Method and system for realizing virtual reality interaction between human and scene

Abstract

The invention belongs to the technical field of archive arrangement, and provides a method and system for realizing virtual reality interaction between a person and a scene, and the method comprises the steps: building a three-dimensional space coordinate system, splitting the virtual reality interaction of the person into a plurality of actions according to the changes of joint points corresponding to all body parts and the action types defined in an action library, the method comprises the following steps of: dividing a virtual action into split actions, quantitatively calculating differences between the split actions and corresponding virtual actions, dividing the actions into high-difference actions and low-difference actions, performing stability analysis on the difference of each action type to judge whether the reason of high difference is unreasonable virtual action parameter setting or not, and formulating a difference strategy for the different actions and the low-difference actions. And the virtual action parameters are adjusted, so that the problem that the difference between the real action and the virtual action is large when a person performs virtual reality interaction due to unreasonable setting of the virtual action parameters is solved, and the experience feeling of a user is improved.

Description

Technical Field

[0001] The present invention belongs to the field of virtual reality interaction technology, and specifically relates to a method and system for realizing virtual reality interaction between a person and a scene. Background Art

[0002] Existing technologies for determining and optimizing the discrepancies between real and virtual actions during virtual reality interaction between humans and scenes present numerous problems. For example, the lack of multi-dimensional quantification of the differences between real and virtual actions can lead to a one-sided measurement of actions, resulting in inaccurate discrepancy calculations. Actions are not grouped according to their differences, making it impossible to determine whether the discrepancy stems from an unreasonable virtual standard or from the user's own inability to complete the action properly. Regarding virtual action optimization, there is a lack of effective methods for determining whether the significant discrepancy between real and virtual actions stems from an unreasonable virtual action standard setting. Consequently, it is impossible to develop differentiation strategies for different discrepancies in actions, resulting in an inability to efficiently and accurately address the significant discrepancy between real and virtual actions, which in turn reduces the user's experience during virtual reality interaction.

[0003] To this end, the present invention provides a method and system for realizing virtual reality interaction between a person and a scene. Summary of the Invention

[0004] In order to make up for the deficiencies of the prior art, at least one technical problem raised in the background technology is solved.

[0005] The technical solution adopted by the present invention to solve its technical problem is:

[0006] A method for realizing virtual reality interaction between a person and a scene, comprising:

[0007] Step 1: Divide the process of human virtual reality interaction into several actions and generate an action set;

[0008] Step 2: Calculate the comprehensive difference coefficient between each action in the action set and the corresponding virtual action, and divide the action set into a high-difference action set and a low-difference action set according to the comprehensive difference coefficient;

[0009] Step 3: Classify all actions in the action set by action type, calculate the stability coefficient of each action type, and determine whether the high difference in each type of action is due to unreasonable virtual action standards based on the stability coefficient;

[0010] Step 4: If the high action variance is due to unreasonable virtual action standards, adjust the virtual actions for the high variance action set and the low variance action set respectively.

[0011] Furthermore, the process of dividing a person's virtual reality interaction into several actions and generating an action set is as follows:

[0012] The state where the human legs stand vertically and the arms are spread out horizontally is regarded as the original state, and a three-dimensional space coordinate system is constructed with the center point of the two feet on the ground as the origin;

[0013] For each joint, the interaction is divided into multiple actions according to the predefined action types in the action library;

[0014] Calculate the single-axis amplitude and total displacement distance between the real-time coordinates and the original state coordinates, record the maximum displacement coordinates reached by the joints during the action, and generate an action set by dividing the actions and the coordinates of the corresponding joints.

[0015] Furthermore, the process of dividing the action set into a high-difference action set and a low-difference action set according to the comprehensive difference coefficient is as follows:

[0016] The comprehensive difference coefficient is obtained by comprehensively calculating the Euclidean distance and the angle between the position vector of each action in the action set and the corresponding virtual action;

[0017] Compare the comprehensive difference coefficient between each action in the action set and the corresponding virtual action with the threshold;

[0018] If the comprehensive difference coefficient is greater than the threshold, the action is classified as a high-difference action;

[0019] If the comprehensive difference coefficient is less than the threshold, the action is classified as a low-difference action.

[0020] Furthermore, the process of determining whether the high difference in each type of action is due to unreasonable virtual action standards based on the stability coefficient is as follows:

[0021] Classify all actions in the action set by type to generate an action type set, calculate the proportion of high-difference actions and the average difference coefficient of each action type, and multiply the proportion of high-difference actions by the average difference coefficient to obtain the stability coefficient;

[0022] The stability coefficient of each type of action is compared with the threshold. If the stability coefficient is greater than the threshold, the large difference between this type of action and the virtual action is due to the unreasonable virtual action standard.

[0023] Furthermore, the calculation process of the high-difference action ratio is as follows:

[0024] For each action type, count the total number of actions and the number of high-difference actions, and calculate the ratio of the number of high-difference actions to the total number of actions to obtain the proportion of high-difference actions;

[0025] Furthermore, the calculation process of the average difference coefficient is:

[0026] For each action type, the comprehensive difference coefficients of all actions are added together and then divided by the number of actions to obtain the average difference coefficient.

[0027] Furthermore, the process of adjusting the virtual actions corresponding to the high-difference action set is as follows:

[0028] For high-difference action sets, the corresponding actions in the virtual library are deleted, and the action with the smallest comprehensive difference coefficient with the corresponding virtual action in each action type in the high-difference set is updated to the virtual library as a new virtual action.

[0029] Furthermore, the process of adjusting the virtual actions corresponding to the low-discrepancy action set is as follows:

[0030] For a low-variance action set, based on any action type, the average value of all action coordinates is calculated to obtain the aggregate coordinates. The adjustment amplitude is calculated according to the comprehensive difference coefficient, and the coordinate components are interpolated to obtain the adjusted virtual action coordinates.

[0031] Furthermore, the process of calculating the adjustment range based on the comprehensive difference coefficient is:

[0032] For any action type in the low-discrepancy action set, the adjustment amplitude is Among them, E max is the maximum comprehensive difference coefficient in this action type, E min is the minimum comprehensive difference coefficient in this action type, E avg is the comprehensive difference coefficient between the aggregate coordinates and the corresponding virtual action coordinates.

[0033] A virtual reality interaction implementation system for people and scenes, including the following modules:

[0034] Action splitting module: splits the process of human virtual reality interaction into several actions and generates an action set;

[0035] Difference classification module: Calculates the comprehensive difference coefficient between each action in the action set and the corresponding virtual action, and divides the action set into a high-difference action set and a low-difference action set based on the comprehensive difference coefficient;

[0036] Analysis and judgment module: classifies all actions in the action set by action type, calculates the stability coefficient of each action type, and determines whether the high difference in each type of action is due to unreasonable virtual action standards based on the stability coefficient;

[0037] Virtual action optimization module: If the high action difference is due to unreasonable virtual action standards, the corresponding virtual actions are adjusted for the high-difference action set and the low-difference action set respectively.

[0038] The beneficial effects of the present invention are as follows:

[0039] By constructing a three-dimensional spatial coordinate system, people's virtual reality interaction is divided into several actions according to the action types in the action library. The differences between the split actions and the corresponding virtual actions are quantified and classified. The action types are combined to determine whether the high difference is due to unreasonable virtual action standards. Finally, the virtual action parameters are adjusted according to the difference, thus solving the problem of differences between real and virtual actions caused by unreasonable virtual parameter settings and improving the user experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] The present invention will be further described below with reference to the accompanying drawings.

[0041] Figure 1 This is a flowchart of the steps of a method for implementing virtual reality interaction between a person and a scene according to an embodiment of the present invention;

[0042] Figure 2 This is a flowchart of a system for implementing virtual reality interaction between a person and a scene according to an embodiment of the present invention. DETAILED DESCRIPTION

[0043] In order to make the technical means, creative features, objectives and effects achieved by the present invention easier to understand, the present invention is further described below in conjunction with specific implementation methods.

[0044] Example 1

[0045] See also Figure 1 As shown, a method for implementing virtual reality interaction between a person and a scene according to an embodiment of the present invention includes the following steps:

[0046] Step 1: Divide the process of human virtual reality interaction into several actions and generate an action set;

[0047] In step 1, the process of dividing the process of a person interacting with virtual reality into several actions includes:

[0048] The state of a person standing with legs vertically and arms extended horizontally is considered the original state. The center point of the two feet on the ground is taken as the origin, the forward and backward direction of the center point of the two feet on the ground is the X-axis, the horizontal direction is the Y-axis, and the vertical upward direction of the middle line of the human body is the Z-axis to construct a three-dimensional space coordinate system;

[0049] Record the original three-dimensional coordinates (x0, y0, z0) of each joint, and obtain the three-dimensional coordinates (x, y, z) of the human body joints in real time through the VR device sensor;

[0050] For each joint point, calculate the single-axis amplitude D between the real-time coordinate and the original state coordinate X =|x-x0|, D Y =|Y-y0|, D Z =|z-z0| and total displacement distance

[0051] The interaction is divided into multiple actions according to the predefined action types in the action library. When the amplitude of any joint exceeds T0, it is judged to enter a new action cycle. When the body part falls below T0 and the duration exceeds Δt, the action cycle is judged to end;

[0052] In each movement cycle, the displacement distance and single-axis amplitude of the joints are monitored in real time, and the maximum displacement coordinates of the joints during the movement are recorded. If a movement type has multiple repeated movements, each repeated movement is regarded as a separate movement;

[0053] Generate an action set by splitting the actions and the coordinates of the corresponding joint points;

[0054] Step 2: Calculate the comprehensive difference coefficient between each action in the action set and the corresponding virtual action, and divide the action set into a high-difference action set and a low-difference action set according to the comprehensive difference coefficient;

[0055] In step 2, the calculation process of the comprehensive difference coefficient between each action in the action set and the corresponding virtual action includes:

[0056] For each action in the action set, the displacement vector of each action and the corresponding virtual action is and Among them, Δx, Δy, and Δz are the coordinate changes of each action in the action set on the x, y, and z axes respectively. Δx std , Δy std , Δz std are the coordinate changes of the corresponding virtual actions on the x, y, and z axes respectively;

[0057] Calculate the Euclidean distance between each action in the action set and the corresponding virtual action The smaller D is, the smaller the position deviation between the two actions is;

[0058] Calculate the angle between each action in the action set and the corresponding virtual action position vector in, is the vector dot product, and is the vector modulus,

[0059] The comprehensive coefficient of variation is Among them, α is the weighting coefficient, D max is the preset maximum allowable position deviation, θ max is the preset maximum allowable directional deviation;

[0060] It can be understood that the physical meaning of the comprehensive difference coefficient is: the comprehensive difference coefficient is a quantitative indicator used to characterize the comprehensive difference between the sum of the position deviations and the sum of the direction deviations of each body part in three-dimensional space between two actions;

[0061] In step 2, the process of dividing the action set into a high-difference action set and a low-difference action set according to the comprehensive difference coefficient includes:

[0062] Compare the comprehensive difference coefficient of each action in the action set with the corresponding virtual action with the threshold. If the comprehensive difference coefficient is greater than the threshold, the action is classified as a high-difference action. If the comprehensive difference coefficient is less than the threshold, the action is classified as a low-difference action.

[0063] Step 3: Classify all actions in the action set by action type, calculate the stability coefficient of each action type, and determine whether the high difference in each type of action is due to unreasonable virtual action standards based on the stability coefficient;

[0064] In step 3, the calculation process of the stability coefficient includes:

[0065] Classify all actions in the action set by type to generate an action type set T = {t1, t2, ..., t m}, where t m is the mth action type, each type t j Contains action collection is the nth action type in the jth action type j Actions

[0066] For each action type t j , calculate the ratio of the number of high-difference actions to the total number of actions of this type, and get the proportion of high-difference actions Among them, j represents the jth action type;

[0067] Calculate the average comprehensive variance coefficient for each action type Among them, E jk is the comprehensive difference coefficient of the kth action in the jth action type;

[0068] For each action type, the stability coefficient S is obtained by multiplying the proportion of high-difference actions by the average comprehensive difference coefficient. j =R high,j ·E avg,j , where S jis the stability coefficient of the j-th action type;

[0069] It is understandable that the physical meaning of the stability coefficient is: a comprehensive quantitative indicator used to measure the consistency and difficulty of achieving the same type of action in VR interaction. It combines the prevalence of high-variance actions with the severity of the degree of variation, helping to distinguish whether action differences are caused by system standard issues or user execution issues.

[0070] In step three, the process of determining whether the difference in each type of movement is due to the virtual movement standard being too high based on the stability coefficient includes:

[0071] Compare the stability coefficient of each type of action with the threshold. If the stability coefficient is greater than the threshold, the large difference between this type of action and the virtual action is due to the unreasonable virtual action standard.

[0072] For example, suppose a user completes a series of "hand-raising" actions (type t1) and "leg-kicking" actions (type t2) in virtual reality, and the system collects the following data:

[0073] Action type t1: hand raising, total number of actions: 10 times (n1=10), number of high-difference actions: 8 times (comprehensive difference coefficient E>0.5), proportion of high-difference actions: The comprehensive coefficient of variation of each action is: [0.6, 0.7, 0.55, 0.8, 0.65, 0.7, 0.68, 0.58, 0.4*, 0.45*] (the two times marked with * are low-variance actions), the average coefficient of variation is:

[0074] Action type t2: kicking, total number of actions: 10 times (n2=10), number of high-difference actions: 2 times (comprehensive difference coefficient E>0.5), proportion of high-difference actions: Average coefficient of variation: E avg,2 =0.3;

[0075] Stability coefficient calculation: Hand raising action (t1): S1 = R high,1 ×E avg,1 =0.8×0.616=0.4928, kicking action (t2): S2=R high,2 ×E avg,2 =0.2×0.3=0.06;

[0076] Assuming the stability coefficient threshold is 0.4, the stability coefficient of the "hand raising" action is greater than the threshold, and the difference is determined to be due to unreasonable virtual action standards. The stability coefficient of the "leg kicking" action is less than the threshold, and the difference is determined to be due to individual user execution problems.

[0077] Step 4: If the high action variance is due to unreasonable virtual action standards, adjust the virtual actions for the high variance action set and the low variance action set respectively;

[0078] In step 4, the process of adjusting the corresponding virtual action includes:

[0079] For any action type with high action variance due to high virtual action standards, if it exists in both the high variance action set and the low variance action set, delete the action of this type in the low variance action set;

[0080] For high-difference action sets, the corresponding actions in the virtual library are deleted, and the action with the smallest comprehensive difference coefficient with the corresponding virtual action in each action type in the high-difference set is updated to the virtual library as the new virtual action;

[0081] For low-discrepancy action sets, based on any action type t j , contains n actions, the coordinates of each action are R k =(R k,x ,R k,y ,R k,z ), where R k,x is the x-coordinate of the k-th action, R k,y is the y coordinate of the kth action, R k,z is the z coordinate of the kth action, and the comprehensive difference coefficient is E k (k=1,2,...,n), calculate the action type t j The average value of all action coordinates;

[0082] average value Get the aggregate coordinates

[0083] The virtual action coordinates corresponding to the aggregate coordinates are V = (V x ,V y ,V z ), adjust the parameters of the virtual action to make it close to the low-discrepancy set t j The average value of all action coordinates in the action type;

[0084] The linear interpolation method is used to adjust the virtual action coordinates. For any action type in the low-discrepancy action set, the adjustment amplitude λ is calculated according to the comprehensive difference coefficient E. Among them, E max is the maximum comprehensive difference coefficient in this action type, E min is the minimum comprehensive difference coefficient in this action type, E avg is the comprehensive difference coefficient between the aggregate coordinates and the corresponding virtual action coordinates;

[0085] Interpolate the coordinate components of the X, Y, and Z axes to obtain the adjusted virtual action coordinates:

[0086]

[0087] The calculated V ’ Replace the original action in the virtual action library j Parameters to complete the adjustment;

[0088] For example, suppose there are two actions, "squat" and "arm swing", in the virtual action library. The "squat" action is determined to be a virtual standard problem and needs to be updated with a high-difference action. The "arm swing" action needs to be interpolated and adjusted with a low-difference action. The coordinates of the virtual action corresponding to the "squat" are (0.8, 1.2, 0.5), and the coordinates of the virtual action corresponding to the "arm swing" are (0.3, 0.6, 0.9).

[0089] The high variance exercise set of the "squat" type includes 5 exercises, and their comprehensive variance coefficients and coordinates are as follows:

[0090] |Action number k|Comprehensive difference coefficient E k |Maximum amplitude coordinate (R x ,R y ,R z) |;

[0091] |1|0.65|(0.7,1.0,0.6)|;

[0092] |2|0.6|(0.75,1.1,0.55)|;

[0093] |3|0.55|(0.8,1.0,0.65)(minimum difference)|;

[0094] |4|0.7|(0.65,0.9,0.7)|;

[0095] |5|0.68|(0.7,0.95,0.68)|;

[0096] The original "squat" action coordinates in the virtual action library are V = (0.8, 1.2, 0.5), and the action E in sequence number 3 k =0.55 is the smallest, and its coordinates are R3=(0.8,1.0,0.65). Replace the "squat" virtual action with the coordinates of R3, that is, the new virtual action coordinates are V new =(0.85,1.3,1.6);

[0097] The "Arm Swing" movement type includes 4 low-variance movements:

[0098] |Action number k|Comprehensive difference coefficient E k|Maximum amplitude coordinate (R x ,R y ,R z )|;

[0099] |1|0.2|(0.4,0.7,1.0)|;

[0100] |2|0.25|(0.35,0.65,0.95)|;

[0101] |3|0.3|(0.45,0.75,1.05)|;

[0102] |4|0.35|(0.5,0.8,1.1)(largest difference)|;

[0103] Calculate the aggregate coordinates:

[0104] Assume that the comprehensive difference coefficient between the aggregated coordinates and the corresponding virtual coordinates is E avg =0.3, calculate the adjustment range

[0105] Linear interpolation adjusts virtual coordinates: V x ′=(1-0.67)×0.3+0.67×0.425≈0.34, V y ′=(1-0.67)×0.6+0.67×0.725≈0.64, V z ′=(1-0.67)×0.9+0.67×1.025≈0.94, and the new virtual coordinates are V′=(0.34,0.64,0.94).

[0106] The technical solution and benefits of the embodiments of the present application lie in breaking down the process of human virtual reality interaction into several actions, generating an action set, calculating the comprehensive difference coefficient between each action in the action set and the corresponding virtual action, dividing the action set into a high-difference action set and a low-difference action set based on the comprehensive difference coefficient, classifying all actions in the action set by action type, calculating the stability coefficient for each action type, and determining whether the high difference in each action type is due to excessively high virtual action standards based on the stability coefficient. For the high-difference action set and the low-difference action set, the corresponding virtual actions are adjusted. This application constructs a three-dimensional spatial coordinate system, breaks down human virtual reality interaction into several actions based on the changes in the corresponding joints of each body part and the action types defined in an action library, and quantifies the difference between the broken actions and the corresponding virtual actions. Based on the difference and action type, it is determined whether the high difference is due to improper virtual action parameter settings. A differentiation strategy is then developed to adjust the virtual action parameters based on the difference, thereby resolving the problem of large differences between real and virtual actions during virtual reality interaction caused by improper virtual action parameter settings, thereby improving the user experience.

[0107] Example 2

[0108] See also Figure 2 As shown, a system for realizing virtual reality interaction between a person and a scene according to an embodiment of the present invention includes:

[0109] Action splitting module: splits the process of human virtual reality interaction into several actions and generates an action set;

[0110] The process of dividing the process of a person interacting with virtual reality into several actions includes:

[0111] The state of a person standing with legs vertically and arms extended horizontally is considered the original state. The center point of the two feet on the ground is taken as the origin, the forward and backward direction of the center point of the two feet on the ground is the X-axis, the horizontal direction is the Y-axis, and the vertical upward direction of the middle line of the human body is the Z-axis to construct a three-dimensional space coordinate system;

[0112] Record the original three-dimensional coordinates (x0, y0, z0) of each joint, and obtain the three-dimensional coordinates (x, y, z) of the human body joints in real time through the VR device sensor;

[0113] For each joint point, calculate the single-axis amplitude D between the real-time coordinate and the original state coordinate X =|x-x0|, D Y =|Y-y0|, D Z =|z-z0| and total displacement distance

[0114] The interaction is divided into multiple actions according to the predefined action types in the action library. When the amplitude of any joint exceeds T0, it is judged to enter a new action cycle. When the body part falls below T0 and the duration exceeds Δt, the action cycle is judged to end;

[0115] In each movement cycle, the displacement distance and single-axis amplitude of the joints are monitored in real time, and the maximum displacement coordinates of the joints during the movement are recorded. If a movement type has multiple repeated movements, each repeated movement is regarded as a separate movement;

[0116] Generate an action set by splitting the actions and the coordinates of the corresponding joint points;

[0117] Difference judgment module: Calculates the comprehensive difference coefficient between each action in the action set and the corresponding virtual action, and divides the action set into a high-difference action set and a low-difference action set based on the comprehensive difference coefficient;

[0118] The calculation process of the comprehensive difference coefficient between each action in the action set and the corresponding virtual action includes:

[0119] For each action in the action set, the displacement vector of each action and the corresponding virtual action is and Among them, Δx, Δy, and Δz are the coordinate changes of each action in the action set on the x, y, and z axes respectively. Δx std , Δy std , Δz std are the coordinate changes of the corresponding virtual actions on the x, y, and z axes respectively;

[0120] Calculate the Euclidean distance between each action in the action set and the corresponding virtual action The smaller D is, the smaller the position deviation between the two actions is;

[0121] Calculate the angle between each action in the action set and the corresponding virtual action position vector in, is the vector dot product, and is the vector modulus,

[0122] The comprehensive coefficient of variation is Among them, α is the weighting coefficient, D max is the preset maximum allowable position deviation, θ max is the preset maximum allowable directional deviation;

[0123] It can be understood that the physical meaning of the comprehensive difference coefficient is: the comprehensive difference coefficient is a quantitative indicator used to characterize the comprehensive difference between the sum of the position deviations and the sum of the direction deviations of each body part in three-dimensional space between two actions;

[0124] The process of dividing the action set into a high-difference action set and a low-difference action set according to the comprehensive difference coefficient includes:

[0125] Compare the comprehensive difference coefficient of each action in the action set with the corresponding virtual action with the threshold. If the comprehensive difference coefficient is greater than the threshold, the action is classified as a high-difference action. If the comprehensive difference coefficient is less than the threshold, the action is classified as a low-difference action.

[0126] Analysis and judgment module: classifies all actions in the action set by action type, calculates the stability coefficient of each action type, and determines whether the high difference in each type of action is due to the excessively high virtual action standard based on the stability coefficient;

[0127] The calculation process of the stability coefficient includes:

[0128] Classify all actions in the action set by type to generate an action type set T = {t1, t2, ..., t m}, where t m is the mth action type, each type t j Contains action collection is the nth action type in the jth action type j Actions

[0129] For each action type t j , calculate the ratio of the number of high-difference actions to the total number of actions of this type, and get the proportion of high-difference actions Among them, j represents the jth action type;

[0130] Calculate the average comprehensive variance coefficient for each action type Among them, E jk is the comprehensive difference coefficient of the kth action in the jth action type;

[0131] For each action type, the stability coefficient S is obtained by multiplying the proportion of high-difference actions by the average comprehensive difference coefficient. j =R high,j ·E avg,j , where S j is the stability coefficient of the j-th action type;

[0132] It is understandable that the physical meaning of the stability coefficient is: a comprehensive quantitative indicator used to measure the consistency and difficulty of achieving the same type of action in VR interaction. It combines the prevalence of high-variance actions with the severity of the degree of variation, helping to distinguish whether action differences are caused by system standard issues or user execution issues.

[0133] The process of judging whether the difference in each type of movement is due to excessively high virtual movement standards based on the stability coefficient includes:

[0134] Compare the stability coefficient of each type of action with the threshold. If the stability coefficient is greater than the threshold, the large difference between this type of action and the virtual action is due to the unreasonable virtual action standard.

[0135] Virtual action optimization module: If the high action variance is due to unreasonable virtual action standards, the virtual actions are adjusted for the high variance action set and the low variance action set respectively;

[0136] The process of adjusting the corresponding virtual action includes:

[0137] For any action type with high action variance due to high virtual action standards, if it exists in both the high variance action set and the low variance action set, delete the action of this type in the low variance action set;

[0138] For high-difference action sets, the corresponding actions in the virtual library are deleted, and the action with the smallest comprehensive difference coefficient with the corresponding virtual action in each action type in the high-difference set is updated to the virtual library as the new virtual action;

[0139] For low-discrepancy action sets, based on any action type t j , contains n actions, the coordinates of each action are R k =(R k,x ,R k,y ,R k,z ), where R k,x is the x-coordinate of the k-th action, R k,y is the y coordinate of the kth action, R k,z is the z coordinate of the kth action, and the comprehensive difference coefficient is E k (k=1,2,...,n), calculate the action type t j The average value of all action coordinates;

[0140] average value Get the aggregate coordinates

[0141] The virtual action coordinates corresponding to the aggregate coordinates are V = (V x ,V y ,Vz ), adjust the parameters of the virtual action to make it close to the low-discrepancy set t j The average value of all action coordinates in the action type;

[0142] The linear interpolation method is used to adjust the virtual action coordinates. For any action type in the low-discrepancy action set, the adjustment amplitude λ is calculated according to the comprehensive difference coefficient E. Among them, E max is the maximum comprehensive difference coefficient in this action type, E min is the minimum comprehensive difference coefficient in this action type, E avg is the comprehensive difference coefficient between the aggregate coordinates and the corresponding virtual action coordinates;

[0143] Interpolate the coordinate components of the X, Y, and Z axes to obtain the adjusted virtual action coordinates:

[0144]

[0145] The calculated V ’ Replace the original action in the virtual action library j Parameters to complete the adjustment.

[0146] The basic principles, main features, and advantages of the present invention are shown and described above. Those skilled in the art should understand that the present invention is not limited to the foregoing embodiments. The foregoing embodiments and descriptions are merely illustrative of the principles of the present invention. Various changes and modifications may be made to the present invention without departing from the spirit and scope of the present invention. Such changes and modifications are intended to fall within the scope of the present invention. The scope of protection claimed in the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for implementing virtual reality interaction between a person and a scene, characterized by: include: Step 1: Divide the process of human virtual reality interaction into several actions and generate an action set; Step 2: Calculate the comprehensive difference coefficient between each action in the action set and the corresponding virtual action, and divide the action set into a high-difference action set and a low-difference action set according to the comprehensive difference coefficient; Step 3: Classify all actions in the action set by action type, calculate the stability coefficient of each action type, and determine whether the high difference in each type of action is due to unreasonable virtual action standards based on the stability coefficient; Step 4: If the high action variance is due to unreasonable virtual action standards, adjust the virtual actions for the high variance action set and the low variance action set respectively.

2. The method for realizing virtual reality interaction between a person and a scene according to claim 1, characterized in that: The method of dividing the process of human virtual reality interaction into several actions and generating an action set is as follows: The state where the human legs stand vertically and the arms are spread out horizontally is regarded as the original state, and a three-dimensional space coordinate system is constructed with the center point of the two feet on the ground as the origin; For each joint, the interaction is divided into multiple actions according to the predefined action types in the action library; Calculate the single-axis amplitude and total displacement distance between the real-time coordinates and the original state coordinates, record the maximum displacement coordinates reached by the joints during the action, and generate an action set by dividing the actions and the coordinates of the corresponding joints.

3. The method for realizing virtual reality interaction between a person and a scene according to claim 1, characterized in that: The process of dividing the action set into a high-difference action set and a low-difference action set according to the comprehensive difference coefficient is as follows: The comprehensive difference coefficient is obtained by comprehensively calculating the Euclidean distance and the angle between the position vector of each action in the action set and the corresponding virtual action; Compare the comprehensive difference coefficient between each action in the action set and the corresponding virtual action with the threshold; If the comprehensive difference coefficient is greater than the threshold, the action is classified as a high-difference action; If the comprehensive difference coefficient is less than the threshold, the action is classified as a low-difference action.

4. The method for realizing virtual reality interaction between a person and a scene according to claim 1, characterized in that: The process of judging whether the high difference of each type of action is due to unreasonable virtual action standards based on the stability coefficient is as follows: Classify all actions in the action set by type to generate an action type set, calculate the proportion of high-difference actions and the average difference coefficient of each action type, and multiply the proportion of high-difference actions by the average difference coefficient to obtain the stability coefficient; The stability coefficient of each type of action is compared with the threshold. If the stability coefficient is greater than the threshold, the large difference between this type of action and the virtual action is due to the unreasonable virtual action standard.

5. The method for realizing virtual reality interaction between a person and a scene according to claim 4, characterized in that: The calculation process of the high-difference action ratio is as follows: For each action type, the total number of actions and the number of high-difference actions are counted, and the proportion of high-difference actions is calculated by dividing the number of high-difference actions by the total number of actions.

6. The method for realizing virtual reality interaction between a person and a scene according to claim 4, characterized in that: The calculation process of the average difference coefficient is: For each action type, the comprehensive difference coefficients of all actions are added together and then divided by the number of actions to obtain the average difference coefficient.

7. The method for realizing virtual reality interaction between a person and a scene according to claim 1, characterized in that: The process of adjusting the virtual actions corresponding to the high-difference action set is as follows: For high-difference action sets, the corresponding actions in the virtual library are deleted, and the action with the smallest comprehensive difference coefficient with the corresponding virtual action in each action type in the high-difference set is updated to the virtual library as a new virtual action.

8. The method for realizing virtual reality interaction between a person and a scene according to claim 1, characterized in that: The process of adjusting the virtual actions corresponding to the low-discrepancy action set is as follows: For a low-variance action set, based on any action type, the average value of all action coordinates is calculated to obtain the aggregate coordinates. The adjustment amplitude is calculated according to the comprehensive difference coefficient, and the coordinate components are interpolated to obtain the adjusted virtual action coordinates.

9. The method for realizing virtual reality interaction between a person and a scene according to claim 7, characterized in that: The process of calculating the adjustment range based on the comprehensive difference coefficient is as follows: For any action type in the low-discrepancy action set, the adjustment amplitude is Among them, E max is the maximum comprehensive difference coefficient in this action type, E min is the minimum comprehensive difference coefficient in this action type, E avg is the comprehensive difference coefficient between the aggregate coordinates and the corresponding virtual action coordinates.

10. A virtual reality interaction system between a person and a scene, characterized by: Includes the following modules: Action splitting module: splits the process of human virtual reality interaction into several actions and generates an action set; Difference classification module: Calculates the comprehensive difference coefficient between each action in the action set and the corresponding virtual action, and divides the action set into a high-difference action set and a low-difference action set based on the comprehensive difference coefficient; Analysis and judgment module: classifies all actions in the action set by action type, calculates the stability coefficient of each action type, and determines whether the high difference in each type of action is due to unreasonable virtual action standards based on the stability coefficient; Virtual action optimization module: If the high action difference is due to unreasonable virtual action standards, the corresponding virtual actions are adjusted for the high-difference action set and the low-difference action set respectively.

Images