
  • the first step 200 is to access the website and join the UNICA user group.
  • an application form needs to be completed in order to receive a startup password.
  • click the "Submit Form” button click the "Submit Form” button, and review and accept the License Agreement.
  • an authentication email will be sent to the email address supplied on the email.
  • the information sent to the user includes a user password and links to the UNICA client software, 3D application converters, and free motion data samples.
  • the next step 220 is to login to the system. The user returns to the website and a prompt appears requesting the User Name and Password. At the prompt, fill in the name from the application form and the password received in the authentication email. The best way to enter the correct password is to highlight your password in the message by left clicking, holding the button down, and dragging the cursor over it. When your password is highlighted, press CTRL-C to copy it to your clipboard. Changing the password is recommended.
  • the client application 220 To install the client application 220 onto the local system it is necessary to download the file.
  • the client application runs on several platforms, and the user selects the appropriate platform, such as the Windows platform.
  • the file is compressed in Zip format, and should be downloaded and saved to a location that will be remembered.
  • the file size is only half a megabyte, so it should not take that long to download.
  • the related files such as the data converters for the most popular 3D applications, including Softimage, Maya, 3D-Studio Max, and Lightwave. With each of these is a Text file describing the command line parameters needed to run the converter.
  • UMF file or UNICA Motion File, as a free sample to use and test out. The user can download one or all of the converters, depending on what 3D application is utilized.
  • connection to the UNICA server is started automatically as soon as the client application is invoked.
  • the application detects existing connections, but if no connections are detected active, the application will try and start an Internet connection. Once the connection is made, a window will appear, prompting you to type in your user name and password. Use the user name created when filling out the application form and the password received in the authentication email. At this point it is advised to verify the personal data and consider altering the password.
  • the UNICA browser will appear 230.
  • the user can select one or more motion sequences from the data library.
  • the user can then customize the motion sequences using the browser controls 240.
  • slider bars the user modifies and fine tunes individual body parts and manipulates the motion sequences until the desired result is achieved.
  • the user can save the data for later usage 250.
  • the data is saved locally in a format that is only readable by the client application and browser, but allows the user to pick-up from an earlier customization job.
  • the user selects a format for exporting the data 260.
  • the user selects from the various payment methods and chooses the o format for downloading the data 260.
  • the browser of the present invention provides the user interface and is shown in
  • the blending palette has numerous elements, including the Tool Bar 300, Blending Area 310 , Status Bar 320, Sequence Selection 330, Viewer Play Controls 340, Frame Slider 350, Source List Icons 355, Bandwidth Usage Meter 360, Viewing Area 370, Total Frame Count 380 and Get 3D Data 390.
  • the Tool Bar 300 (located in the upper left-hand corner) is the area where the user can change general UNICA browser settings. The menu breakdown is as follows: FILE, CONNECTION, MOTION, VIEW and HELP.
  • Hidden files Displays hidden files in the current location.
  • Cancel Stops the save process and brings you back to the Browser.
  • Get 3D data - This option allows the user to enter the sequence name.
  • a payment/account summary screen is presented and the user can proceed to making payment on the new creation or can cancel out of the payment segment.
  • Skeleton - A Viewing mode of a motion sequence (source or blend) which shows the correct skeletal structure under the markers 420, including the right offsets and joint rotation values FIG. 5.
  • a default sequence is the natural walk.
  • the user can simply click on the Sequence Selector 330 in the upper right-hand corner of the browser window. Sequences are continuously being added to the library and posted on the server. Sequences such as wrestling, golfing, fight scenes, and animals provide an array of possibilities for the creative animator.
  • the Viewing Area 370 (located in the left-hand side) is the area that allows the user to change general browser settings.
  • the Viewing Area menu breakdown is as follows:
  • Viewer Play Controls 340 (located below the Viewing area) allow the user to play, pause or stop a blending loop. This is useful if a single frame is to be focused on during the blending process.
  • Frame Slider 350 (located to the right of the Viewer play controls) shows the frame id currently being viewed in the Viewing area. Frames can be individually selected by simply o dragging this slider to the left or right.
  • Total Frame Count 380 (located to the right of the Frame Slider 340) shows the total number of frames in this blending sequence. The number shown here will vary depending on the sequence that is being blended. 5
  • Source List Icons 355 By clicking on any of the Source List Icons 355 user is requesting from the server to add the selected source to the display list. If a source is already open (default head source) a blend between the default and the new selected source will be seen. To view sources without any blending make sure that you are viewing one source only, (e.g.: remove the 0 default source first.)
  • the Status Bar 320 provides process information to the animator. Such items displayed here are: number of frames received from the UNICA server, starting file paths, etc...
  • the Blending Area 310 allows the user to make changes to the data in the blending process. This is done using sliders and blend types in the Blending Area 310.
  • the model is broken down by area of the body (i.e. head, shoulders, hands, feet, etc.). Within each area of the body a further breakdown can be accomplished.
  • the slider bar and value display are the key interfaces in the blending process.
  • the user can change the values of each characteristic by clicking or dragging on the associated slider bar. Note that the numerical value to the right adjusts accordingly.
  • the Blending Type controls the amount of "leakage" of motion from the active (currently under editing) body part, to the rest of the body. Since all the sources are made from full body motion capture sequences, the amount and the path of the leakage will create different blends.
  • Constant refers to a full leakage, similar to average mix and is illustrated in FIG. 7.
  • the leakage icon 450 shows the representation for constant decay under the blending type. Linear Decay, wherein depending on the active body part, motion will "leak" down the skeleton hierarchy so that the body parts furthest away from the active body part will be compensated minimally is shown in FIG. 8.
  • the leakage icon 450 shows the representation for linear decay under the blending type.
  • the leakage icon 450 shows the representation for exponential decay under the blending type.
  • Cutoff equates to No leakage at all and is shown in FIG. 10.
  • the Cutoff can cause unrealistic results, and be used in developing unrealistic modes or "clown modes".
  • Using cutoff blend type for multiple sources for the same body part, will result in linear blend between them.
  • the leakage icon 450 shows the representation for cutoff under the blending type.
  • a Bandwidth Usage Meter 360 (located at the lower right-hand corner of the browser) is a gauge of how much of the available bandwidth is being used by the browser during the blending process. Acceptable levels are shown in green. As less bandwidth is available, the meter indicator turns yellow. At poor levels, the meter turns red. The Bandwidth Usage meter is located in the lower right-hand corner of the browser, next to the "Get 3Ddata" button 390.
  • the motion capture databases of the present invention are the sources from which the motion sequences to blend are selected. This can be analogized by looking at one page from a book that has many pages, wherein this book is on a bookshelf containing many different books, and there are many bookshelves. The page in the book you are looking at currently is dedicated to one motion, generated by one character, in one speed, inside one environment. If the same character would execute the same motion in the same speed but in a different environment (for example; strong headwind), the motion will change totally.
  • the user can blend in real-time, intuitively building a full body motion sequence. And, the user can do that by working on blending motions from different body parts.
  • the body parts which are generic to the 5 present structure are: Head, Shoulders, Chest, Hips, Elbows, Knees Hands and Feet.
  • the blending type controls the amount of "leakage” of motion from the active (currently under editing) body part, to the rest of the body. Since all the sources are made from full body motion capture sequences, the amount and the path of the leakage will o create different blends.
  • the default sequence is typically a source pertaining to the Head.
  • the name of the sequence is printed next to the blend slider. At any time the user can remove, change or 0 add other sources. If you click on other body parts from the list, you should see the referring body part changing color on the moving figure, corresponding to the selected body part.
  • UPF UNICA Parameter File
  • UPF blend parameter file
  • UNICA Motion Files are the result of the blends. They contain the capture marker positions and the joint rotation values of the skeleton. Those files can be converted to formats used by most 3D Vendors. Currently in the UNICA Beta UMF output files are compatible with SOFTIMAGE 3D SP2, SOFTIMAGE SUMATRA (Beta), Maya, Lightwave, and 3D-Max.
  • the user can download the converters from the website for different operating systems, and use them locally. Installation instructions are available at the converter download page.
  • a typical converter command will look like: Umf2soft ⁇ umf filename> ⁇ output filename> (in case of Softimage files a .dsc scene file)
  • the UMF file conversion will result in a scene file containing a hierarchy of nulls representing the motion capture markers, and a skeleton containing joint rotation values.
  • the customized animation is by the frame.
  • the following table shows a purchasing methodology, as well as demonstrating the price in comparison to the thousands of dollars normally associated with customized animation.
  • the present invention in one embodiment is a method of customizing animation in realtime over a network, comprising the steps of opening a client application on a local system, establishing a network connection to a server, wherein the server is interconnected to a motion capture library and to a blending means, selecting a motion sequence from the motion capture library of the server, customizing the motion sequence in realtime using the blending means, and downloading the motion sequence onto the local system.
  • Other variations include purchasing the motion sequence prior to downloading.
  • the motion sequences can be singular or a plurality of motion sequences.
  • the downloading comprises a step of converting the motion sequences into a user selected format. Furthermore, there is a step of saving the motion sequence on a local system in a format only readable by the server.
  • the invention also has a method of purchasing animation in realtime over a network, comprising the steps of opening a client application on a local system, establishing a network connection to an animation server, selecting a motion sequence from the server, customizing the motion sequence in realtime, purchasing the motion sequence on-line, downloading the motion sequence onto the local system. Additionally, a method, wherein the step of purchasing is based upon the number of frames.
  • the invention is a customizable animation system, comprising a local computer system with a memory means and network means, an animation server with a network means, wherein the server is interconnected to a motion capture library, and wherein the server is connected to the local computer system by the network means, a means of selecting a motion sequence by the local computer system, a means of customizing said motion sequence, and a means of storing the motion sequence into a useable form onto the local computer system.
  • the means of blending inco ⁇ orates environmental information, and is a blending algorithm, wherein the blending algorithm uses inverse kinematics and/or genetics algorithm.

Landscapes

  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Processing Or Creating Images (AREA)

Abstract

L'invention concerne un système et un procédé permettant de fournir une animation personnalisable. Une base de données de capture (20) de mouvements héberge une bibliothèque de mouvements humains accessible via une connexion client-serveur. Des utilisateurs (150) peuvent sélectionner des séquences de mouvements, et modifier ces mouvements en temps réel sur une connexion de réseau, afin de créer de nouvelles oeuvres d'art originales. Dans un mode de réalisation, on utilise un logiciel mélangeur (130) afin d'effectuer une interpolation et une extrapolation à partir des données existantes dans la bibliothèque, de façon à créer de nouvelles séquences de mouvements originales développées de façon à satisfaire aux lois de la bio-mécanique, de la physique et de l'anatomie humaine. Dans un autre mode de réalisation, les utilisateurs peuvent contourner les lois de la bio-mécanique, de la physique et de l'anatomie humaine, et créer des mutations des données de capture. Cette invention permet de manipuler une animation en temps réel, et d'exécuter cette manipulation sur un réseau à l'aide d'une application client, d'un algorithme de mélange et par accès à la bibliothèque de capture de mouvements. L'animation ainsi créée s'achète sur l'Internet, ce qui fournit des moyens de création et d'achat d'animation bon marché.
PCT/US2000/022232 1999-08-10 2000-08-10 Systeme d'animation personnalisable WO2001011570A1 (fr)

Priority Applications (1)

Application Number Priority Date Filing Date Title
AU67703/00A AU6770300A (en) 1999-08-10 2000-08-10 Customizable animation system

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
US14803299P 1999-08-10 1999-08-10
US60/148,032 1999-08-10

Publications (1)

Publication Number Publication Date
WO2001011570A1 true WO2001011570A1 (fr) 2001-02-15

Family

ID=22523936

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/US2000/022232 WO2001011570A1 (fr) 1999-08-10 2000-08-10 Systeme d'animation personnalisable

Country Status (2)

Country Link
AU (1) AU6770300A (fr)
WO (1) WO2001011570A1 (fr)

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5561745A (en) * 1992-10-16 1996-10-01 Evans & Sutherland Computer Corp. Computer graphics for animation by time-sequenced textures
US5923791A (en) * 1991-09-18 1999-07-13 Sarnoff Corporation Video merging employing pattern-key insertion
US6088042A (en) * 1997-03-31 2000-07-11 Katrix, Inc. Interactive motion data animation system
US6115052A (en) * 1998-02-12 2000-09-05 Mitsubishi Electric Information Technology Center America, Inc. (Ita) System for reconstructing the 3-dimensional motions of a human figure from a monocularly-viewed image sequence

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5923791A (en) * 1991-09-18 1999-07-13 Sarnoff Corporation Video merging employing pattern-key insertion
US5561745A (en) * 1992-10-16 1996-10-01 Evans & Sutherland Computer Corp. Computer graphics for animation by time-sequenced textures
US6088042A (en) * 1997-03-31 2000-07-11 Katrix, Inc. Interactive motion data animation system
US6115052A (en) * 1998-02-12 2000-09-05 Mitsubishi Electric Information Technology Center America, Inc. (Ita) System for reconstructing the 3-dimensional motions of a human figure from a monocularly-viewed image sequence

Also Published As

Publication number Publication date
AU6770300A (en) 2001-03-05

Similar Documents

Publication Publication Date Title
US6738065B1 (en) Customizable animation system
Wang et al. Multi-weight enveloping: least-squares approximation techniques for skin animation
Gleicher Motion path editing
Rose et al. Verbs and adverbs: Multidimensional motion interpolation
US5933150A (en) System for image manipulation and animation using embedded constraint graphics
US7650381B2 (en) Network based system design of custom products with live agent support
Lee et al. Motion fields for interactive character locomotion
US7515155B2 (en) Statistical dynamic modeling method and apparatus
Magnenat-Thalmann et al. Virtual humans: thirty years of research, what next?
US9968850B2 (en) System for providing virtual spaces for access by users
Coros et al. Synthesis of constrained walking skills
Cooper et al. Active learning for real-time motion controllers
Colas et al. Interaction fields: Intuitive sketch‐based steering behaviors for crowd simulation
US7091975B1 (en) Shape and animation methods and systems using examples
Wampler Fast and reliable example-based mesh IK for stylized deformations
US20190139288A1 (en) Dynamic motion simulation methods and systems
WO2001011570A1 (fr) Systeme d'animation personnalisable
US7742901B2 (en) Method and system for virtual object generation
Hwang et al. Simplified physical model‐based balance‐preserving motion re‐targeting for physical simulation
Ijiri et al. ProcDef: Local‐to‐global Deformation for Skeleton‐free Character Animation
WO2004104935A1 (fr) Procede et appareil de modelisation dynamique statistique
Song Development of a Previsualization Proxy Plug-in Tool
WO2005076225A1 (fr) Analyse de posture et de mouvement a partir de quaternions
WO2005076226A1 (fr) Techniques d'animation utilisant des quaternions
Cremonesi et al. Toward a new fashion concepts design tool: The vMannequin framework

Legal Events

Date Code Title Description
AK Designated states

Kind code of ref document: A1

Designated state(s): AE AG AL AM AT AU AZ BA BB BG BR BY BZ CA CH CN CR CU CZ DE DK DM DZ EE ES FI GB GD GE GH GM HR HU ID IL IN IS JP KE KG KP KR KZ LC LK LR LS LT LU LV MA MD MG MK MN MW MX MZ NO NZ PL PT RO RU SD SE SG SI SK SL TJ TM TR TT TZ UA UG US UZ VN YU ZA ZW

AL Designated countries for regional patents

Kind code of ref document: A1

Designated state(s): GH GM KE LS MW MZ SD SL SZ TZ UG ZW AM AZ BY KG KZ MD RU TJ TM AT BE CH CY DE DK ES FI FR GB GR IE IT LU MC NL PT SE BF BJ CF CG CI CM GA GN GW ML MR NE SN TD TG

121 Ep: the epo has been informed by wipo that ep was designated in this application
DFPE Request for preliminary examination filed prior to expiration of 19th month from priority date (pct application filed before 20040101)
REG Reference to national code

Ref country code: DE

Ref legal event code: 8642

122 Ep: pct application non-entry in european phase
NENP Non-entry into the national phase

Ref country code: JP