Die penetration detection method and device

Abstract

The invention provides a mold penetrating detection method and device, and relates to the technical field of motion capture. The method comprises the following steps: acquiring first skeletal animation data of a digital person, wherein the first skeletal animation data comprises first position information and first motion information of a plurality of joints of the digital person in a first frame of a digital person animation; according to the first skeletal animation data, a detection area in a second frame of the digital human animation is determined, and a model of the digital human in the detection area is predicted to have model penetration; acquiring second skeletal animation data of the digital human; in the detection area in the second frame, according to the second skeletal animation data, carrying out cross-model detection on the model of the digital human; and when it is detected that the model of the digital person passes through the model, feedback information is sent, and the feedback information is used for indicating the real person to adjust the action. Due to the fact that the mold penetrating detection speed is high, real-time mold penetrating detection can be achieved, a real person is reminded to adjust actions in time, and mold penetrating of the digital human model is avoided.

Description

[0001] This application claims priority to the Chinese patent application filed with the State Intellectual Property Office of China on September 11, 2023, with application number 202311167036.4 and invention name “A collision detection method, device and computing device cluster”, the entire contents of which are incorporated by reference in this application. Technical Field

[0002] The present application relates to the field of motion capture technology, and more specifically, to a method and device for detecting a penetration pattern. Background Art

[0003] Motion capture (MC) can be called mocap or motion capture. Motion capture refers to the technology of recording and processing the movements of people or other objects. Motion capture is widely used in military, entertainment, sports, medical and other fields. Motion capture involves many fields such as computer vision (CV) and robotics.

[0004] In the scene driven by 3D digital human, skeletal animation data is an important way to drive the movement of digital human. The driven digital human model includes skeleton and skin. For digital human, the skeleton is composed of rigid joints with multiple levels, and the skin can be understood as the surface of the digital human, which can include multiple meshes.

[0005] The surface of the driven digital human model is prone to penetration. The relevant technical solution is to determine the penetration by intersecting the facets, and then the animator will modify it later. In this way, the number of intersecting facets required for penetration detection is large, the penetration detection is time-consuming, and the cost of later modification is high.

[0006] Therefore, how to avoid the penetration of digital human models is an urgent problem to be solved. Summary of the invention

[0007] The present application provides a model penetration detection method and device, which can prevent the model penetration of a digital human model.

[0008] In a first aspect, a model penetration detection method is provided, comprising: obtaining first skeletal animation data of a digital human, the first skeletal animation data comprising first position information and first motion information of multiple joints of the digital human in a first frame of the digital human animation, the first skeletal animation data being determined according to first motion data of a real person collected by a motion capture device; determining a detection area in a second frame of the digital human animation according to the first skeletal animation data, wherein the model of the digital human in the detection area is expected to penetrate the model, and the second frame is after the first frame; obtaining second skeletal animation data of the digital human, the second skeletal animation data comprising second position information and second motion information of the multiple joints of the digital human in the second frame of the digital human animation, the second skeletal animation data being determined according to second motion data of the real person collected by the motion capture device; performing model penetration detection on the model of the digital human according to the second skeletal animation data within the detection area in the second frame; and sending feedback information when it is detected that the model of the digital human penetrates the model, the feedback information comprising position information of the model penetration area of ​​the model of the digital human in the second frame, the feedback information being used to instruct the real person to adjust the action.

[0009] In the embodiment of the present application, the range of the area where the collision may occur is first predicted by the skeletal animation data, and then the penetration detection is performed on the range of the area where the collision may occur, which reduces the amount of calculation, thereby increasing the speed of the penetration detection. In addition, since the penetration detection speed is fast, the method provided in the embodiment of the present application can realize real-time penetration detection, and promptly remind real people to adjust their movements, thereby avoiding the penetration of the digital human model. In addition, since the method provided in the embodiment of the present application reduces the penetration of the skeletal animation data obtained by motion capture, the skeletal animation data can be applied to the digital human that is expected to be driven, without the need for manual modification by technical personnel, thereby reducing time and cost.

[0010] In combination with the first aspect, in certain implementations of the first aspect, determining a detection area in a second frame of the digital human animation based on the first skeletal animation data includes: determining predicted positions of the multiple joints in the second frame of the digital human animation based on first position information and first motion information of the multiple joints of the digital human; and determining a detection area in the second frame of the digital human animation based on the predicted positions of the multiple joints in the second frame of the digital human animation.

[0011] The above scheme first determines the predicted position of the second frame and then determines the detection area of ​​the second frame, which can accurately determine the detection area, thereby improving the accuracy of mold penetration detection.

[0012] In combination with the first aspect, in certain implementations of the first aspect, the multiple joints include a first joint and a second joint, wherein determining the detection area in the second frame of the digital human animation based on the predicted positions of the multiple joints in the second frame of the digital human animation includes: determining the predicted distance between the first joint and the second joint based on the predicted positions of the multiple joints in the second frame of the digital human animation; and when the predicted distance is less than or equal to a preset threshold, determining that the detection area in the second frame of the digital human animation includes the area between the predicted position of the first joint and the predicted position of the second joint.

[0013] The above scheme can determine the detection area of ​​the second frame by comparing the predicted distances between multiple joints with the preset threshold, thereby improving the accuracy of mold penetration detection.

[0014] In combination with the first aspect, in some implementations of the first aspect, the second frame is a next frame of the first frame.

[0015] Based on the above scheme, the detection area of ​​the next frame is predicted based on the skeletal animation data of the current frame. Since the position information and speed information of the joints in the current frame are highly correlated with the area where the collision may occur in the next frame, the method provided in the embodiment of the present application can improve the accuracy of determining the detection area, so that the detection area can better reflect the range of the area where the collision may occur, thereby improving the accuracy of the penetration detection.

[0016] In combination with the first aspect, in certain implementations of the first aspect, the model of the digital human includes multiple face patches on the surface of the digital human, wherein, within the detection area in the second frame, based on the second skeletal animation data, the model of the digital human is subjected to model penetration detection, including: based on the second skeletal animation data, performing an intersection operation on a target face patch among the multiple face patches on the surface of the digital human to perform model penetration detection, wherein the target face patch is a face patch within the detection area of ​​the second frame.

[0017] Based on the above solution, the intersection operation is performed on the facets within the area where the collision may occur, which reduces the amount of calculation, thereby improving the speed of mold penetration detection. In addition, due to the high speed of mold penetration detection, the method provided in the embodiment of the present application can realize real-time mold penetration detection.

[0018] In combination with the first aspect, in certain implementations of the first aspect, the method also includes: obtaining third skeletal animation data of the digital human, the third skeletal animation data including third position information and third motion information of the multiple joints of the digital human in a third frame of the digital human animation, the third skeletal animation data being determined based on third motion data of a real person collected by the motion capture device, wherein the third frame is before the first frame; determining a detection area in a second frame of the digital human animation based on the first skeletal animation data, including: determining the detection area in the second frame of the digital human animation based on the first skeletal animation data and the third skeletal animation data.

[0019] In the embodiment of the present application, the detection area can be determined based on multiple frames of skeletal animation data before the frame to be detected. Compared with the solution of determining the detection area based on only one frame of skeletal animation data, the method provided by the present application improves the accuracy of determining the detection area, so that the detection area can better reflect the range of the area where the collision may occur, thereby improving the accuracy of the penetration detection.

[0020] In combination with the first aspect, in some implementations of the first aspect, the motion capture device includes at least one of an optical sensor, an inertial sensor, or a visual sensor.

[0021] In the embodiments of the present application, the motion data of a real person obtained by an optical sensor, an inertial sensor or a visual sensor has high precision and can improve the accuracy of motion capture.

[0022] In a second aspect, an embodiment of the present application provides a mold penetration detection device, which includes a module for implementing the first aspect or any possible implementation method of the first aspect.

[0023] In a third aspect, a computing device cluster is provided, comprising at least one computing device, each computing device comprising a processor and a memory; the processor of the at least one computing device is used to execute instructions stored in the memory of the at least one computing device, so that the computing device cluster executes the mold penetration detection method of the first aspect or any possible implementation of the first aspect.

[0024] In a fourth aspect, a computer program product comprising instructions is provided. When the instructions are executed by a computing device cluster, the computing device cluster executes the mold penetration detection method of the first aspect or any possible implementation of the first aspect.

[0025] In a fifth aspect, a computer-readable storage medium is provided, comprising computer program instructions. When the computer program instructions are executed by a computing device cluster, the computing device cluster executes the mold penetration detection method of the first aspect or any possible implementation of the first aspect. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 This is a schematic diagram of the digital human model wearing the mold.

[0027] Figure 2 It is a schematic flow chart of a mold penetration detection method provided in an embodiment of the present application.

[0028] Figure 3 It is a schematic block diagram of a motion capture system provided in an embodiment of the present application.

[0029] Figure 4 It is a schematic block diagram of a mold penetration detection device provided in an embodiment of the present application.

[0030] Figure 5 It is a schematic block diagram of a computing device provided in an embodiment of the present application.

[0031] Figure 6 It is a schematic block diagram of a computing device cluster provided in an embodiment of the present application.

[0032] Figure 7 It is a schematic block diagram of another computing device cluster provided in an embodiment of the present application. DETAILED DESCRIPTION

[0033] The technical solution in this application will be described below in conjunction with the accompanying drawings.

[0034] Digital human (DH) can also be called virtual human. Digital human refers to a digital human with human or human-like appearance and behavior patterns created through technological means such as modeling, motion capture, and artificial intelligence (AI), and presented through display devices.

[0035] In a scene driven by a 3D digital human, skeletal animation data is an important way to drive the movement of the digital human. The driven digital human model can include bones and skins. The bones can also be called armatures.

[0036] For digital humans, skeletons can be understood as human-like bone frames composed of digital humans' pelvis, waist bones, sternum, etc. Skeletons are composed of rigid joints with multiple levels. Among the multiple levels of joints, joints at a higher level can be called parent joints, and joints at a lower level can be called child joints. Among them, the movement of the parent joint can affect the movement of the child joint, and the movement of the child joint does not affect the movement of the parent joint. For example, taking the human skeleton as a reference, the elbow joint and the wrist joint are connected by the ulna and radius. The elbow joint can be regarded as the parent joint, and the wrist joint can be regarded as the child joint. The movement of the elbow joint often affects the movement of the wrist joint, while the movement of the wrist joint does not affect the movement of the elbow joint. At least two joints can be connected by bones. The relationship between the joints and bones of digital humans is similar to the relationship between the joints and bones of humans. The joints of digital humans can rotate, move, and so on, so that the bones also move with the movement of the joints, and the entire skeleton also moves with the movement of the joints. Therefore, it can be understood that the joints drive the bones and bones.

[0037] The skin can be understood as the surface of the digital human, similar to human skin. Just as human skin moves with the movement of the skeleton, the skin also moves with the movement of the skeleton, so it can be understood that the skeleton drives the skin. The skin can include multiple patches, and the patch can be a polygon composed of at least three vertices. In some embodiments, multiple patches constitute the skin.

[0038] Motion capture technology can collect real-life motion data or movement information through optical sensors, inertial sensors, or visual sensors (e.g., cameras), and then generate skeletal animation data of a digital human corresponding to the real person (e.g., rotation information of joints, position information, etc.). Based on the skeletal animation data, the digital human model can be driven. However, the surface of the driven digital human model is prone to the problem of penetration.

[0039] Figure 1 This is a schematic diagram of the digital human model wearing the model. Figure 1 , Figure 1 The left part of the figure shows the skeleton obtained by motion capture and the preview of the digital human model corresponding to the skeleton. Figure 1 As shown in the circle on the left part of the figure, there is an obvious penetration between the left hand of the digital human and the body. If the penetration detection can detect this problem, the real person corresponding to the digital human can be reminded to adjust the action so that the left hand of the digital human model does not penetrate the body. The preview image of the skeleton without penetration and the digital human corresponding to the skeleton is shown in the figure below. Figure 1 shown in the right part of .

[0040] The related technical solution can detect mold penetration by using the facet intersection method. Facet intersection can calculate the intersection of two faces to determine whether there is an intersection between the two faces. If there is an intersection, it can be determined that mold penetration has occurred.

[0041] However, the number of facets in the digital human is huge, which results in a huge amount of computation for facet intersection and slow detection of penetration. This makes it impossible for the real person corresponding to the digital human to know in real time whether penetration has occurred during the motion capture process. This makes it impossible to directly apply the skeletal animation data obtained through motion capture to the digital human to be driven, and it needs to be manually modified by technicians, which increases time and cost.

[0042] Therefore, how to avoid the penetration of digital human models is an urgent problem to be solved.

[0043] Figure 2 2 is a schematic flow chart of a mold penetration detection method 200 provided in an embodiment of the present application. The method 200 can prevent the digital human model from being penetrated. The method 200 can be executed by any device with computing capabilities, such as a computer. Figure 2 Method 200 is introduced.

[0044] S210, obtaining first skeleton animation data of a digital human.

[0045] The first skeletal animation data may include first position information and first motion information of multiple joints of the digital human in a first frame of the digital human animation, and the first skeletal animation data may be determined based on first motion data of a real person collected by a motion capture device.

[0046] The first skeleton animation data may be skeleton animation data in the first frame of the digital human animation. The first frame may be any frame in the digital human animation. The first position information and the first motion information may be position information and motion information included in the first skeleton animation data, respectively.

[0047] The position information may include joint coordinates and joint directions. The joint coordinates of a joint are used to indicate the position of the joint relative to the coordinate origin, and the joint direction of a joint is used to indicate the numerical value of the angle difference between the joint and the parent joint of the joint. In addition, the position information of the joint may be the position information of the joint in a three-dimensional (3D) space, or it may be the position information in a two-dimensional (2D) plane. The above-mentioned 2D plane may be the projection of the skeleton in the 3D space under multiple perspectives. For example, multiple perspectives may include front, back, left, right and other perspectives.

[0048] The motion information may also be referred to as a motion vector. The motion information of a joint may include linear velocity, angular velocity, acceleration, etc. In a first frame, the motion information of a joint may indicate the motion trend of the joint in the first frame.

[0049] Exemplarily, the multiple joints of the digital human include joint #1 and joint #2. The first skeletal animation data may include position information and motion information of joint #1 in the first frame of the digital human animation, and the first skeletal animation data may also include position information and motion information of joint #2 in the first frame of the digital human animation.

[0050] It should be noted that the present application does not limit the skeletal animation data to only include the position information and speed information of the joints, and the skeletal animation data may also include other information. For example, the skeletal animation data may also include information such as the labels of the joints.

[0051] In some embodiments, the motion capture device includes at least one of an optical sensor, an inertial sensor, or a visual sensor.

[0052] The process of motion capture using an optical sensor can also be called optical motion capture. In the process of optical motion capture, multiple marker points can be set at key parts of the physical object (such as the crotch and elbow of a person). The optical sensor can receive light signals from each marker point, and the skeletal animation data can be determined based on the received light signals. Among them, the marker point can be actively luminous, such as using a light emitting diode (LED) as a marker point; the marker point can be passively luminous, such as using a reflective material as a marker point. Exemplarily, the optical sensor can be an infrared optical camera.

[0053] The process of using inertial sensors for motion capture can also be called inertial motion capture. In the process of inertial motion capture, inertial sensors can be set at key parts of the physical object (such as the hips and elbows of a person). When the physical object moves, the inertial sensor can collect inertial data such as angular velocity, acceleration, velocity direction, etc. of the key parts, and the skeletal animation data can be determined based on these inertial data.

[0054] The visual sensor can also be called a camera. The visual sensor can record the physical object visually to obtain video data. The skeletal animation data can be determined by analyzing the video data.

[0055] It should be noted that the present application does not limit the motion capture device to only include optical sensors, inertial sensors or visual sensors. The motion capture device may also include other devices, such as electromagnetic sensors, acoustic sensors, etc. The electromagnetic sensor can determine the change of the electromagnetic field, and thus determine the motion data according to the change of the electromagnetic field. The acoustic sensor can receive sound, and thus determine the motion data according to the received sound.

[0056] In the embodiments of the present application, the motion data of a real person obtained by an optical sensor, an inertial sensor or a visual sensor has high precision and can improve the accuracy of motion capture.

[0057] S220: Determine a detection area in a second frame of the digital human animation according to the first skeletal animation data.

[0058] The model of the digital human in the detection area is expected to be penetrated, and the second frame may be after the first frame.

[0059] In some optional embodiments, the second frame may be the next frame of the first frame. The next frame of the first frame may be understood as a frame after the first frame and adjacent to the first frame. That is to say, the detection area of ​​the next frame is predicted based on the skeletal animation data of the current frame. Since the position information and motion information of the joints in the current frame are highly correlated with the area where the collision may occur in the next frame, the above scheme can improve the accuracy of determining the detection area, so that the detection area can better reflect the range of the area where the collision may occur, thereby improving the accuracy of the penetration detection.

[0060] It should be noted that the present application does not limit the second frame to only be the next frame of the first frame. In other optional embodiments, the second frame can be any frame after the first frame. It can be understood by those skilled in the art that the position information and motion information of the joint in the current frame have a certain correlation with the area where the collision may occur in any frame after the current frame.

[0061] The detection area of ​​the second frame can be understood as the mold penetration detection range of the second frame. In other words, in the second frame, the mold penetration detection is performed on the detection area. The detection area indicates the spatial range where the mold penetration may occur.

[0062] It should be noted that the detection area can be a one-dimensional linear area, a two-dimensional plane area, or a three-dimensional stereoscopic area.

[0063] For example, the position information and motion information included in the first skeleton animation data can be used as input, and the detection area of ​​the second frame can be output through the neural network. The present application does not limit the specific form of the neural network, for example, it can be a long short term memory (LSTM) neural network.

[0064] In some embodiments, S220 includes: determining the predicted positions of the multiple joints in the second frame of the digital human animation based on the first position information and the first motion information of the multiple joints of the digital human; and determining the detection area in the second frame of the digital human animation based on the predicted positions of the multiple joints in the second frame of the digital human animation.

[0065] It should be noted that the predicted position is different from the actual position. The predicted position of the second frame can be the possible position of the second frame predicted by the first skeletal animation data of the first frame. When the predicted position of the second frame is obtained, the time has not yet reached the second frame. For example, the time may still stay in the first frame, or a frame between the first frame and the second frame. The actual position of the second frame is indicated by the second position information included in the second skeletal animation data of the second frame. When the actual position of the second frame is obtained, the time is in the second frame.

[0066] For example, the first position information and the first motion information can be understood as initial conditions, and the predicted position information corresponding to the initial conditions can be calculated according to a physical formula. For example, the following physical formula can be used for calculation.

[0067]

[0068] Wherein, S2 represents the predicted position information of the second frame, S1 represents the joint coordinates in the first position information, V1 represents the linear velocity in the first motion information, t represents the time difference between the first frame and the second frame, and a represents the acceleration in the first motion information. It can be understood that the above linear velocity and acceleration are directional vectors corresponding to the joint coordinates.

[0069] For another example, the first position information and the first motion information may be used as input, and the predicted position of the second frame may be output through a neural network. The present application does not limit the specific form of the neural network, for example, it may be an LSTM neural network.

[0070] After the predicted position of the second frame is obtained, the detection area can be determined based on the predicted position. For example, if the predicted position of a joint is close to the predicted positions of other joints, it can be considered that the area around the positions indicated by these predicted positions is the detection area. For another example, based on the predicted position and the preset range, it can be determined that the area outside the predicted position and within the preset range is the detection area. The preset range can be one or more.

[0071] The above scheme first determines the predicted position of the second frame and then determines the detection area of ​​the second frame, which can accurately determine the detection area, thereby improving the accuracy of mold penetration detection.

[0072] In some embodiments, the multiple joints include a first joint and a second joint, wherein determining a detection area in the second frame of the digital human animation based on the predicted positions of the multiple joints in the second frame of the digital human animation includes: determining a predicted distance between the first joint and the second joint based on the predicted positions of the multiple joints in the second frame of the digital human animation; and determining, when the predicted distance is less than or equal to a preset threshold, that the detection area in the second frame of the digital human animation includes an area between the predicted position of the first joint and the predicted position of the second joint.

[0073] The first joint may be any one of the multiple joints of the digital human. The second joint may be any one of the multiple joints of the digital human. The first joint and the second joint may be different joints.

[0074] Exemplarily, the predicted position of the first joint is (0, 1), and the predicted position of the second joint is (0, 5), then the predicted distance between the first joint and the second joint may be 4.

[0075] The above scheme can be understood as follows: when the predicted distance between two joints is small, the computer can determine that the two joints may collide, and thus perform a penetration detection between the two joints.

[0076] The preset threshold may be preset or determined according to a preset rule.

[0077] Different joint combinations may correspond to different preset thresholds. For example, the distance between the facet corresponding to the joint of the digital human's "elbow" and the joint is small, so the joint may correspond to a lower preset threshold. For another example, the distance between the facet corresponding to the joint of the digital human's "knee" and the joint is medium, so the joint may correspond to a medium preset threshold. For another example, the distance between the facet corresponding to the joint of the digital human's "hip" and the joint is large, so the joint may correspond to a larger preset threshold. In this way, when the joint of the "elbow" and the joint of the "knee" are combined as a joint, the preset threshold is small; when the joint of the "elbow" and the joint of the "hip" are combined as a joint, the preset threshold is large.

[0078] In other words, since the "arms" of the digital human are thinner than the "legs" of the digital human, and the "legs" of the digital human are thinner than the "waist" of the digital human. Therefore, when the predicted distance between the "arms" and "legs" of the digital human is not too large, the "arms" and "legs" may not penetrate the model, and there is no need to use the area between the "arms" and "legs" as the detection area. When the predicted distance between the "arms" and "waist" of the digital human is the same as the predicted distance between the "arms" and "legs" mentioned above, the "arms" and "waist" may have penetrated the model, and therefore the area between the "arms" and "waist" needs to be used as the detection area. The above description is only for ease of understanding and does not constitute a limitation of this application.

[0079] Different digital humans may correspond to different preset thresholds. For example, a digital human with a fatter body may correspond to a larger preset threshold, while a digital human with a thinner body may correspond to a smaller preset threshold.

[0080] The detection area includes the area between the predicted position of the first joint and the predicted position of the second joint. For example, the detection area may include the line between the predicted position of the first joint and the predicted position of the second joint. However, the present application does not limit the detection area to only include the area between the predicted position of the first joint and the predicted position of the second joint. For example, the detection area may be the predicted position of the first joint, the predicted position of the second joint, and the up, down, left, right, front, and back extensions of the line between the predicted position of the first joint and the predicted position of the second joint.

[0081] The above scheme can determine the detection area of ​​the second frame by comparing the predicted distances between multiple joints with the preset threshold, thereby improving the accuracy of mold penetration detection.

[0082] In some embodiments, the multiple joints include a first joint, a second joint, and a third joint, and the first joint and the second joint are connected by a bone, wherein determining the detection area in the second frame of the digital human animation based on the predicted positions of the multiple joints in the second frame of the digital human animation includes: determining the predicted distance between the third joint and the predicted position of the first point on the bone based on the predicted positions of the multiple joints in the second frame of the digital human animation; and when the predicted distance is less than or equal to a preset threshold, determining that the detection area in the second frame of the digital human animation includes the area between the predicted position of the third joint and the predicted position of the first point on the bone.

[0083] The first point on the bone can be any point on the bone. In some optional embodiments, the distance from the first point to the predicted position of the third joint is less than or equal to the distance from other points on the bone to the predicted position of the third joint. In other words, the first point can be the intersection of a vertical line from the predicted position of the third joint to the bone and the bone.

[0084] The above scheme can be understood as follows: when the predicted distance between a joint and a bone is small, the computer can determine that the joint may collide with the bone, and thus perform a penetration detection between the joint and the bone.

[0085] The preset threshold may be pre-set or determined according to a preset rule. Different combinations of joints and bones may correspond to different preset thresholds.

[0086] The detection area includes the area between the predicted position of the third joint and the predicted position of the first point on the bone. For example, the detection area may include the line between the predicted position of the third joint and the predicted position of the first point. However, the present application does not limit the detection area to only include the area between the predicted position of the third joint and the predicted position of the first point. For example, the detection area can be the predicted position of the third joint, the predicted position of the first point, and the up, down, left, right, front, and back extensions of the line between the predicted position of the third joint and the predicted position of the first point.

[0087] S230, obtaining second skeletal animation data of the digital human.

[0088] The second skeletal animation data may include second position information and second motion information of the multiple joints of the digital human in the second frame of the digital human animation, and the second skeletal animation data may be determined based on second motion data of the real person collected by the motion capture device.

[0089] The meanings of the second skeletal animation data, the second position information, the second motion information, and the second action data are similar to those of the aforementioned first skeletal animation data, the first position information, the first motion information, and the first action data. The difference is that the second skeletal animation data is the skeletal animation data of the second frame of the digital human animation. The second position information and the second motion information are the position information and motion information included in the second skeletal animation data. The second action data can be used to determine the second skeletal animation data. For other descriptions, please refer to the aforementioned S210, which will not be repeated here.

[0090] S240: Performing a penetration detection on the digital human model in the detection area of ​​the second frame according to the second skeletal animation data.

[0091] It is worth noting that in the second frame, the model penetration detection is not performed on the entire area of ​​the digital human, but on the model of the digital human in the detection area, thereby reducing the amount of calculation.

[0092] The second skeletal animation data can drive the digital human model, so that the present solution can perform model penetration detection on the digital human model.

[0093] In some embodiments, the detection area of ​​the second frame is also determined based on the skeletal animation data of the third frame, and the third frame is before the first frame. In other words, the method 200 also includes: obtaining the third skeletal animation data of the digital human, the third skeletal animation data includes the third position information and the third motion information of the multiple joints of the digital human in the third frame of the digital human animation, and the third skeletal animation data is determined based on the third motion data of the real person collected by the motion capture device, wherein the third frame is before the first frame; S220, including: determining the detection area in the second frame of the digital human animation based on the first skeletal animation data and the third skeletal animation data.

[0094] The meanings of the third skeletal animation data, the third position information, the third motion information and the third action data are similar to the meanings of the aforementioned first skeletal animation data, the first position information, the first motion information and the first action data. The difference is that the third skeletal animation data is the skeletal animation data of the third frame of the digital human animation. The third position information and the third motion information are the position information and motion information included in the third skeletal animation data. The third action data can be used to determine the third skeletal animation data. For other descriptions, please refer to the aforementioned S210, which will not be repeated here.

[0095] The third frame is any frame before the first frame, or any multiple frames. In some embodiments, the third frame is the previous frame of the first frame. The previous frame of the first frame can be understood as a frame before the first frame and adjacent to the first frame.

[0096] It is understandable that the detection area of ​​the second frame can be determined not only by the skeletal animation data of the first frame, but also by the skeletal animation data before the first frame. In some embodiments, the detection area of ​​the second frame can be determined by the skeletal animation data of at least one frame before the second frame. In some embodiments, the detection area of ​​the second frame can be determined by the skeletal animation data of at least two frames before the second frame. In some embodiments, the detection area of ​​the second frame can be determined by the skeletal animation data of the first frame and the third frame before the second frame, wherein the first frame is the previous frame of the second frame, and the third frame is the previous frame of the first frame.

[0097] In the embodiment of the present application, the detection area can be determined based on multiple frames of skeletal animation data before the frame to be detected. Compared with the solution of determining the detection area based on only one frame of skeletal animation data, the method provided by the present application improves the accuracy of determining the detection area, so that the detection area can better reflect the range of the area where the collision may occur, thereby improving the accuracy of the penetration detection.

[0098] In some embodiments, the model of the digital human includes multiple face patches on the surface of the digital human, and S230 includes: according to the second skeletal animation data, performing an intersection operation on a target face patch among the multiple face patches on the surface of the digital human to perform model penetration detection, wherein the target face patch is a face patch within the detection area of ​​the second frame.

[0099] It should be noted that the target face is the face within the detection area of ​​the second frame. In other words, the face intersection is for the target face within the detection area of ​​the second frame. It is understandable that the target face can represent a part of all the faces of the digital human model. In some embodiments, the face outside the detection area is not subjected to the intersection operation.

[0100] The face intersection calculation can be used to calculate the intersection of two faces to determine whether there is an intersection between the two faces. If there is an intersection, it can be determined that a mold penetration has occurred.

[0101] The second skeletal animation data can drive the digital human model, thereby determining the positions of multiple patches on the surface of the digital human.

[0102] Based on the above solution, the intersection operation is performed on the facets within the area where the collision may occur, which reduces the amount of calculation, thereby improving the speed of mold penetration detection. In addition, due to the high speed of mold penetration detection, the method provided in the embodiment of the present application can realize real-time mold penetration detection.

[0103] S250: When it is detected that the model of the digital human is penetrated, feedback information is sent.

[0104] The feedback information may include position information of a penetration area of ​​the digital human model in the second frame, and the feedback information may be used to instruct the real person to adjust the action.

[0105] Feedback information can be sent to any device. For example, feedback information can be sent to a sensor, which can inform the real person corresponding to the digital human by making sounds, vibrating, etc., thereby guiding the real person to change actions, adjust postures or movement trajectories to avoid penetration. For another example, feedback information can be sent to a device that processes skeletal animation data. Further, the feedback information can include the position of the surface of the digital human where penetration occurs, or the feedback information can include the part of the real person corresponding to the position of the surface of the digital human where penetration occurs. The feedback information can be displayed to the user in the form of a digital human preview interface.

[0106] As an example, the feedback information can directly indicate the adjustment action of the real person corresponding to the digital person. As another example, the feedback information can be used to warn the real person corresponding to the digital person of the model penetration, thereby indirectly indicating the adjustment action of the real person corresponding to the digital person.

[0107] In the embodiment of the present application, the range of the area where the collision may occur is first predicted by the skeletal animation data, and then the penetration detection is performed on the range of the area where the collision may occur, which reduces the amount of calculation, thereby increasing the speed of the penetration detection. In addition, since the penetration detection speed is fast, the method provided in the embodiment of the present application can realize real-time penetration detection, and promptly remind real people to adjust their movements, thereby avoiding the penetration of the digital human model. In addition, since the method provided in the embodiment of the present application reduces the penetration of the skeletal animation data obtained by motion capture, the skeletal animation data can be applied to the digital human that is expected to be driven, without the need for manual modification by technical personnel, thereby reducing time and cost.

[0108] Figure 3 It is a schematic block diagram of a motion capture system 300 provided in an embodiment of the present application. Figure 3 It includes a data acquisition module 310, a data processing module 320, a mold penetration detection module 330 and a visualization module 340.

[0109] The data acquisition module 310 can collect motion data of a real person. For example, the data acquisition module 310 can include an optical sensor, an inertial sensor, a visual sensor, and the like.

[0110] The data processing module 320 can process the motion data of a real person to obtain skeletal animation data. For example, the data processing module 320 can determine the position information and motion information of the joints according to the motion data.

[0111] The mold penetration detection module 330 may execute the above method 200. For example, the mold penetration detection module 330 may determine the detection area and perform the operation of patch intersection.

[0112] The visualization module 340 can perform visualization processing based on the skeletal animation data. For example, operations such as skeletal skinning rendering and animation rendering can be performed on the skeletal animation data to present a preview image to the user. Figure 1 Some results of the visualization process are shown.

[0113] The method embodiments of the present application are described in detail above. Figures 4 to 7 The device embodiment of the embodiment of the present application is described. The device embodiment corresponds to the method embodiment, so the parts that are not described in detail can refer to the previous method embodiments. The device can implement any possible implementation method of the above method.

[0114] Figure 4 4 is a schematic block diagram of a mold penetration detection device 400 provided in an embodiment of the present application. The device 400 may be the mold penetration detection module 330 described above, and the device 400 may execute the method 200 described above.

[0115] The penetration detection device 400 includes: an acquisition module 410, which is used to acquire first skeletal animation data of a digital human, wherein the first skeletal animation data includes first position information and first motion information of multiple joints of the digital human in a first frame of the digital human animation, and the first skeletal animation data is determined according to first motion data of a real person collected by a motion capture device; a processing module 420, which is used to determine a detection area in a second frame of the digital human animation according to the first skeletal animation data, wherein the model of the digital human in the detection area is expected to penetrate the model, and the second frame is after the first frame; the acquisition module 410 is also used to acquire a second skeletal animation data of the digital human The second skeletal animation data includes second position information and second motion information of the multiple joints of the digital human in the second frame of the digital human animation, and the second skeletal animation data is determined according to the second motion data of the real person collected by the motion capture device; a detection module 430 is used to perform a penetration detection on the model of the digital human according to the second skeletal animation data within the detection area in the second frame; and a feedback module 440 is used to send feedback information when it is detected that the model of the digital human has penetrated the model, and the feedback information includes the position information of the penetration area of ​​the model of the digital human in the second frame, and the feedback information is used to instruct the real person to adjust the action.

[0116] In a possible embodiment, the processing module 420 is specifically used to: determine the predicted positions of the multiple joints in the second frame of the digital human animation according to the first position information and the first motion information of the multiple joints of the digital human; and determine the detection area in the second frame of the digital human animation according to the predicted positions of the multiple joints in the second frame of the digital human animation.

[0117] In a possible embodiment, the multiple joints include a first joint and a second joint, wherein the processing module 410 is specifically used to: determine the predicted distance between the first joint and the second joint according to the predicted positions of the multiple joints in the second frame of the digital human animation; and when the predicted distance is less than or equal to a preset threshold, determine that the detection area in the second frame of the digital human animation includes the area between the predicted position of the first joint and the predicted position of the second joint.

[0118] In a possible embodiment, the model of the digital human includes multiple face patches on the surface of the digital human, wherein the detection module 430 is specifically used to: perform an intersection operation on a target face patch among the multiple face patches on the surface of the digital human according to the second skeletal animation data to perform model penetration detection, wherein the target face patch is a face patch within the detection area of ​​the second frame.

[0119] In a possible embodiment, the acquisition module 410 is also used to acquire third skeletal animation data of the digital human, the third skeletal animation data including third position information and third motion information of the multiple joints of the digital human in a third frame of the digital human animation, the third skeletal animation data being determined based on third motion data of a real person collected by the motion capture device, wherein the third frame is before the first frame; wherein the processing module 420 is specifically used to determine the detection area in the second frame of the digital human animation based on the first skeletal animation data and the third skeletal animation data.

[0120] Specifically, the specific implementation of various operations of the mold penetration detection method performed by the mold penetration detection device 400 can refer to the description of the relevant contents in the above method embodiment, which will not be repeated here.

[0121] Among them, the acquisition module 410, the processing module 420, the detection module 430 and the feedback module 440 can all be implemented by software, or can be implemented by hardware. Exemplarily, the implementation of the acquisition module 410 is introduced below by taking the acquisition module 410 as an example. Similarly, the implementation of the processing module 420, the detection module 430 and the feedback module 440 can refer to the implementation of the acquisition module 410.

[0122] As an example of a software functional unit, the acquisition module 410 may include code running on a computing instance. Among them, the computing instance may include at least one of a physical host (computing device), a virtual machine, and a container. Further, the above-mentioned computing instance may be one or more. For example, the acquisition module 410 may include code running on multiple hosts / virtual machines / containers. It should be noted that the multiple hosts / virtual machines / containers used to run the code may be distributed in the same region (region) or in different regions. Furthermore, the multiple hosts / virtual machines / containers used to run the code may be distributed in the same availability zone (AZ) or in different AZs, each AZ including a data center or multiple data centers with close geographical locations. Among them, usually a region may include multiple AZs.

[0123] Similarly, multiple hosts / virtual machines / containers used to run the code can be distributed in the same virtual private cloud (VPC) or in multiple VPCs. Usually, a VPC is set up in a region. For cross-region communication between two VPCs in the same region and between VPCs in different regions, a communication gateway needs to be set up in each VPC to achieve interconnection between VPCs through the communication gateway.

[0124] As an example of a hardware functional unit, the acquisition module 410 may include at least one computing device, such as a server, etc. Alternatively, the acquisition module 410 may also be a device implemented using an application-specific integrated circuit (ASIC) or a programmable logic device (PLD). The PLD may be a complex programmable logical device (CPLD), a field-programmable gate array (FPGA), a generic array logic (GAL) or any combination thereof.

[0125] The multiple computing devices included in the acquisition module 410 can be distributed in the same region or in different regions. The multiple computing devices included in the acquisition module 410 can be distributed in the same AZ or in different AZs. Similarly, the multiple computing devices included in the acquisition module 410 can be distributed in the same VPC or in multiple VPCs. The multiple computing devices can be any combination of computing devices such as servers, ASICs, PLDs, CPLDs, FPGAs, and GALs.

[0126] It should be noted that, in other embodiments, the acquisition module 410 can be used to execute any step in the mold penetration detection method, the processing module 420 can be used to execute any step in the mold penetration detection method, the detection module 430 can be used to execute any step in the mold penetration detection method, and the feedback module 440 can be used to execute any step in the mold penetration detection method. The steps that the acquisition module 410, the processing module 420, the detection module 430, and the feedback module 440 are responsible for implementing can be specified as needed, and the acquisition module 410, the processing module 420, the detection module 430, and the feedback module 440 respectively implement different steps in the mold penetration detection method to achieve all the functions of the mold penetration detection device 400.

[0127] The present application also provides a computing device 500. Figure 5 As shown, computing device 500 includes: bus 502, processor 504, memory 506 and communication interface 508. Processor 504, memory 506 and communication interface 508 communicate through bus 502. Computing device 500 can be a server or a terminal device. It should be understood that the present application does not limit the number of processors and memories in computing device 500.

[0128] The bus 502 may be a peripheral component interconnect (PCI) bus or an extended industry standard architecture (EISA) bus. The bus may be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 5 The bus 502 is represented by only one line, but it does not mean that there is only one bus or one type of bus. The bus 502 may include a path for transmitting information between various components of the computing device 500 (eg, the memory 506, the processor 504, and the communication interface 508).

[0129] The processor 504 may include any one or more of a central processing unit (CPU), a graphics processing unit (GPU), a microprocessor (MP), or a digital signal processor (DSP).

[0130] The memory 506 may include a volatile memory, such as a random access memory (RAM). The memory 506 may also include a non-volatile memory, such as a read-only memory (ROM), a flash memory, a hard disk drive (HDD), or a solid state drive (SSD).

[0131] The memory 506 stores executable program codes, and the processor 504 executes the executable program codes to respectively implement the functions of the aforementioned acquisition module 410, processing module 420, detection module 430, and feedback module 440, thereby implementing the mold penetration detection method 200. That is, the memory 506 stores instructions for executing the mold penetration detection method 200.

[0132] The communication interface 508 uses a transceiver module such as, but not limited to, a network interface card or a transceiver to implement communication between the computing device 500 and other devices or a communication network.

[0133] The embodiment of the present application also provides a computing device cluster. The computing device cluster includes at least one computing device. The computing device can be a server, such as a central server, an edge server, or a local server in a local data center. In some embodiments, the computing device can also be a terminal device such as a desktop computer, a laptop computer, or a smart phone.

[0134] like Figure 6 As shown, the computing device cluster includes at least one computing device 500. The memory 506 in one or more computing devices 500 in the computing device cluster may store the same instructions for executing the mold penetration detection method 200.

[0135] In some embodiments, the memory 506 of one or more computing devices 500 in the computing device cluster may also respectively store partial instructions for executing the mold penetration detection method 200. In other words, the combination of one or more computing devices 500 may jointly execute instructions for executing the mold penetration detection method 200.

[0136] It should be noted that the memory 506 in different computing devices 500 in the computing device cluster can store different instructions, which are respectively used to execute part of the functions of the mold penetration detection device 400. That is, the instructions stored in the memory 506 in different computing devices 500 can realize the functions of one or more modules among the acquisition module 410, the processing module 420, the detection module 430 and the feedback module 440.

[0137] In some embodiments, one or more computing devices in the computing device cluster may be connected via a network, which may be a wide area network or a local area network, etc.

[0138] Figure 7 A possible implementation method is proposed. Figure 7 As shown, two computing devices 500A and 500B are connected via a network. Specifically, the network is connected via a communication interface in each computing device. In this type of possible implementation, the memory 506 in the computing device 500A stores instructions for executing the functions of the acquisition module 410, the detection module 430, and the feedback module 440. At the same time, the memory 506 in the computing device 500B stores instructions for the functions of the processing module 420.

[0139] Figure 7 The connection method between the computing device clusters shown may be based on the consideration that the mold penetration detection method 200 provided in the present application requires a large number of detection areas to be determined, and therefore the functions implemented by the processing module 420 may be handed over to the computing device 500B for execution.

[0140] It should be understood that Figure 7The functions of the computing device 500A shown in FIG. 5A may also be completed by multiple computing devices 500. Similarly, the functions of the computing device 500B may also be completed by multiple computing devices 500.

[0141] In some embodiments, the memory 506 of one or more computing devices 500 in the computing device cluster may also respectively store partial instructions for executing the mold penetration detection method 200. In other words, the combination of one or more computing devices 500 may jointly execute the instructions for executing the mold penetration detection method 200.

[0142] The present application also provides a computer program product including instructions. The computer program product may be software or a program product including instructions that can be run on a computing device or stored in any available medium. When the computer program product is run on at least one computing device, the at least one computing device executes the above-mentioned mold penetration detection method 200.

[0143] The embodiment of the present application also provides a computer-readable storage medium. The computer-readable storage medium can be any available medium that can be stored by a computing device or a data storage device such as a data center that contains one or more available media. The available medium can be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium (e.g., a solid-state hard disk). The computer-readable storage medium includes instructions that instruct the computing device to execute the above-mentioned penetration detection method.

[0144] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0145] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0146] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0147] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0148] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0149] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in the various embodiments of the present application. The aforementioned storage media include: various media that can store program codes, such as USB flash drives, mobile hard disks, read-only memories (ROM), random access memories (RAM), magnetic disks or optical disks.

[0150] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A mold penetration detection method, characterized in that: The method comprises: Acquire first skeleton animation data of a digital human, wherein the first skeleton animation data includes first position information and first motion information of a plurality of joints of the digital human in a first frame of the digital human animation, and the first skeleton animation data is determined according to first motion data of a real person collected by a motion capture device; Determine, according to the first skeletal animation data, a detection area in a second frame of the digital human animation, wherein the model of the digital human in the detection area is expected to be penetrated, and the second frame is after the first frame; Acquire second skeletal animation data of the digital human, wherein the second skeletal animation data includes second position information and second motion information of the plurality of joints of the digital human in a second frame of the digital human animation, and the second skeletal animation data is determined according to second motion data of the real person collected by the motion capture device; In the detection area in the second frame, performing a penetration detection on the model of the digital human according to the second skeletal animation data; When it is detected that the model of the digital human is penetrated, feedback information is sent, the feedback information includes position information of the penetrated area of ​​the model of the digital human in the second frame, and the feedback information is used to instruct the real person to adjust the action.

2. The method according to claim 1, characterized in that The step of determining the detection area in the second frame of the digital human animation according to the first skeletal animation data comprises: Determining predicted positions of the multiple joints in a second frame of the digital human animation according to the first position information and the first motion information of the multiple joints of the digital human; According to the predicted positions of the multiple joints in the second frame of the digital human animation, a detection area in the second frame of the digital human animation is determined.

3. The method according to claim 2, characterized in that The multiple joints include a first joint and a second joint, wherein determining the detection area in the second frame of the digital human animation according to the predicted positions of the multiple joints in the second frame of the digital human animation includes: Determining a predicted distance between the first joint and the second joint according to predicted positions of the plurality of joints in a second frame of the digital human animation; When the predicted distance is less than or equal to a preset threshold, it is determined that the detection area in the second frame of the digital human animation includes an area between the predicted position of the first joint and the predicted position of the second joint.

4. The method according to any one of claims 1 to 3, characterized in that The model of the digital human includes a plurality of face patches on the surface of the digital human, wherein, in the detection area in the second frame, the model of the digital human is subjected to a penetration detection according to the second skeletal animation data, including: According to the second skeletal animation data, an intersection operation is performed on a target facet among the multiple facets on the surface of the digital human to perform mold penetration detection, wherein the target facet is a facet within the detection area of ​​the second frame.

5. The method according to any one of claims 1 to 4, characterized in that The method further comprises: Acquire third skeletal animation data of the digital human, the third skeletal animation data comprising third position information and third motion information of the plurality of joints of the digital human in a third frame of the digital human animation, the third skeletal animation data being determined according to third motion data of a real person collected by the motion capture device, wherein the third frame is before the first frame; The step of determining the detection area in the second frame of the digital human animation according to the first skeletal animation data comprises: The detection area in the second frame of the digital human animation is determined according to the first skeletal animation data and the third skeletal animation data.

6. The method according to any one of claims 1 to 5, characterized in that The motion capture device includes at least one of an optical sensor, an inertial sensor or a visual sensor.

7. A mold penetration detection device, characterized in that: The device comprises: An acquisition module, used for acquiring first skeleton animation data of a digital human, wherein the first skeleton animation data comprises first position information and first motion information of a plurality of joints of the digital human in a first frame of the digital human animation, and the first skeleton animation data is determined according to first motion data of a real person collected by a motion capture device; A processing module, configured to determine, based on the first skeletal animation data, a detection area in a second frame of the digital human animation, wherein the model of the digital human in the detection area is expected to be penetrated, and the second frame is after the first frame; The acquisition module is further used to acquire second skeletal animation data of the digital human, wherein the second skeletal animation data includes second position information and second motion information of the plurality of joints of the digital human in a second frame of the digital human animation, and the second skeletal animation data is determined according to second motion data of the real person collected by the motion capture device; A detection module, configured to perform a penetration detection on the model of the digital human according to the second skeletal animation data within a detection area in the second frame; The feedback module is used to send feedback information when it is detected that the model of the digital human is penetrated, wherein the feedback information includes the position information of the penetrated area of ​​the model of the digital human in the second frame, and the feedback information is used to instruct the real person to adjust the action.

8. The device according to claim 7, characterized in that The processing module is specifically used for: Determining predicted positions of the multiple joints in a second frame of the digital human animation according to the first position information and the first motion information of the multiple joints of the digital human; According to the predicted positions of the multiple joints in the second frame of the digital human animation, a detection area in the second frame of the digital human animation is determined.

9. The device according to claim 8, characterized in that The plurality of joints include a first joint and a second joint, wherein the processing module is specifically used for: Determining a predicted distance between the first joint and the second joint according to predicted positions of the plurality of joints in a second frame of the digital human animation; When the predicted distance is less than or equal to a preset threshold, it is determined that the detection area in the second frame of the digital human animation includes an area between the predicted position of the first joint and the predicted position of the second joint.

10. The device according to any one of claims 7 to 9, characterized in that The model of the digital human includes a plurality of patches on the surface of the digital human, wherein the detection module is specifically used for: According to the second skeletal animation data, an intersection operation is performed on a target facet among the multiple facets on the surface of the digital human to perform mold penetration detection, wherein the target facet is a facet within the detection area of ​​the second frame.

11. The device according to any one of claims 7 to 10, characterized in that The acquisition module is further used to acquire third skeletal animation data of the digital human, wherein the third skeletal animation data includes third position information and third motion information of the plurality of joints of the digital human in a third frame of the digital human animation, and the third skeletal animation data is determined according to third motion data of a real person collected by the motion capture device, wherein the third frame is before the first frame; wherein, The processing module is specifically used for: The detection area in the second frame of the digital human animation is determined according to the first skeletal animation data and the third skeletal animation data.

12. The device according to any one of claims 7 to 11, characterized in that The motion capture device includes at least one of an optical sensor, an inertial sensor or a visual sensor.

13. A computing device cluster, characterized in that: comprising at least one computing device, each computing device comprising a processor and a memory; The processor of the at least one computing device is configured to execute instructions stored in the memory of the at least one computing device, so that the computing device cluster executes the method according to any one of claims 1 to 6.

14. A computer program product comprising a computer program or instructions, characterized in that When the computer program or instruction is executed by a computing device cluster, the computing device cluster is caused to perform the method according to any one of claims 1 to 6.

15. A computer-readable storage medium, characterized in that: The method comprises computer program instructions. When the computer program instructions are executed by a computing device cluster, the computing device cluster performs the method according to any one of claims 1 to 6.

Images