WO2023087753A1 - 动作数据获取方法、***、装置、设备、存储介质和计算机程序产品 - Google Patents

动作数据获取方法、***、装置、设备、存储介质和计算机程序产品 Download PDF

Info

Publication number
WO2023087753A1
WO2023087753A1 PCT/CN2022/105816 CN2022105816W WO2023087753A1 WO 2023087753 A1 WO2023087753 A1 WO 2023087753A1 CN 2022105816 W CN2022105816 W CN 2022105816W WO 2023087753 A1 WO2023087753 A1 WO 2023087753A1
Authority
WO
WIPO (PCT)
Prior art keywords
joints
motion data
skeletal
bone
robot
Prior art date
Application number
PCT/CN2022/105816
Other languages
English (en)
French (fr)
Inventor
付强
马世奎
彭飞
王博玉
Original Assignee
达闼科技(北京)有限公司
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 达闼科技(北京)有限公司 filed Critical 达闼科技(北京)有限公司
Publication of WO2023087753A1 publication Critical patent/WO2023087753A1/zh

Links

Images

Classifications

    • AHUMAN NECESSITIES
    • A63SPORTS; GAMES; AMUSEMENTS
    • A63FCARD, BOARD, OR ROULETTE GAMES; INDOOR GAMES USING SMALL MOVING PLAYING BODIES; VIDEO GAMES; GAMES NOT OTHERWISE PROVIDED FOR
    • A63F13/00Video games, i.e. games using an electronically generated display having two or more dimensions
    • A63F13/55Controlling game characters or game objects based on the game progress
    • A63F13/58Controlling game characters or game objects based on the game progress by computing conditions of game characters, e.g. stamina, strength, motivation or energy level

Definitions

  • the present invention relates to the field of robots, in particular to a motion data acquisition method, system, device, equipment, storage medium and computer program product.
  • the action data usually includes the angular velocity, acceleration, motion trajectory and so on of the robot joints.
  • sensing devices such as motion capture devices and inertial measurement units (IMU for short) can be used to collect motion data.
  • IMU inertial measurement units
  • embodiments of the present invention provide a motion data acquisition method, system, device, device, storage medium, and computer program product, so as to reduce the difficulty of motion data acquisition.
  • an embodiment of the present invention provides a method for acquiring action data, including:
  • an action data acquisition device including:
  • An acquisition module configured to acquire the first bone structure of the virtual character, the first motion data of the joints in the first bone structure, and the second bone structure of the robot;
  • An action data determining module configured to determine a second action of a joint in the second skeletal structure according to the correspondence between the first skeletal structure and the joints in the second skeletal structure and the first action data data.
  • an embodiment of the present invention provides an action data acquisition system, including: a robot and a server;
  • the server is used to obtain the first skeleton structure of the virtual character, the first action data of the joints in the first skeleton structure, and the second skeleton structure of the robot; according to the first skeleton structure and the second skeleton structure The corresponding relationship between the joints, and the first motion data, determine the second motion data of the joints in the second skeletal structure;
  • the robot is configured to receive the second motion data sent by the server; and move according to the second motion data.
  • an embodiment of the present invention provides an electronic device, including a processor and a memory, the memory is used to store one or more computer instructions, wherein, when the one or more computer instructions are executed by the processor Realize the action data acquisition method in the first aspect above.
  • the electronic device may also include a communication interface for communicating with other devices or a communication network.
  • the embodiment of the present invention provides a non-transitory machine-readable storage medium, the non-transitory machine-readable storage medium stores executable code, when the executable code is executed by the processor of the electronic device During execution, the processor can at least realize the action data acquisition method as described in the first aspect.
  • an embodiment of the present invention provides a computer program product, the computer program product includes a computer program/instruction, and when the computer program/instruction is executed by a processor, the motion data acquisition method as described in the first aspect is implemented .
  • the action data acquisition method acquires the first bone structure of the virtual character, the first action data of the joints in the first bone structure, and the second bone structure of the robot. Then, according to the corresponding relationship between the joints in the two bone structures and the first motion data, the second motion data of each joint in the second bone structure is determined, that is, the motion data suitable for the robot is obtained. It can be seen that the above method provides a method for obtaining motion data suitable for a robot according to the motion data of the virtual character. Since the motion data of the virtual character can be obtained directly, the acquisition difficulty is relatively low. Therefore, the motion data suitable for the robot directly obtained through the motion data of the virtual character is also relatively difficult to obtain, which reduces the difficulty of motion data. acquisition cost.
  • FIG. 1 is a flow chart of a method for acquiring action data provided by an embodiment of the present invention
  • Fig. 2 is a schematic diagram of the joint hierarchy in the skeletal structure provided by the embodiment of the present invention.
  • Fig. 3 is a schematic diagram of an operation interface corresponding to the way of establishing correspondence between joints in the skeletal structure provided by the embodiment of the present invention
  • FIG. 4 is a flow chart of another motion data acquisition method provided by an embodiment of the present invention.
  • FIG. 5 is a schematic structural diagram of an action data acquisition system provided by an embodiment of the present invention.
  • FIG. 6 is a schematic structural diagram of an action data acquisition device provided by an embodiment of the present invention.
  • FIG. 7 is a schematic structural diagram of an electronic device corresponding to the motion data acquisition device provided by the embodiment shown in FIG. 6 .
  • the words “if”, “if” as used herein may be interpreted as “at” or “when” or “in response to determining” or “in response to identifying”.
  • the phrases “if determined” or “if identified (the stated condition or event)” could be interpreted as “when determined” or “in response to the determined” or “when identified (the stated condition or event) )” or “in response to recognition of (a stated condition or event)”.
  • Fig. 1 is a flowchart of a motion data acquisition method provided by an embodiment of the present invention, and the motion data acquisition method provided by the embodiment of the present invention can be executed by an acquisition device.
  • the acquisition device may be implemented as software, or a combination of software and hardware, such as a server maintained by a robot developer. As shown in Figure 1, the method includes the following steps:
  • the server may directly obtain the first bone structure of the virtual character and the first motion data of the joints in the first bone structure from the established motion data database.
  • the first motion data may include angular velocity, acceleration, motion track, etc. of joints in the skeleton structure.
  • the server can also obtain the second bone structure of the robot designed by the robot developer.
  • the action data database may store pre-designed bone structures of different virtual characters and action data associated with each bone structure.
  • the virtual character can be designed by the character designer, specifically, it can be a game character or an animation character, etc., and the action data associated with the skeleton structure of the virtual character can also be provided by the character designer.
  • the robot may be a humanoid robot, and optionally, both the first skeleton structure and the second skeleton structure may also be humanoid skeleton structures. And the first skeletal structure and the second skeletal structure often also have a similar skeletal structure.
  • the similarity between the two can be embodied in that the two bone structures have the same number of joints and/or the corresponding joints in the two bone structures can have the same joint hierarchy, for example, the corresponding joints in the two bone structures
  • the joints are all parent joints or all are child joints. Among them, the relationship between the parent-child joints can be understood in combination with the skeleton structure shown in FIG. 2 .
  • circles represent joints
  • triangles represent bones connected to joints.
  • a corresponding relationship between joints in the two skeletal structures can also be established. Based on the established corresponding relationship and based on the first motion data, second motion data of joints in the second skeletal structure are further determined.
  • the corresponding joints in the two skeletal structures can be directly determined as having the same motion data.
  • the first motion data of joint 1 in the first skeletal structure may be directly determined as the second motion data of joint 2 in the second skeletal structure, where joint 1 and joint 2 have a corresponding relationship.
  • the establishment of the corresponding relationship between the joints in the two skeletal structures it can optionally be established manually, or automatically established by the server according to the names of the joints in the two skeletal structures and the joint hierarchical relationship.
  • the establishment of the corresponding relationship may be performed by a game development engine deployed in the server, such as Unreal Engine 4 (UE4 for short).
  • UE4 Unreal Engine 4
  • the server can also send it to the robot, and the robot can control itself to perform the same motion as the virtual character by using the second motion data.
  • the first motion data of the virtual character is often more accurate, and since the second motion data is obtained based on the first motion data, the second motion data can also be guaranteed.
  • the accuracy of the motion data further makes the robot's motion have a higher degree of anthropomorphism, which can ensure the expressiveness of the robot's motion.
  • the first skeleton structure of the virtual character, the first motion data of the joints in the first skeleton structure, and the second skeleton structure of the robot are acquired. Then, according to the corresponding relationship between the joints in the two bone structures and the first motion data, the second motion data of each joint in the second bone structure is determined, that is, the motion data suitable for the robot is obtained. It can be seen that the above method provides a method for obtaining motion data suitable for a robot according to the motion data of the virtual character. Since the motion data of the virtual character can be obtained directly, it is less difficult to obtain. Therefore, the motion data applicable to the robot directly obtained through the motion data of the virtual character also has relatively low difficulty and cost of obtaining.
  • the actions of the virtual character are often not limited, and the virtual character can perform actions with a higher degree of anthropomorphism according to the first action data.
  • the robot is composed of multiple physical mechanical structures, and optionally, the second skeleton structure is designed by the robot developer according to the mechanical mechanism of the robot. Therefore, the first skeleton structure and the second skeleton structure Although the two bone structures are similar, there are still differences. This difference is manifested in the fact that the angles of limit motion and/or setup angles of the corresponding joints in the two bone structures are different.
  • the limit motion angle of the joint is adjusted to the first motion data, so as to obtain the second motion data more suitable for the robot. That is to say, an adjustment process of motion data is implied in the process of determining the second motion data according to the correspondence between the joints and the first motion data. After adjustment, the anthropomorphic degree of the robot's motion can be ensured to the greatest extent without exceeding the limit of the robot's motion.
  • the process of adjusting the first action data may be performed by an adjustment algorithm preset in the server.
  • the adjustment to the first motion data may optionally be adjusted according to the extreme motion angle of the joint in the second skeletal structure.
  • the first motion data indicates that the limit motion angle of the elbow joint in the second skeletal structure is 180°, and considering the mechanical structure of the robot, the limit motion angle of the elbow joint is 150°, therefore, the first The movement data of the elbow joint in the first movement data is adjusted to 150° to obtain the second movement data.
  • the first motion data can also be adjusted according to the setting angles of the joints in the second skeletal structure.
  • the main factors affecting the level of similarity include the setting angles of the joints in the second bone structure, that is, the angle difference between the corresponding joints in the two bone structures. This angle difference is usually produced in consideration of the volume of the mechanical structure so that multiple mechanical structures can be assembled into a robot smoothly.
  • the shoulder joint and the bones connected to it should be in a horizontal direction.
  • the mechanical structure corresponding to the shoulder joint is usually set at a predetermined acute angle with the horizontal direction, such as 20° , then when the angle of the shoulder joint in the first motion data is 20°, it can be corrected to 40°.
  • step S102 after obtaining the adjustment result of the first action data, that is, the second action data, it can be further sent to the robot, so that the robot can act according to the second action data, Thereby ensuring the anthropomorphic degree of robot action.
  • the server can also store the second action data in the form of an action data file, so that the robot can call the file at any time to control the robot to perform actions with a high degree of anthropomorphism.
  • step S102 the corresponding relationship between the joints in the two bone structures can be manually established.
  • the character developer will name each joint in the first bone structure and declare the hierarchical relationship, so as to obtain the first attribute file containing the joint name and hierarchical relationship, and this first attribute file will be obtained when the first bone structure Get it together with the structure.
  • the robot developer can also name each joint in the skeletal structure and declare the hierarchical relationship of the joints, and generate a second property file containing joint names and hierarchical relationships.
  • the server responds to the display operation triggered by the robot developer, and at the same time displays the joint names contained in the first property file and the second property file on the server.
  • the relationship establishment operation triggered by the robot developer, the corresponding relationship between the joints in the two bone structures is artificially established.
  • the specific operation interface can be shown in FIG. 3 .
  • the relationship establishment operation may be a connection operation, that is, the corresponding relationship between joints can be established through manual connection on the operation interface shown in FIG. 3 .
  • the server may also automatically establish a corresponding relationship. Specifically, after the server obtains the above-mentioned property files corresponding to the two skeleton structures, it can automatically establish correspondence according to the joint names and joint hierarchical relationships of the joints in the skeleton structures. For example, if there are joints with the same name in the first property file and the second property file, and the joints with the same name also have the same hierarchical relationship, a corresponding relationship between the two joints is established.
  • the server can directly obtain the first bone structure of the virtual character from the database of motion data. It is easy to understand that the database of action data can contain multiple candidate bone structures, and optionally, the bone structure closest to the second bone structure can be determined as the first bone structure from the candidate bone structures by manual selection .
  • the server may also automatically select the first bone structure through a configured selection algorithm. Specifically, the server may separately count the number of joints of the plurality of candidate bone structures, and determine the bone structure having the same number of joints as the second bone structure as the first bone structure.
  • the server can also obtain the candidate bone structure and the hierarchical relationship of each joint in the second bone structure, and according to the candidate bone structure and the second bone structure Hierarchical relationships of the same joints in the structure, determining the first bone structure from the candidate bone structures. For example, if the target bone structure has the same hierarchical relationship as shoulder joints, elbow joints, wrist joints, knee joints, etc. in the second bone structure, it can be determined that the target bone structure is the first bone model, wherein the target bone structure is any of the bone structures in the alternative bone structure.
  • the number of joints in the bone structure and the hierarchical relationship of the joints may also be considered to select the first bone structure from the candidate bone structures.
  • Fig. 4 is a flow chart of another action data acquisition method provided by the embodiment of the present invention. As shown in Fig. 4, the method may include the following steps:
  • the server can directly obtain the first skeletal animation from the skeletal animation database, and then obtain multiple frames of the first skeletal image in the first skeletal animation by sampling the first skeletal animation, and simultaneously extract Get the motion data of the joints in the skeleton image.
  • the motion data of the joints in multiple frames of the first skeleton image may constitute the first motion data in the embodiment shown in FIG. 1 .
  • the server may also acquire the second skeleton structure of the robot designed by the robot developer and the second skeleton image including the second skeleton structure.
  • the server may also generate a plurality of second skeleton images containing the second skeleton structure based on the sampling results of the first skeleton animation, wherein the number of the second skeleton images is equal to that of the first skeleton images.
  • the motion data of the joints in the second bone image is determined, and the second bone image and the joint motion data in the second bone image The motion data generates a second skeletal animation.
  • the motion data of the joints in the first skeleton image may be directly determined as the motion data of the joints in the second skeleton image.
  • the motion data of the joints in the first bone image can also be adjusted to obtain the motion of the joints in the second bone image suitable for the robot data.
  • the specific adjustment process reference may be made to the relevant description above, which will not be repeated here.
  • the server may also store the motion data of the joints in the second skeleton image in the form of motion files.
  • the action data of the joints in each frame of the skeletal image can be obtained by sampling and extraction, and then with the help of two
  • the corresponding relationship between the joints in the skeleton structure adjust the motion data of the joints in the first skeleton image to obtain the motion data of the joints in the second skeleton image corresponding to the robot, that is, to obtain the motion data suitable for the robot, so that Greatly reduce the difficulty and cost of acquiring motion data.
  • FIG. 5 is a schematic structural diagram of an action data acquisition system provided by an embodiment of the present invention.
  • the system includes: servers and robots.
  • the server may first obtain the first bone structure of the virtual character, the first motion data of the joints in the first bone structure, and at the same time obtain the second bone structure of the robot provided by the robot developer.
  • the second bone structure may be designed by the robot developer according to the actual mechanical structure of the robot.
  • the server establishes a correspondence between the joints in the first skeleton structure and the second skeleton structure, and finally obtains the second movement data of the joints in the second skeleton structure according to the correspondence and the first movement data.
  • joints having a corresponding relationship may be directly determined as having the same motion data.
  • the corresponding relationship between the two skeleton structures can be established manually or automatically by the server. For details, please refer to the relevant description above, which will not be repeated here.
  • the server may send it to the robot, so that the robot can move according to the second motion data.
  • the first action data can also be processed according to the difference between the two skeletal structures. Adjust to get the second motion data that more closely matches the robot.
  • the difference between the two skeletal structures can be specifically reflected in the difference in the extreme motion angles and/or the setting angles of the corresponding joints in the two skeletal structures, and the adjustment process for the first motion data can be referred to the above-mentioned related description and will not be repeated here.
  • the first bone structure for the selection of the first bone structure, it can be selected manually, or can be selected according to the hierarchical relationship of the same joint in the bone structure.
  • the specific selection process can also refer to the relevant description above, and will not be repeated here.
  • the server may also store it in the form of an action data file.
  • the file can be called directly to control the robot to make the corresponding action.
  • the action data can also be displayed with skeletal animation as the carrier, then the server can obtain the first action data corresponding to the virtual character by sampling and extracting the skeletal animation, and then based on the correspondence between the joints in the two skeletal structures Second motion data applicable to the robot is determined.
  • Fig. 6 is a schematic structural diagram of an action data acquisition device provided by an embodiment of the present invention. As shown in Fig. 6, the device includes:
  • the acquisition module 11 is configured to acquire the first skeleton structure of the virtual character, the first motion data of the joints in the first skeleton structure, and the second skeleton structure of the robot.
  • An action data determining module 12 configured to determine the second position of the joint in the second skeletal structure according to the corresponding relationship between the first skeletal structure and the joints in the second skeletal structure and the first action data. action data.
  • the action data determination module 12 is specifically configured to: according to the angle difference between joints having the corresponding relationship in the first bone structure and the second bone structure and/or the second bone structure The limit motion angles of the joints in the structure are adjusted to obtain the second motion data by adjusting the first motion data.
  • the acquiring module 11 is configured to acquire a first skeleton animation including the first skeleton structure and the first motion data, and the second skeleton structure.
  • the action data determining module 12 is specifically configured to: sample the first skeletal animation to obtain the first skeletal image and the action data of joints in the first skeletal image; Action data of the joints in the skeletal image, determining the action data of the joints in the second skeletal image including the second skeletal structure, so as to obtain a second skeletal animation including the second skeletal structure and the second action data.
  • the device further includes: a storage module 13, configured to store motion data of joints in the second skeleton image.
  • the device further includes: a bone structure determining module 14, configured to determine, among the candidate bone structures, a bone structure having the same number of joints as the second bone structure as the first bone structure.
  • a bone structure determining module 14 configured to determine, among the candidate bone structures, a bone structure having the same number of joints as the second bone structure as the first bone structure.
  • the first bone structure is determined from the candidate bone structure according to the hierarchical relationship between the candidate bone structure and the same joint in the second bone structure.
  • the device further includes: a creation module 15, configured to create the second bone structure according to the mechanical structure of the robot.
  • the device further includes: a corresponding relationship establishment module 16, configured to establish the corresponding relationship according to the joint hierarchical relationship of the joints in the first bone structure and the second bone structure.
  • a corresponding relationship establishment module 16 configured to establish the corresponding relationship according to the joint hierarchical relationship of the joints in the first bone structure and the second bone structure.
  • the device further includes: a sending module 17, configured to send the second action data to the robot, so as to control the robot to perform actions according to the second action data.
  • a sending module 17 configured to send the second action data to the robot, so as to control the robot to perform actions according to the second action data.
  • the device shown in FIG. 6 can execute the method of the embodiment shown in FIG. 1 to FIG. 4 .
  • the parts not described in detail in this embodiment refer to the relevant description of the embodiment shown in FIG. 1 to FIG. 4 .
  • the structure of the motion data acquisition device can be implemented as an electronic device.
  • the electronic device can include: a processor 21 and a memory twenty two.
  • the memory 22 is used to store a program that supports the electronic device to execute the motion data acquisition method provided in the embodiment shown in FIGS. 1 to 4 above, and the processor 21 is configured to execute the stored program.
  • the program includes one or more computer instructions, wherein, when the one or more computer instructions are executed by the processor 21, the following steps can be realized:
  • the processor 21 is further configured to execute all or part of the steps in the foregoing embodiments shown in FIG. 1 to FIG. 4 .
  • the structure of the electronic device may further include a communication interface 23 for the electronic device to communicate with other devices or a communication network.
  • an embodiment of the present invention provides a computer storage medium, which is used to store computer software instructions used by the above-mentioned electronic device, which includes instructions for executing the action data acquisition method in the above-mentioned method embodiments shown in FIGS. 1 to 4 . program.
  • An embodiment of the present invention provides a computer program product, which includes computer programs/instructions for executing the method for acquiring action data in the method embodiments shown in FIGS. 1 to 4 above.

Landscapes

  • Engineering & Computer Science (AREA)
  • Multimedia (AREA)
  • Theoretical Computer Science (AREA)
  • Human Computer Interaction (AREA)
  • Processing Or Creating Images (AREA)

Abstract

本发明实施例提供一种动作数据获取方法、***、装置、设备、存储介质和计算机程序产品,该方法包括:获取虚拟角色的第一骨骼结构、第一骨骼结构中关节的第一动作数据以及机器人的第二骨骼结构。再根据两个骨骼结构中关节之间的对应关系,以及第一动作数据,确定第二骨骼结构中各关节的第二动作数据,也即是得到了适用于机器人的动作数据。可见,上述方法提供了一种根据虚拟角色的动作数据获取适用于机器人的动作数据的方法。由于虚拟角色的动作数据是能够直接获取的,其的获取难度较低,因此,通过虚拟角色的动作数据直接得到的适用于机器人的动作数据,也具有较低的获取难度。

Description

动作数据获取方法、***、装置、设备、存储介质和计算机程序产品
交叉引用
本申请要求2021年11月19日递交的、申请号为“2021113998745”、发明名称为“动作数据获取方法、***、装置、设备和存储介质”的中国专利申请的优先权,其全部内容通过引用结合在本申请中。
技术领域
本发明涉及机器人领域,尤其涉及一种动作数据获取方法、***、装置、设备、存储介质和计算机程序产品。
背景技术
传统机器人特别是人型机器人,通常是利用动作数据来控制其的动作。其中,动作数据通常包括机器人关节的角速度、加速度、运动轨迹等等内容。
现有技术中,可以使用动作捕捉设备、惯性测量单元(Inertial Measurement Unit,简称IMU)等传感设备采集动作数据。但传感设备的成本往往较高,从而增大适用于机器人的动作数据的获取难度。因此,如何降低动作数据获取的难度就成为一个亟待解决的问题。
发明内容
有鉴于此,本发明实施例提供一种动作数据获取方法、***、装置、设备、存储介质和计算机程序产品,用以降低动作数据的获取难度。
第一方面,本发明实施例提供一种动作数据获取方法,包括:
获取虚拟角色的第一骨骼结构、所述第一骨骼结构中关节的第一动作数据以及机器人的第二骨骼结构;
根据所述第一骨骼结构和所述第二骨骼结构中关节之间的对应关系,以 及所述第一动作数据,确定所述第二骨骼结构中关节的第二动作数据。
第二方面,本发明实施例提供一种动作数据获取装置,包括:
获取模块,用于获取虚拟角色的第一骨骼结构、所述第一骨骼结构中关节的第一动作数据以及机器人的第二骨骼结构;
动作数据确定模块,用于根据所述第一骨骼结构和所述第二骨骼结构中关节之间的对应关系,以及所述第一动作数据,确定所述第二骨骼结构中关节的第二动作数据。
第三方面,本发明实施例提供一种动作数据获取***,包括:机器人和服务器;
所述服务器,用于获取虚拟角色的第一骨骼结构、所述第一骨骼结构中关节的第一动作数据以及机器人的第二骨骼结构;根据所述第一骨骼结构和所述第二骨骼结构中关节之间的对应关系,以及所述第一动作数据,确定所述第二骨骼结构中关节的第二动作数据;
所述机器人,用于接收所述服务器发送的所述第二动作数据;按照所述第二动作数据运动。
第四方面,本发明实施例提供一种电子设备,包括处理器和存储器,所述存储器用于存储一条或多条计算机指令,其中,所述一条或多条计算机指令被所述处理器执行时实现上述第一方面中的动作数据获取方法。该电子设备还可以包括通信接口,用于与其他设备或通信网络通信。
第五方面,本发明实施例提供了一种非暂时性机器可读存储介质,所述非暂时性机器可读存储介质上存储有可执行代码,当所述可执行代码被电子设备的处理器执行时,使所述处理器至少可以实现如第一方面所述的动作数据获取方法。
第六方面,本发明实施例提供了一种计算机程序产品,所述计算机程序产品包括计算机程序/指令,所述计算机程序/指令被处理器执行时实现如第一方面所述的动作数据获取方法。
本发明实施例提供的动作数据获取方法,获取虚拟角色的第一骨骼结构、第一骨骼结构中关节的第一动作数据以及机器人的第二骨骼结构。再根据两个骨骼结构中关节之间的对应关系,以及第一动作数据,确定第二骨骼结构中各关节的第二动作数据,也即是得到了适用于机器人的动作数据。可见,上述方法提供了一种根据虚拟角色的动作数据获取适用于机器人的动作数据的方法。由于虚拟角色的动作数据是能够直接获取的,其的获取难度较低,因此,通过虚拟角色的动作数据直接得到的适用于机器人的动作数据也具有较低的获取难度,也就降低了动作数据获取的成本。
附图说明
为了更清楚地说明本发明实施例或现有技术中的技术方案,下面将对实施例或现有技术描述中所需要使用的附图作一简单地介绍,显而易见地,下面描述中的附图是本发明的一些实施例,对于本领域普通技术人员来讲,在不付出创造性劳动的前提下,还可以根据这些附图获得其他的附图。
图1为本发明实施例提供的一种动作数据获取方法的流程图;
图2为本发明实施例提供的骨骼结构中关节层次的示意图;
图3为本发明实施例提供的骨骼结构中关节之间对应关系建立方式对应的操作界面示意图;
图4为本发明实施例提供的另一种动作数据获取方法的流程图;
图5为本发明实施例提供的一种动作数据获取***的结构示意图;
图6为本发明实施例提供的一种动作数据获取装置的结构示意图;
图7为与图6所示实施例提供的动作数据获取装置对应的电子设备的结构示意图。
具体实施方式
为使本发明实施例的目的、技术方案和优点更加清楚,下面将结合本发明实施例中的附图,对本发明实施例中的技术方案进行清楚、完整地描述, 显然,所描述的实施例是本发明一部分实施例,而不是全部的实施例。基于本发明中的实施例,本领域普通技术人员在没有作出创造性劳动前提下所获得的所有其他实施例,都属于本发明保护的范围。
在本发明实施例中使用的术语是仅仅出于描述特定实施例的目的,而非旨在限制本发明。在本发明实施例和所附权利要求书中所使用的单数形式的“一种”、“所述”和“该”也旨在包括多数形式,除非上下文清楚地表示其他含义,“多种”一般包含至少两种,但是不排除包含至少一种的情况。
应当理解,本文中使用的术语“和/或”仅仅是一种描述关联对象的关联关系,表示可以存在三种关系,例如,A和/或B,可以表示:单独存在A,同时存在A和B,单独存在B这三种情况。另外,本文中字符“/”,一般表示前后关联对象是一种“或”的关系。
取决于语境,如在此所使用的词语“如果”、“若”可以被解释成为“在……时”或“当……时”或“响应于确定”或“响应于识别”。类似地,取决于语境,短语“如果确定”或“如果识别(陈述的条件或事件)”可以被解释成为“当确定时”或“响应于确定”或“当识别(陈述的条件或事件)时”或“响应于识别(陈述的条件或事件)”。
还需要说明的是,术语“包括”、“包含”或者其任何其他变体意在涵盖非排他性的包含,从而使得包括一系列要素的商品或者***不仅包括那些要素,而且还包括没有明确列出的其他要素,或者是还包括为这种商品或者***所固有的要素。在没有更多限制的情况下,由语句“包括一个……”限定的要素,并不排除在包括所述要素的商品或者***中还存在另外的相同要素。
下面结合附图对本发明的一些实施方式作详细说明。在各实施例之间不冲突的情况下,下述的实施例及实施例中的特征可以相互组合。另外,下述各方法实施例中的步骤时序仅为一种举例,而非严格限定。
图1为本发明实施例提供的一种动作数据获取方法的流程图,本发明实施 例提供的该动作数据获取方法可以获取设备执行。可以理解的是,该获取设备可以实现为软件、或者软件和硬件的组合,比如可以是由机器人开发方维护的服务器。如图1所示,该方法包括如下步骤:
S101,获取虚拟角色的第一骨骼结构、第一骨骼结构中关节的第一动作数据以及机器人的第二骨骼结构。
服务器可以从已经建立的动作数据的数据库直接中获取虚拟角色的第一骨骼结构以及第一骨骼结构中关节的第一动作数据。其中,正如背景技术中描述的,第一动作数据可以包括骨骼结构中关节的角速度、加速度、运动轨迹等等。同时,服务器还可以获取由机器人开发方设计的机器人的第二骨骼结构。
其中,动作数据的数据库可以存储有预先设计的、不同虚拟角色各自的骨骼结构以及与各骨骼结构关联的动作数据。虚拟角色可以由角色设计方设计,具体可以是游戏角色或者是动画角色等等,并且与此虚拟角色的骨骼结构关联的动作数据也可以由角色设计方一并提供。
在实际中,机器人可以是人型机器人,则可选地,第一骨骼结构和第二骨骼结构也均可以是人型骨骼结构。并且第一骨骼结构和第二骨骼结构往往还具有相似的骨骼结构。可选地,二者之间的相似性可以具体体现为两个骨骼结构具有相同的关节数量和/或两个骨骼结构中对应的关节可以具有相同的关节层次关系,比如两个骨骼结构中对应的关节都是父关节或者都是子关节。其中,对于父子关节之间的关系可以结合图2所示的骨骼结构理解。在图2中,圆形代表关节,三角形代表与关节相连的骨骼。
S102,根据第一骨骼结构和第二骨骼结构中关节之间的对应关系,以及第一动作数据,确定第二骨骼结构中关节的第二动作数据。
基于步骤S101获取到的两个骨骼结构,还可以建立两个骨骼结构中各关节之间的对应关系。基于建立的对应关系,以第一动作数据为基础,进一步确定第二骨骼结构中关节的第二动作数据。
由于步骤S101中已经说明两个骨骼结构具有相似性,因此,可选地,可 以直接将两个骨骼结构中具有对应关系的关节确定为具有相同的运动数据。举例来说,可以将第一骨骼结构中关节1的第一动作数据直接确定为第二骨骼结构中关节2的第二动作数据,其中,关节1和关节2具有对应关系。
而对于两个骨骼结构中关节之间对应关系的建立,可选地,可以人为建立,或者由服务器根据两个骨骼结构中各关节的名称和关节层次关系自动建立。具体建立过程可以参见下述实施例中的描述。可选地,对应关系的建立可以由服务器中部署的游戏开发引擎,比如虚幻4引擎(Unreal Engine 4,简称UE4)来执行。
对于得到的第二动作数据,可选地,服务器还可以将其发送至机器人,机器人使用第二动作数据便可以控制自身做出与虚拟角色相同的动作。同时,相比于机器人开发方利用运动规划算法得到的动作数据,虚拟角色的第一动作数据往往更加准确,并且由于第二动作数据是根据第一动作数据得到的,因此,也能够保证第二动作数据的准确性,进一步使得机器人的动作具有更高的拟人化程度,能够保证机器人动作的表现性。
本实施例中,获取虚拟角色的第一骨骼结构、第一骨骼结构中关节的第一动作数据以及机器人的第二骨骼结构。再根据两个骨骼结构中关节之间的对应关系,以及第一动作数据,确定第二骨骼结构中各关节的第二动作数据,也即是得到了适用于机器人的动作数据。可见,上述方法提供了一种根据虚拟角色的动作数据获取适用于机器人的动作数据的方法。由于虚拟角色的动作数据是能够直接获取的,其的获取难度较低,因此,通过虚拟角色的动作数据直接得到的适用于机器人的动作数据,也具有较低的获取难度和成本。
在实际中,考虑到虚拟角色的虚拟性,虚拟角色的动作往往不受限,虚拟角色按照第一动作数据能够做出的拟人化程度更高的动作。但与虚拟角色不同的是,机器人是由多个实体的机械结构构成的,并且可选地,第二骨骼结构是机器人开发方按照机器人具有的机械机构设计的,因此,第一骨骼结构和第二骨骼结构虽然相似,但还是存在差异。这种差异体现在:两个骨骼 结构中对应关节的极限运动角度和/或设置角度不同。因此,为了进一步提供第二动作数据与机器人的适配性,在得到第一动作数据后,还可以根据两个骨骼结构中具有对应关系的关节之间的角度差异和/或第二骨骼结构中关节的极限运动角度,对第一动作数据进行调整,从而得到更适用于机器人的第二动作数据。也即是在根据关节之间的对应关系以及第一动作数据确定第二动作数据的过程中暗含一个动作数据的调整过程。经过调整能够使在不超出机器人动作极限的情况下,最大程度上保证机器人动作的拟人化程度。
可选地,调整第一动作数据的过程可以由服务器中预设的调整算法执行。
对于对第一动作数据的调整,可选地,可以根据第二骨骼结构中关节的极限运动角度进行调整。举例来说,第一运动数据表明第二骨骼结构中手肘关节的极限运动角度是180°,而考虑到机器人的机械结构,该手肘关节的极限运动角度是150°,因此,可以将第一动作数据中手肘关节的动作数据调整为150°,以得到第二动作数据。
可选地,还可以根据第二骨骼结构中各关节的设置角度进行第一动作数据的调整。具体地,上述描述中已经提及第一骨骼结构和第二骨骼结构之间具有相似性,但二者存在差异。而影响相似性高低的主要因素包括第二骨骼结构中各关节的设置角度,即两个骨骼结构中对应关节之间的角度差异。而这种角度差异通常是考虑到机械结构的体积,为使多个机械结构能够顺利组装成机器人才产生的。
举例来说,对于骨骼结构中的肩关节,正常情况下,肩关节及其相连的骨骼应该处于水平方向。而考虑机械结构的体积,为了使肩关节对应的机械结构能够于手臂关节对应的机械结构拼接在一起,肩关节对应的机械结构通常会设置于与水平方向成某一预设锐角,比如20°,则当第一动作数据中肩关节所呈的角度为20°时,则可以将其修正为40°。
可选地,与步骤S102中的内容类似的,在得到第一动作数据的调整结果即第二动作数据后,可以进一步将其发送至机器人,以使所述机器人按照第二动作数据进行动作,从而保证机器人动作拟人化程度。
可选地,服务器还可以将第二动作数据以动作数据文件的形式进行存储,以使机器人能够随时调用该文件,控制机器人做出拟人化程度高的动作。
步骤S102中已经提及可以人工建立两个骨骼结构中各关节之间的对应关系。则具体地,角色开发方会对第一骨骼结构中的各关节进行命名并声明层次关系,以得到包含关节名称和层次关系的第一属性文件,并且此第一属性文件会在获取第一骨骼结构时一并得到。类似的,机器人开发方在设计第二骨骼结构时也可以对骨骼结构中的各关节进行命名同时声明该关节的层次关系,并生成包含关节名称和层次关系的第二属性文件。
则服务器响应于机器人开发方触发显示操作,同时将第一属性文件和第二属性文件中包含的关节名称展示在服务器上。再响应于机器人开发方触发的关系建立操作,从而人工建立两个骨骼结构中关节之间的对应关系。具体操作界面可以如图3所示。可选地,关系建立操作可以是连线操作,也即是在图3所示的操作界面上通过人工连线的方式能够建立关节之间的对应关系。
除了上述方式,可选地,服务器还可以自动建立对应关系。具体地,服务器在得到两个骨骼结构各自对应的上述属性文件后,可以根据骨骼结构中各关节的关节名称和关节层次关系自动对应建立。比如,若第一属性文件和第二属性文件中具有相同名称的关节,并且具有相同名称的关节也具有相同层次关系,从而建立其这两个关节之间的对应关系。
图1所示实施例中已经公开:服务器可以从动作数据的数据库直接中获取虚拟角色的第一骨骼结构。容易理解的,动作数据的数据库可以包含多个备选骨骼结构,则可选地,可以通过人工选择的方式从备选骨骼结构中确定与第二骨骼结构最接近的骨骼结构为第一骨骼结构。
可选地,服务器还可以通过配置的选择算法自动选择第一骨骼结构。具体地,服务器可以分别统计多个备选骨骼结构各自的关节数量,并将与第二骨骼结构具有相同关节数量的骨骼结构确定为第一骨骼结构。
进一步地,为了使选择出的第一骨骼结构与第二骨骼结构更相似,服务器还可以获取备选骨骼结构以及第二骨骼结构中各关节的层次关系,并根据备选骨骼结构和第二骨骼结构中相同关节的层次关系,从备选骨骼结构中确定第一骨骼结构。比如,若目标骨骼结构与第二骨骼结构中的肩关节、肘关节、腕关节、膝关节等具有相同的层次关系,则可以确定此目标骨骼结构为第一骨骼模型,其中,此目标骨骼结构是备选骨骼结构中的任一骨骼结构。
可选地,在实际中还可以同时考虑骨骼结构中关节的数量以及关节的层次关系从备选骨骼结构中选择第一骨骼结构。
可选地,虚拟角色和机器人做出的动作通常可以借助一段骨骼动画来表现,此骨骼动画中包含了骨骼结构和骨骼结构中各关节的动作数据。则图4为本发明实施例提供的另一种动作数据获取方法的流程图,如图4所示,该方法可以包括如下步骤:
S201,获取包含虚拟角色的第一骨骼结构、第一骨骼结构中关节的第一动作数据的第一骨骼动画以及机器人的第二骨骼结构。
S202,对第一骨骼动画进行采样,以得到第一骨骼图像以及第一骨骼图像中关节的动作数据。
服务器可以直接从骨骼动画数据库中获取第一骨骼动画,再通过对第一骨骼动画进行采样,以得到第一骨骼动画中的多帧第一骨骼图像,并同时从每张第一骨骼图像中提取出该骨骼图像中关节的动作数据。多帧第一骨骼图像中关节的动作数据可以构成图1所示实施例中的第一动作数据。
另外,服务器还可以获取由机器人开发方设计的机器人的第二骨骼结构以及包含第二骨骼结构的第二骨骼图像。服务器还可以基于第一骨骼动画的采样结果生成包含第二骨骼结构的多张第二骨骼图像,其中第二骨骼图像和第一骨骼图像的数量相等。
S203,根据第一骨骼结构和第二骨骼结构中关节之间的对应关系和第一骨骼图像中关节的动作数据,确定包含第二骨骼结构的第二骨骼图像中关节 的动作数据,以生成包含第二骨骼结构和第二动作数据的第二骨骼动画。
接着,根据两个骨骼结构中关节之间对应关系,以及第一骨骼图像中关节的动作数据,确定第二骨骼图像中关节的动作数据,并由第二骨骼图像和第二骨骼图像中关节的动作数据生成第二骨骼动画。
可选地,可以直接将第一骨骼图像中关节的动作数据确定为第二骨骼图像中关节的动作数据。
可选地,考虑到第一骨骼结构与第二骨骼结构之间的差异,还可以对第一骨骼图像中关节的动作数据进行调整,以得到适用于机器人的、第二骨骼图像中关节的动作数据。具体调整过程可以参见上述的相关描述,在此不再赘述。
可选地,服务器还可以以动作文件的形式存储第二骨骼图像中关节的动作数据。
本实施例中,当服务器从数据库中获取到的是骨骼动画,以及该骨骼动画对应的动作数据时,可以通过采样并提取的方式得到每一帧骨骼图像中关节的动作数据,再借助两个骨骼结构中关节之间的对应关系,对第一骨骼图像中关节的动作数据进行调整,以得到机器人对应的第二骨骼图像中关节的动作数据,也即是得到适用于机器人的动作数据,从而大大降低动作数据的获取难度和获取成本。
另外,本实施例未详细描述的部分,可参考对图1至图4所示实施例的相关说明。该技术方案的执行过程和技术效果参见图1至图4所示实施例中的描述,在此不再赘述。
上述各实施例已经从服务器的角度对获取适用于机器人的动作数据的过程进行描述,在此基础上,还可以从动作数据获取***的角度,描述***中各设备的具体工作过程。
则图5为本发明实施例提供的一种动作数据获取***的结构示意图。该***包括:服务器和机器人。
服务器可以先从获取虚拟角色的第一骨骼结构、第一骨骼结构中关节的第一动作数据,同时获取机器人开发方提供的机器人的第二骨骼结构。可选地,第二骨骼结构可以由机器人开发方按照机器人实际的机械结构设计。
接着,服务器建立第一骨骼结构和第二骨骼结构中关节之间的对应关系,并根据此对应关系,以及第一动作数据,最终得到第二骨骼结构中关节的第二动作数据。可选地,可以直接将具有对应关系的关节确定为具有相同的动作数据。可选地,两个骨骼结构之间的对应关系可以人工或者由服务器自动建立,具体内容可以参见上述相关描述,在此不再赘述。
对于得到的第二动作数据,可选地,服务器可以将其发送至机器人,以使机器人可以按照第二动作数据运动。
可选地,由于第二骨骼结构是按照机器人实际的机械结构设计的,其与第一骨骼结构虽然相似但存在差异,因此,还可以根据两个骨骼结构之间的差异对第一动作数据进行调整,以得到与机器人更加匹配的第二动作数据。
可选地,两个骨骼结构之间的差异具体可以体现在两个骨骼结构中对应关节的极限运动角度的不同和/或设置角度不同,而对第一动作数据的调整过程可以参见上述的相关描述,在此不再赘述。
可选地,对于第一骨骼结构的选择,可以人工选择,也可以根据骨骼结构中相同关节的层次关系进行选择,具体选择过程也可以参加上述的相关描述,在此不再赘述。
可选地,服务器在得到第二动作数据后还可以以动作数据文件的形式进行存储。当需要控制机器人再次做出该动作时,可以直接调用该文件,以控制机器人做出相应动作。
可选地,动作数据也可以以骨骼动画为载体进行展示,则服务器可以通过对骨骼动画进行采样并提取以得到虚拟角色对应的第一动作数据,再通过根据两个骨骼结构中关节的对应关系确定适用于机器人的第二动作数据。
本实施例中未详细描述的部分,可参考对图1至图4所示实施例的相关说明。该技术方案的执行过程和技术效果参见图1至图4所示实施例中的描 述,在此不再赘述。
以下将详细描述本发明的一个或多个实施例的动作数据获取装置。本领域技术人员可以理解,这些蒙皮处理装置均可使用市售的硬件组件通过本方案所教导的步骤进行配置来构成。
图6为本发明实施例提供的一种动作数据获取装置的结构示意图,如图6所示,该装置包括:
获取模块11,用于获取虚拟角色的第一骨骼结构、所述第一骨骼结构中关节的第一动作数据以及机器人的第二骨骼结构。
动作数据确定模块12,用于根据所述第一骨骼结构和所述第二骨骼结构中关节之间的对应关系,以及所述第一动作数据,确定所述第二骨骼结构中关节的第二动作数据。
可选地,所述动作数据确定模块12具体用于:根据所述第一骨骼结构和所述第二骨骼结构中具有所述对应关系的关节之间的角度差异和/或所述第二骨骼结构中关节的极限运动角度,调整所述第一动作数据,以得到所述第二动作数据。
可选地,所述获取模块11,用于获取包含所述第一骨骼结构和所述第一动作数据的第一骨骼动画,以及所述第二骨骼结构。
所述动作数据确定模块12具体用于:对所述第一骨骼动画进行采样,以得到第一骨骼图像以及所述第一骨骼图像中关节的动作数据;根据所述对应关系和所述第一骨骼图像中关节的动作数据,确定包含所述第二骨骼结构的第二骨骼图像中关节的动作数据,以得到包含所述的第二骨骼结构和所述第二动作数据的第二骨骼动画。
所述装置还包括:存储模块13,用于存储所述第二骨骼图像中关节的动作数据。
可选地,所述装置还包括:骨骼结构确定模块14,用于在备选骨骼结构中,确定与所述第二骨骼结构具有相同数量关节的骨骼结构为所述第一骨骼 结构。
和/或,
根据所述备选骨骼结构与所述第二骨骼结构中相同关节的层次关系,从所述备选骨骼结构中确定所述第一骨骼结构。
可选地,所述装置还包括:创建模块15,用于按照所述机器人具有的机械结构,创建所述第二骨骼结构。
可选地,所述装置还包括:对应关系建立模块16,用于根据所述第一骨骼结构和所述第二骨骼结构中关节的关节层次关系,建立所述对应关系。
可选地,所述装置还包括:发送模块17,用于发送所述第二动作数据至所述机器人,以控制所述机器人按照所述第二动作数据进行动作。
图6所示装置可以执行图1至图4所示实施例的方法,本实施例未详细描述的部分,可参考对图1至图4所示实施例的相关说明。该技术方案的执行过程和技术效果参见图1至图4所示实施例中的描述,在此不再赘述。
以上描述了动作数据获取装置的内部功能和结构,在一个可能的设计中,动作数据获取装置的结构可实现为一电子设备,如图7所示,该电子设备可以包括:处理器21和存储器22。其中,所述存储器22用于存储支持该电子设备执行上述图1至图4所示实施例中提供的动作数据获取方法的程序,所述处理器21被配置为用于执行所述存储器22中存储的程序。
所述程序包括一条或多条计算机指令,其中,所述一条或多条计算机指令被所述处理器21执行时能够实现如下步骤:
获取虚拟角色的第一骨骼结构、所述第一骨骼结构中关节的第一动作数据以及机器人的第二骨骼结构;
根据所述第一骨骼结构和所述第二骨骼结构中关节之间的对应关系,以及所述第一动作数据,确定所述第二骨骼结构中关节的第二动作数据。
可选地,所述处理器21还用于执行前述图1至图4所示实施例中的全部或部分步骤。
其中,所述电子设备的结构中还可以包括通信接口23,用于该电子设备与其他设备或通信网络通信。
另外,本发明实施例提供了一种计算机存储介质,用于储存上述电子设备所用的计算机软件指令,其包含用于执行上述图1至图4所示方法实施例中动作数据获取方法所涉及的程序。
本发明实施例提供了一种计算机程序产品,该产品包括用于执行上述图1~图4所示方法实施例中动作数据获取方法所涉及的计算机程序/指令。
最后应说明的是:以上各实施例仅用以说明本发明的技术方案,而非对其限制;尽管参照前述各实施例对本发明进行了详细的说明,本领域的普通技术人员应当理解:其依然可以对前述各实施例所记载的技术方案进行修改,或者对其中部分或者全部技术特征进行等同替换;而这些修改或者替换,并不使相应技术方案的本质脱离本发明各实施例技术方案的范围。

Claims (13)

  1. 一种动作数据获取方法,其特征在于,包括:
    获取虚拟角色的第一骨骼结构、所述第一骨骼结构中关节的第一动作数据以及机器人的第二骨骼结构;
    根据所述第一骨骼结构和所述第二骨骼结构中关节之间的对应关系,以及所述第一动作数据,确定所述第二骨骼结构中关节的第二动作数据。
  2. 根据权利要求1所述的方法,其特征在于,所述根据所述第一骨骼结构和所述第二骨骼结构中关节之间的对应关系,以及所述第一动作数据,确定所述第二骨骼结构中关节的第二动作数据,包括:
    根据所述第一骨骼结构和所述第二骨骼结构中具有所述对应关系的关节之间的角度差异和/或所述第二骨骼结构中关节的极限运动角度,调整所述第一动作数据,以得到所述第二动作数据。
  3. 根据权利要求1所述的方法,其特征在于,所述获取虚拟角色的第一骨骼结构、所述第一骨骼结构中关节的第一动作数据以及机器人的第二骨骼结构,包括:
    获取包含所述第一骨骼结构和所述第一动作数据的第一骨骼动画,以及所述第二骨骼结构;
    所述根据所述第一骨骼结构和所述第二骨骼结构中关节之间的对应关系,以及所述第一动作数据,确定所述第二骨骼结构中关节的第二动作数据,包括:
    对所述第一骨骼动画进行采样,以得到第一骨骼图像以及所述第一骨骼图像中关节的动作数据;
    根据所述对应关系和所述第一骨骼图像中关节的动作数据,确定包含所述第二骨骼结构的第二骨骼图像中关节的动作数据,以得到包含所述的第二骨骼结构和所述第二动作数据的第二骨骼动画。
  4. 根据权利要求3所述的方法,其特征在于,所述方法还包括:
    存储所述第二骨骼图像中关节的动作数据。
  5. 根据权利要求1所述的方法,其特征在于,所述方法还包括:
    在备选骨骼结构中,确定与所述第二骨骼结构具有相同数量关节的骨骼结构为所述第一骨骼结构;
    和/或,
    根据所述备选骨骼结构与所述第二骨骼结构中相同关节的层次关系,从所述备选骨骼结构中确定所述第一骨骼结构。
  6. 根据权利要求1中所述的方法,其特征在于,所述方法还包括:
    按照所述机器人具有的机械结构,创建所述第二骨骼结构。
  7. 根据权利要求1所述的方法,其特征在于,所述方法还包括:
    根据所述第一骨骼结构和所述第二骨骼结构中关节的层次关系,建立所述对应关系。
  8. 根据权利要求1中所述的方法,其特征在于,所述方法还包括:
    发送所述第二动作数据至所述机器人,以控制所述机器人按照所述第二动作数据进行动作。
  9. 一种动作数据获取***,其特征在于,包括:机器人和服务器;
    所述服务器,用于获取虚拟角色的第一骨骼结构、所述第一骨骼结构中关节的第一动作数据以及机器人的第二骨骼结构;根据所述第一骨骼结构和所述第二骨骼结构中关节之间的对应关系,以及所述第一动作数据,确定所述第二骨骼结构中关节的第二动作数据;
    所述机器人,用于接收所述服务器发送的所述第二动作数据;按照所述第二动作数据运动。
  10. 一种动作数据获取装置,其特征在于,包括:
    获取模块,用于获取虚拟角色的第一骨骼结构、所述第一骨骼结构中关节的第一动作数据以及机器人的第二骨骼结构;
    动作数据确定模块,用于根据所述第一骨骼结构和所述第二骨骼结构中关节之间的对应关系,以及所述第一动作数据,确定所述第二骨骼结构中关节的第二动作数据。
  11. 一种电子设备,其特征在于,包括:存储器、处理器;其中,所述存储器上存储有可执行代码,当所述可执行代码被所述处理器执行时,使所述处理器执行如权利要求1至8中任一项所述的动作数据获取方法。
  12. 一种非暂时性机器可读存储介质,其特征在于,所述非暂时性机器可读存储介质上存储有可执行代码,当所述可执行代码被电子设备的处理器执行时,使所述处理器执行如权利要求1至8中任一项所述的动作数据获取方法。
  13. 一种计算机程序产品,其特征在于,包括计算机程序/指令,所述计算机程序/指令被处理器执行时实现如权利要求1至8中任一项所述的动作数据获取方法。
PCT/CN2022/105816 2021-11-19 2022-07-14 动作数据获取方法、***、装置、设备、存储介质和计算机程序产品 WO2023087753A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202111399874.5 2021-11-19
CN202111399874.5A CN114225420A (zh) 2021-11-19 2021-11-19 动作数据获取方法、***、装置、设备和存储介质

Publications (1)

Publication Number Publication Date
WO2023087753A1 true WO2023087753A1 (zh) 2023-05-25

Family

ID=80750700

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2022/105816 WO2023087753A1 (zh) 2021-11-19 2022-07-14 动作数据获取方法、***、装置、设备、存储介质和计算机程序产品

Country Status (2)

Country Link
CN (1) CN114225420A (zh)
WO (1) WO2023087753A1 (zh)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114225420A (zh) * 2021-11-19 2022-03-25 达闼科技(北京)有限公司 动作数据获取方法、***、装置、设备和存储介质

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20120143374A1 (en) * 2010-12-03 2012-06-07 Disney Enterprises, Inc. Robot action based on human demonstration
US20120239196A1 (en) * 2011-03-15 2012-09-20 Microsoft Corporation Natural Human to Robot Remote Control
CN106965183A (zh) * 2017-05-02 2017-07-21 南京大学 一种基于景深感知机制的机器人操控***及其工作方法
CN109816773A (zh) * 2018-12-29 2019-05-28 深圳市瑞立视多媒体科技有限公司 一种虚拟人物的骨骼模型的驱动方法、插件及终端设备
CN112215930A (zh) * 2020-10-19 2021-01-12 珠海金山网络游戏科技有限公司 一种数据处理方法与装置
CN112873166A (zh) * 2021-01-25 2021-06-01 之江实验室 一种生成机器人肢体动作的方法、装置、电子设备及介质
CN114225420A (zh) * 2021-11-19 2022-03-25 达闼科技(北京)有限公司 动作数据获取方法、***、装置、设备和存储介质

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110413110A (zh) * 2019-07-05 2019-11-05 深圳市工匠社科技有限公司 虚拟角色的控制方法及相关产品
CN111402290B (zh) * 2020-02-29 2023-09-12 华为技术有限公司 一种基于骨骼关键点的动作还原方法以及装置
CN113313794B (zh) * 2021-05-19 2022-11-08 深圳市慧鲤科技有限公司 动画迁移方法和装置、设备及存储介质

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20120143374A1 (en) * 2010-12-03 2012-06-07 Disney Enterprises, Inc. Robot action based on human demonstration
US20120239196A1 (en) * 2011-03-15 2012-09-20 Microsoft Corporation Natural Human to Robot Remote Control
CN106965183A (zh) * 2017-05-02 2017-07-21 南京大学 一种基于景深感知机制的机器人操控***及其工作方法
CN109816773A (zh) * 2018-12-29 2019-05-28 深圳市瑞立视多媒体科技有限公司 一种虚拟人物的骨骼模型的驱动方法、插件及终端设备
CN112215930A (zh) * 2020-10-19 2021-01-12 珠海金山网络游戏科技有限公司 一种数据处理方法与装置
CN112873166A (zh) * 2021-01-25 2021-06-01 之江实验室 一种生成机器人肢体动作的方法、装置、电子设备及介质
CN114225420A (zh) * 2021-11-19 2022-03-25 达闼科技(北京)有限公司 动作数据获取方法、***、装置、设备和存储介质

Also Published As

Publication number Publication date
CN114225420A (zh) 2022-03-25

Similar Documents

Publication Publication Date Title
US20200250889A1 (en) Augmented reality system
JP7182919B2 (ja) 映像処理方法、コンピュータプログラムおよび記録媒体
US9904664B2 (en) Apparatus and method providing augmented reality contents based on web information structure
US20200184726A1 (en) Implementing three-dimensional augmented reality in smart glasses based on two-dimensional data
CN109242978B (zh) 三维模型的视角调整方法和装置
US20230185868A1 (en) Automatic website data migration
WO2019217159A1 (en) Immersive feedback loop for improving ai
US11854231B2 (en) Localizing an augmented reality device
WO2002021753A2 (en) Method and apparatus for transferring data during automated data processing
WO2023087753A1 (zh) 动作数据获取方法、***、装置、设备、存储介质和计算机程序产品
KR102069366B1 (ko) 교육용 로봇과 상호 작용하는 증강 현실 구현 장치 및 그 방법
WO2022174574A1 (zh) 基于传感器的裸手数据标注方法及***
KR101864717B1 (ko) 오브젝트 조립 형상에 따른 무(無)마커 맞춤공간표출형 증강현실 컨텐츠 형성장치 및 방법
CN113066125A (zh) 一种增强现实方法及其相关设备
TWM594733U (zh) 人工智慧輔助擴增實境系統
TWI798514B (zh) 人工智慧輔助擴增實境系統與方法、電腦程式產品
CN115170765A (zh) 一种模型处理***、方法及装置
Agushinta Augmented reality design of Indonesia fruit recognition
TWM596380U (zh) 人工智慧擴增實境輔助系統
JP2020510936A (ja) 補正パターン分析による映像補正方法およびシステム
Sun et al. Bridging semantics with physical objects using augmented reality
Korovin et al. Human pose estimation applying ANN while RGB-D cameras video handling
US20240046521A1 (en) Concurrent camera calibration and bundle adjustment
WO2024050961A1 (zh) 建图方法、装置、设备及存储介质
WO2024140962A1 (zh) 用于确定相对位姿的方法、装置、***、设备和介质

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 22894296

Country of ref document: EP

Kind code of ref document: A1