A 3D digital human generation method based on AIGC
By comparing the 3D digital human joint trajectory curve and adjusting the rendering magnification and data volume, the problems of low operation accuracy and generation efficiency of 3D digital humans in the prior art are solved, and more efficient and accurate digital human generation is achieved.
Patent Information
- Application Number
- CN202410347327.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-03-26
- Publication Date
- 2025-04-04
- Estimated Expiration
- 2044-03-26
AI Technical Summary
The prior art fails to consider adjusting the generation parameters based on the operation of each joint of the 3D digital human, which affects the accuracy and generation efficiency of the 3D digital human operation.
Through data collection and preprocessing, key feature points are extracted to generate the initial model of 3D digital humans, and the joint trajectory curve is compared according to preset standards, and the rendering magnification and data volume are adjusted to improve the accuracy and generation efficiency of joint operation.
It improves the accuracy and generation efficiency of 3D digital human operation, ensures the smoothness and accuracy of joint movement, and reduces trajectory deviations caused by lag.
Smart Images

Figure CN118521685B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of virtual digital human, and in particular to a 3D digital human generation method based on AIGC. Background Art
[0002] A virtual human model is a computer-generated virtual entity used to simulate and express human images and movements. In the fields of virtual reality, game development, film and television special effects, virtual human models are widely used in human-computer interaction, role-playing, scene reconstruction, etc. Traditional methods for generating virtual human models usually use manual design or physical simulation-based methods, but these methods are difficult to consider the details of human body shape and movement performance during the generation process.
[0003] Chinese Patent Publication No.: CN116630495A discloses a virtual digital human model planning system based on AIGC algorithm, including a data acquisition unit, a virtual digital human generation unit, an editing and storage unit and a data optimization unit. The data acquisition unit is used to collect human image data, the virtual digital human generation unit is used to collect data acquisition unit images, and generate a virtual digital human by analyzing the image data through the AIGC algorithm. The editing and storage unit is used to edit the appearance and clothing of the virtual digital human generated by the virtual digital human generation unit and classify and store them, and the data optimization unit is used to run the virtual digital human edited by the editing and storage unit. It can be seen that the prior art has the following problems: it does not take into account the adjustment of the generation parameters of the 3D digital human according to the operation status of each joint of the 3D digital human, which affects the accuracy of the 3D digital human operation and further affects the generation efficiency of the 3D digital human. Summary of the invention
[0004] To this end, the present invention provides a 3D digital human generation method based on AIGC to overcome the problem in the prior art that the generation parameters of the 3D digital human are not adjusted according to the operation conditions of each joint of the 3D digital human, which affects the accuracy of the 3D digital human operation and further affects the generation efficiency of the 3D digital human.
[0005] To achieve the above object, the present invention provides a 3D digital human generation method based on AIGC, comprising:
[0006] Data collection and preprocessing of the collected data, including data cleaning, reconstruction, and alignment; feature extraction of the preprocessed data to generate the initial model of the 3D digital human based on the extracted key feature points;
[0007] Animation generation: adding animation to the 3D digital human model to make it move according to pre-designed actions; rendering settings: setting the preset rendering magnification in the rendering engine to control the clarity of the image;
[0008] Data transmission: the server downloads the 3D digital human model data, sets a cache area with a preset data volume to cache the data locally, parses and processes the transmitted data, caches the calculation results during the calculation of skeletal animation, generates a 3D digital human and presents it on the screen;
[0009] Data adjustment: establishing coordinates for the generated 3D digital human, and comparing the trajectory curve of a single joint of the 3D digital human in the coordinates with the corresponding preset trajectory curve, and determining whether the operation of the single joint meets the preset standard according to the obtained overlap, and when it is preliminarily determined that the operation of the single joint does not meet the preset standard, determining whether to adjust the preset rendering magnification according to the distance between the end point of the obtained trajectory curve of the single joint and the end point of the preset trajectory curve;
[0010] Data generation: When the operation of each joint of the 3D digital human meets the preset standards, an output is generated to indicate that the generation of the 3D digital human is complete.
[0011] Further, the analysis module establishes coordinates for the generated 3D digital human, and the analysis module compares the trajectory curve of a single joint of the 3D digital human in the coordinates with the corresponding preset trajectory curve, and determines whether the operation of the single joint meets the preset standard according to the obtained overlap, and when it is determined that the operation of the single joint does not meet the preset standard, determines the processing method for the single joint according to the difference between the first preset overlap and the overlap, or when it is preliminarily determined that the operation of the single joint does not meet the preset standard, re-determines whether the operation of the single joint meets the preset standard according to the distance between the end point of the acquired trajectory curve of the single joint and the end point of the preset trajectory curve;
[0012] The overlap degree is the ratio of the length of the overlapped trajectory curve to the length of the preset trajectory curve.
[0013] Furthermore, the analysis module calculates the ratio of the distance between the end point of the trajectory curve of a single joint and the end point of the preset trajectory curve to the length of the preset trajectory curve, and records it as the distance ratio. The analysis module determines whether the operation of the single joint meets the preset standard based on the obtained distance ratio, and when it is determined that the operation of the single joint does not meet the preset standard, the preset rendering magnification is adjusted to a corresponding value based on the difference between the distance ratio and the preset distance ratio.
[0014] Furthermore, the analysis module records the difference between the first preset overlap and the overlap as the overlap difference, and determines the operation processing method for the single joint based on the obtained overlap difference, including adjusting the preset rendering magnification to a corresponding value according to the number of areas with overlap abnormalities, and determining whether to adjust the relative angle between the single joint and the superior joint to a corresponding value according to the judgment result of the superior joint of the single joint, and adjusting the distance between the two joints to a corresponding value according to the ratio of the length of the trajectory curve to the length of the preset trajectory curve. The analysis module redetermines the coordinates of the single joint based on the adjusted distance between the two joints, and re-determines whether the operation of the single joint meets the preset standard based on the overlap according to the redetermined coordinates of the single joint.
[0015] Furthermore, the analysis module obtains the overlapping graph of the trajectory curve and the corresponding preset trajectory curve, and marks each area with overlapping abnormalities as an abnormal area. The analysis module is provided with a plurality of rendering adjustment methods for the preset rendering magnification based on the number of abnormal areas obtained, and the adjustment range of each rendering adjustment method for the preset rendering magnification is different.
[0016] Further, when the analysis module completes the adjustment for the rendering magnification, it re-determines whether the operation of the single joint meets the preset standard based on the re-acquired single joint. If the analysis module still determines to adjust the rendering magnification, the analysis module determines to use the current rendering magnification as the generation parameter for the 3D digital human, and adjusts the preset data volume to a corresponding value according to the difference between the second preset overlap and the re-acquired overlap;
[0017] The analysis module is provided with a plurality of data adjustment methods for the preset data volume based on the difference between the second preset overlap degree and the re-acquired overlap degree, and the adjustment range of each data adjustment method for the preset data volume is different.
[0018] Further, the analysis module obtains the determination result of the superior joint of the single joint; when it is determined that the operation of the superior joint meets the preset standard, the relative angle between the two joints is adjusted to a corresponding value according to the second pre-overlap difference and the lower level difference of the overlap difference; when it is determined that the operation of the superior joint does not meet the preset standard, the superior joint is processed accordingly, and the single joint is re-detected after the processing is completed, and the analysis mechanism determines whether the operation of the single joint meets the preset standard according to the re-acquired overlap;
[0019] Among them, the relative angle between the two joints is the angle between the two line segments whose intersection points are the upper joints obtained by connecting the joints.
[0020] Furthermore, the analysis module is provided with a plurality of angle adjustment methods for the relative angle based on a lower level difference between the second pre-overlap difference and the overlap difference, and each angle adjustment method has a different adjustment range for the relative angle between the two joints.
[0021] Furthermore, the analysis module records the ratio of the length of the trajectory curve to the length of the preset trajectory curve as the length ratio. Based on the obtained length ratio, the analysis module provides several adjustment methods for the distance between the two joints, and the adjustment range of each distance adjustment method for the distance between the two joints is different.
[0022] Furthermore, the analysis module records the difference between the distance ratio and the preset distance ratio as the distance difference, and the analysis module provides several model adjustment methods for the preset rendering magnification based on the obtained distance difference, and the adjustment range of each model adjustment method for the preset rendering magnification is different.
[0023] Compared with the prior art, the actual operation status of each joint of the 3D digital human is determined, and whether the operation of a single joint meets the preset standard is determined for the trajectory curve of a single joint. When the overlap is less than or equal to the second preset overlap and greater than the first preset overlap, the distance between the end point of the trajectory curve of the single joint and the end point of the preset trajectory curve is obtained to determine whether the joint fails to run to the corresponding point within the time period of obtaining the trajectory curve due to jamming. When the distance ratio is less than or equal to the preset distance ratio, the single joint completes the operation of the preset trajectory; when the distance ratio is greater than the preset distance ratio, the analysis module determines to lower the preset rendering magnification, so as to reduce the calculation burden, improve the operation fluency of the digital human, and further effectively improve the generation efficiency of the 3D digital human.
[0024] Furthermore, when the overlap degree is less than or equal to the first preset overlap degree, the corresponding processing method is specifically determined according to the overlap difference. When the overlap difference is less than or equal to the first preset overlap difference, the running trajectory only has a small range of deviations. In this case, the 3D digital human is stuck, and the preset rendering magnification is lowered; when the overlap difference is less than or equal to the second preset overlap difference and greater than the first preset overlap difference, the running trajectory of a single joint has a large range of deviations. In this case, the parent joint of the single joint is specifically determined, so that when the operation of the parent joint does not meet the preset standard, the parent joint can be optimized in time to avoid abnormalities caused by the parent joint. Affects the accuracy of the judgment of the operation of a single joint; when the operation of the upper joint meets the preset standard, in this case, the relative angle between the single joint and the upper joint is abnormal, resulting in a deviation between the trajectory curve and the preset trajectory curve, so the relative angle between the single joint and the upper joint is adjusted; when the overlap difference is greater than the second preset overlap difference, the trajectory curve deviation is too large, in this case, the joint selection is abnormal, the joint is re-determined, and whether the operation of the joint meets the preset standard is judged again based on the re-determined joint; while effectively improving the accuracy of the 3D digital human operation, the generation efficiency of the 3D digital human is further effectively improved. BRIEF DESCRIPTION OF THE DRAWINGS
[0025] Figure 1 This is a flowchart of the steps of the method for generating a 3D digital human based on AIGC according to an embodiment of the present invention;
[0026] Figure 2 This is a flow chart of a joint determination method in which the analysis module of an embodiment of the present invention determines whether the operation of a single joint meets a preset standard based on the degree of overlap. DETAILED DESCRIPTION
[0027] In order to make the objects and advantages of the present invention more clearly understood, the present invention is further described below in conjunction with embodiments; it should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.
[0028] The preferred embodiments of the present invention are described below with reference to the accompanying drawings. It should be understood by those skilled in the art that these embodiments are only used to explain the technical principles of the present invention and are not intended to limit the protection scope of the present invention.
[0029] It should be noted that, in the description of the present invention, terms such as "up", "down", "left", "right", "inside" and "outside" indicating directions or positional relationships are based on the directions or positional relationships shown in the drawings. This is merely for the convenience of description and does not indicate or imply that the device or element must have a specific orientation, be constructed and operated in a specific orientation. Therefore, it cannot be understood as a limitation on the present invention.
[0030] In addition, it should be noted that in the description of the present invention, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For those skilled in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.
[0031] See also Figure 1 as well as Figure 2 As shown, they are respectively a flowchart of the steps of the 3D digital human generation method based on AIGC in an embodiment of the present invention, and a flowchart of the joint determination method in which the analysis module determines whether the operation of a single joint meets the preset standard based on the overlap; a 3D digital human generation method based on AIGC in an embodiment of the present invention comprises:
[0032] S1, data collection, collects data used to generate 3D digital humans, including human body models, motion capture data, and facial expression data;
[0033] S2, data preprocessing, preprocessing the collected data, including data cleaning, reconstruction, and alignment;
[0034] S3, feature extraction, performs feature extraction on the preprocessed data to extract key feature points and motion trajectories of the human body model;
[0035] S4, generating a model, generating an initial model of a 3D digital human based on the extracted key feature points;
[0036] S5, texture mapping, adding texture mapping to the generated 3D digital human model to increase the realism and details of the model;
[0037] S6, animation generation, adding animation to the 3D digital human model so that it moves according to pre-designed actions; rendering settings, setting the preset rendering magnification in the rendering engine to control the clarity of the image;
[0038] S7, model presentation, presenting the generated 3D digital human model in a virtual environment;
[0039] S8, data transmission, the server downloads the 3D digital human model data, sets a cache area with a preset data volume to cache the data locally, parses and processes the transmitted data, caches the calculation results during the calculation of skeletal animation, generates a 3D digital human and presents it on the screen;
[0040] S9, data adjustment, establishing coordinates for the generated 3D digital human, aligning and comparing the trajectory curve of a single joint of the 3D digital human in the coordinates with the corresponding preset trajectory curve, determining whether the operation of the single joint meets the preset standard according to the obtained overlap, and when it is preliminarily determined that the operation of the single joint does not meet the preset standard, determining whether to adjust the preset rendering magnification according to the distance between the end point of the acquired trajectory curve of the single joint and the end point of the preset trajectory curve;
[0041] S10, data generation, when the operation of each joint of the 3D digital human meets the preset standard, an identification indicating that the generation of the 3D digital human is completed is output.
[0042] In order to make the objects and advantages of the present invention more clearly understood, the present invention is further described below in conjunction with embodiments; it should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.
[0043] It should be pointed out that the data in this embodiment are obtained by comprehensive analysis and evaluation of the historical test data and the corresponding historical test results of the three months before this test by the method of the present invention. Before this test, the method of the present invention comprehensively determines the values of the preset parameter standards for this test based on the 1753 3D digital humans and the joint trajectory curves corresponding to the 3D digital humans, the rendering data of the 3D digital humans, and the cache area. It can be understood by those skilled in the art that the method of the present invention can determine the single parameter of the above parameters by selecting the highest value as the preset standard parameter according to the data distribution, using weighted summation to use the obtained value as the preset standard parameter, substituting each historical data into a specific formula and using the value obtained by the formula as the preset standard parameter or other selection methods, as long as the method of the present invention can clearly define the different specific situations in the single determination process through the obtained value.
[0044] Specifically, the analysis module establishes coordinates for the generated 3D digital human, and the analysis module compares the trajectory curve of a single joint of the 3D digital human in the coordinates with the corresponding preset trajectory curve, and determines whether the operation of the single joint meets the preset standard joint determination method according to the obtained overlap, wherein:
[0045] The first joint determination method is that the analysis module determines that the operation of a single joint does not meet the preset standard, and determines the processing method for the single joint according to the difference between the first preset overlap and the overlap; the first joint determination method satisfies that the overlap is less than or equal to the first preset overlap;
[0046] The second joint determination method is that the analysis module preliminarily determines that the operation of a single joint does not meet the preset standard, and re-determines whether the operation of a single joint meets the preset standard based on the distance between the end point of the acquired trajectory curve of the single joint and the end point of the preset trajectory curve; the second joint determination method satisfies that the overlap is less than or equal to the second preset overlap and greater than the first preset overlap, and the first preset overlap is less than the second preset overlap;
[0047] The third joint determination method is that the analysis module determines that the operation of a single joint meets the preset standard, and detects the next trajectory curve of the single joint; the third joint determination method satisfies that the overlap is greater than the second preset overlap;
[0048] The overlap degree is the ratio of the length of the overlapped trajectory curve to the length of the preset trajectory curve.
[0049] Among them, the first preset overlap degree is 0.82, and the second preset overlap degree is 0.9.
[0050] Specifically, the analysis module calculates the ratio of the distance between the end point of the trajectory curve of a single joint and the end point of the preset trajectory curve to the length of the preset trajectory curve under the second joint determination mode, and records it as the distance ratio. The analysis module determines whether the operation of the single joint meets the preset standard joint secondary determination mode according to the obtained distance ratio, wherein:
[0051] The first joint secondary determination method is that the analysis module determines that the operation of a single joint meets the preset standard, and detects the next trajectory curve of the single joint; the first joint secondary determination method satisfies that the distance ratio is less than or equal to the preset distance ratio;
[0052] The second joint secondary judgment method is that the analysis module determines that the operation of a single joint does not meet the preset standard, and adjusts the preset rendering magnification to a corresponding value according to the difference between the distance ratio and the preset distance ratio; the second joint secondary judgment method satisfies that the distance ratio is greater than the preset distance ratio.
[0053] Among them, the preset distance ratio is 0.03.
[0054] Determine the actual operation status of each joint of the 3D digital human, determine whether the operation of a single joint meets the preset standard for the trajectory curve of a single joint, and when the overlap is less than or equal to the second preset overlap and greater than the first preset overlap, obtain the distance between the end point of the trajectory curve of the single joint and the end point of the preset trajectory curve to determine whether the joint fails to run to the corresponding point within the time period of obtaining the trajectory curve due to jamming. When the distance ratio is less than or equal to the preset distance ratio, the single joint completes the operation of the preset trajectory; when the distance ratio is greater than the preset distance ratio, the analysis module determines to lower the preset rendering magnification, so as to reduce the calculation burden, improve the running smoothness of the digital human, and further effectively improve the generation efficiency of the 3D digital human.
[0055] Specifically, the analysis module records the difference between the first preset overlap and the overlap as the overlap difference under the first joint determination mode, and determines the operation processing mode for a single joint according to the obtained overlap difference, wherein:
[0056] The first operation processing mode is that the analysis module adjusts the preset rendering magnification to a corresponding value according to the number of regions with abnormal overlap; the first operation processing mode satisfies that the overlap difference is less than or equal to the first preset overlap difference;
[0057] The second operation processing mode is that the analysis module determines whether to adjust the relative angle between the single joint and the upper joint to the corresponding value according to the determination result of the upper joint of the single joint; the second operation processing mode satisfies that the overlap difference is less than or equal to the second preset overlap difference and greater than the first preset overlap difference, and the first preset overlap difference is less than the second preset overlap difference;
[0058] The third operation processing method is that the analysis module adjusts the distance between the two joints to a corresponding value according to the ratio of the length of the trajectory curve to the length of the preset trajectory curve, and the analysis module redetermines the coordinates of the single joint based on the adjusted distance between the two joints, and re-determines whether the operation of the single joint meets the preset standard based on the overlap degree according to the redetermined coordinates of the single joint; the third operation processing method satisfies that the overlap difference is greater than the second preset overlap difference.
[0059] Among them, the first preset overlap difference is 0.2, and the second preset overlap difference is 0.43.
[0060] When the overlap degree is less than or equal to the first preset overlap degree, the corresponding processing method is specifically determined according to the overlap difference. When the overlap difference is less than or equal to the first preset overlap difference, the running trajectory only has a small range of deviations. In this case, the 3D digital human is stuck, and the preset rendering magnification is lowered; when the overlap difference is less than or equal to the second preset overlap difference and greater than the first preset overlap difference, the running trajectory of a single joint has a large range of deviations. In this case, the parent joint of the single joint is specifically determined, so that when the operation of the parent joint does not meet the preset standard, the parent joint is optimized in time to avoid the accuracy of the judgment of the operation of the single joint being affected by the abnormality of the parent joint; when the operation of the parent joint meets the preset standard, in this case, the trajectory curve deviates from the preset trajectory curve due to the abnormality of the relative angle between the single joint and the parent joint, so the relative angle between the single joint and the parent joint is adjusted, while effectively improving the accuracy of the operation of the 3D digital human, the generation efficiency of the 3D digital human is further effectively improved.
[0061] Specifically, the analysis module obtains the overlapped graph of the trajectory curve and the corresponding preset trajectory curve under the first operation processing mode, and marks each region with abnormal overlap as an abnormal region. The analysis module determines the rendering adjustment mode for the preset rendering magnification according to the number of abnormal regions obtained, wherein:
[0062] The first rendering adjustment method is that the analysis module uses a first preset rendering adjustment coefficient to adjust the preset rendering magnification to a corresponding value; the first rendering adjustment method satisfies that the number of abnormal areas is less than or equal to the first preset number;
[0063] The second rendering adjustment method is that the analysis module uses a second preset rendering adjustment coefficient to adjust the preset rendering magnification to a corresponding value; the second rendering adjustment method satisfies that the number of abnormal areas is less than or equal to the second preset number and greater than the first preset number, and the first preset number is less than the second preset number;
[0064] The third rendering adjustment method is that the analysis module uses a third preset rendering adjustment coefficient to adjust the preset rendering magnification to a corresponding value; the third rendering adjustment method satisfies that the number of abnormal areas is greater than the second preset number.
[0065] Among them, the first preset number is 2, the second preset number is 4, the first preset rendering adjustment coefficient is 0.9, the second preset rendering adjustment coefficient is 0.81, and the third preset rendering adjustment coefficient is 0.72.
[0066] Specifically, when the analysis module completes the adjustment of the rendering magnification, it re-determines whether the operation of the single joint meets the preset standard based on the re-acquired single joint. If the analysis module still determines to adjust the rendering magnification, the analysis module determines to use the current rendering magnification as the generation parameter for the 3D digital human, and adjusts the preset data volume to a corresponding value according to the difference between the second preset overlap and the re-acquired overlap;
[0067] The analysis module records the difference between the calculated second preset overlap and the re-acquired overlap as an update difference, and the analysis module determines a data adjustment method for the preset data amount according to the obtained update difference, wherein:
[0068] The first data adjustment method is that the analysis module uses a first preset data adjustment coefficient to adjust the preset data amount to a corresponding value; the first data adjustment method satisfies that the update difference is less than or equal to the first preset update difference;
[0069] The second data adjustment method is that the analysis module uses the second preset data adjustment coefficient to adjust the preset data amount to the corresponding value; the second data adjustment method satisfies that the update difference is less than or equal to the second preset update difference and greater than the first preset update difference, and the first preset update difference is less than the second preset update difference;
[0070] The third data adjustment method is that the analysis module uses a third preset data adjustment coefficient to adjust the preset data amount to a corresponding value; the third data adjustment method satisfies that the update difference is greater than the second preset update difference.
[0071] Among them, the first preset update difference is 0.25, the second preset update difference is 0.34, the first preset data adjustment coefficient is 1.12, the second preset data adjustment coefficient is 1.21, and the third preset data adjustment coefficient is 1.31.
[0072] Specifically, the analysis module obtains the determination result of the superior joint of the single joint under the second operation processing mode; when it is determined that the operation of the superior joint meets the preset standard, the relative angle between the two joints is adjusted to the corresponding value according to the second pre-overlapping difference and the lower level difference of the overlap difference; when it is determined that the operation of the superior joint does not meet the preset standard, the superior joint is processed accordingly, and the single joint is re-detected after the processing is completed, and whether the operation of the single joint meets the preset standard is determined according to the re-acquired overlap;
[0073] The relative angle between the two joints is the angle between the two line segments whose intersection points are obtained by connecting the joints. (If there are two adjacent joints of the superior joint, the superior joint of the superior joint is selected as one end of one of the line segments, and a single joint is selected as one end of the other line segment).
[0074] (Using the upright form of the 3D digital human as a reference, each joint is determined, and the level is determined in descending order of joint height. Joints at the same height are marked as the same level. For a single joint, the joint that is higher than the single joint and closest to it is the superior joint of the single joint.)
[0075] Specifically, the analysis module determines the angle adjustment method for the relative angle according to a lower level difference between the second pre-coincidence difference and the coincidence difference, wherein:
[0076] The first angle adjustment method is that the analysis module uses the first preset angle adjustment coefficient to adjust the relative angle between the two joints to a corresponding value; the first angle adjustment method satisfies that the lower level difference is less than or equal to the first preset lower level difference;
[0077] The second angle adjustment method is that the analysis module uses the second preset angle adjustment coefficient to adjust the relative angle between the two joints to the corresponding value; the second angle adjustment method satisfies that the lower level difference is less than or equal to the second preset lower level difference and greater than the first preset lower level difference, and the first preset lower level difference is less than the second preset lower level difference;
[0078] The third angle adjustment method is that the analysis module uses the third preset angle adjustment coefficient to adjust the relative angle between the two joints to a corresponding value; the third angle adjustment method satisfies that the lower level difference is greater than the second preset lower level difference.
[0079] Among them, the first preset lower level difference is 0.17, the second preset lower level difference is 0.23, the first preset angle adjustment coefficient is 1.28, the second preset angle adjustment coefficient is 1.21, and the third preset angle adjustment coefficient is 1.13.
[0080] Specifically, the analysis module records the ratio of the length of the trajectory curve to the length of the preset trajectory curve as the length ratio in the third operation processing mode, and the analysis module determines the adjustment method for the distance between the two joints according to the obtained length ratio, wherein:
[0081] The first distance adjustment method is that the analysis module uses a first preset distance adjustment coefficient to adjust the distance between the two joints to a corresponding value; the first distance adjustment method satisfies that the length ratio is less than or equal to the first preset length ratio;
[0082] The second distance adjustment method is that the analysis module uses the second preset distance adjustment coefficient to adjust the distance between the two joints to a corresponding value; the second distance adjustment method satisfies that the length ratio is less than or equal to the second preset length ratio and greater than the first preset length ratio, and the first preset length ratio is less than the second preset length ratio;
[0083] The third distance adjustment method is that the analysis module uses the third preset distance adjustment coefficient to adjust the distance between the two joints to a corresponding value; the third distance adjustment method satisfies that the length ratio is greater than the second preset length ratio.
[0084] Among them, the first preset length ratio is 0.47, the second preset length ratio is 0.62, the first preset distance adjustment coefficient is 1.1, the second preset distance adjustment coefficient is 1.17, and the third preset distance adjustment coefficient is 1.28.
[0085] Specifically, the analysis module records the difference between the distance ratio and the preset distance ratio as the distance difference under the second joint secondary determination mode, and the analysis module determines the model adjustment mode for the preset rendering magnification based on the obtained distance difference, wherein:
[0086] The first model adjustment method is that the analysis module uses the first preset model adjustment coefficient to adjust the preset rendering magnification to a corresponding value; the first model adjustment method satisfies that the distance difference is less than or equal to the first preset distance difference;
[0087] The second model adjustment method is that the analysis module uses the second preset model adjustment coefficient to adjust the preset rendering magnification to a corresponding value; the second model adjustment method satisfies that the distance difference is less than or equal to the second preset distance difference and greater than the first preset distance difference, and the first preset distance difference is less than the second preset distance difference;
[0088] The third model adjustment method is that the analysis module uses a third preset model adjustment coefficient to adjust the preset rendering magnification to a corresponding value; the third model adjustment method satisfies that the distance difference is greater than the second preset distance difference.
[0089] Among them, the first preset distance difference is 0.065, the second preset distance difference is 0.131, the first preset model adjustment coefficient is 0.91, the second preset model adjustment coefficient is 0.88, and the third preset model adjustment coefficient is 0.84.
[0090] So far, the technical solutions of the present invention have been described in conjunction with the preferred embodiments shown in the accompanying drawings. However, it is easy for those skilled in the art to understand that the protection scope of the present invention is obviously not limited to these specific embodiments. Without departing from the principle of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will fall within the protection scope of the present invention.
[0091] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.
Claims
1. A 3D digital human generation method based on AIGC, characterized in that: include: Data collection and preprocessing of the collected data, including data cleaning, reconstruction, and alignment; Perform feature extraction on the preprocessed data and generate the initial model of the 3D digital human based on the extracted key feature points; Animation generation: adding animation to the 3D digital human model to make it move according to pre-designed actions; rendering settings: setting the preset rendering magnification in the rendering engine to control the clarity of the image; Data transmission: the server downloads the 3D digital human model data, sets a cache area with a preset data volume to cache the data locally, parses and processes the transmitted data, caches the calculation results during the calculation of skeletal animation, generates a 3D digital human and presents it on the screen; Data adjustment: establishing coordinates for the generated 3D digital human, and comparing the trajectory curve of a single joint of the 3D digital human in the coordinates with the corresponding preset trajectory curve, and determining whether the operation of the single joint meets the preset standard according to the obtained overlap, and when it is preliminarily determined that the operation of the single joint does not meet the preset standard, determining whether to adjust the preset rendering magnification according to the distance between the end point of the obtained trajectory curve of the single joint and the end point of the preset trajectory curve; Data generation: when the operation of each joint of the 3D digital human meets the preset standards, an indication of the completion of the generation of the 3D digital human is output; The analysis module establishes coordinates for the generated 3D digital human, and compares the trajectory curve of a single joint of the 3D digital human in the coordinates with the corresponding preset trajectory curve, and determines whether the operation of the single joint meets the preset standard according to the obtained overlap, and when it is determined that the operation of the single joint does not meet the preset standard, determines the processing method for the single joint according to the difference between the first preset overlap and the overlap, or when it is preliminarily determined that the operation of the single joint does not meet the preset standard, re-determines whether the operation of the single joint meets the preset standard according to the distance between the end point of the acquired trajectory curve of the single joint and the end point of the preset trajectory curve; The overlap degree is the ratio of the length of the overlapped trajectory curve to the length of the preset trajectory curve.
2. The AIGC-based 3D digital human generation method according to claim 1, characterized in that: The analysis module calculates the ratio of the distance between the end point of the trajectory curve of a single joint and the end point of the preset trajectory curve to the length of the preset trajectory curve, and records it as the distance ratio. The analysis module determines whether the operation of the single joint meets the preset standard based on the obtained distance ratio, and when it is determined that the operation of the single joint does not meet the preset standard, the preset rendering magnification is adjusted to a corresponding value based on the difference between the distance ratio and the preset distance ratio.
3. The AIGC-based 3D digital human generation method according to claim 2, characterized in that: The analysis module records the difference between the first preset overlap and the overlap as the overlap difference, and determines the operation processing method for the single joint based on the obtained overlap difference, including adjusting the preset rendering magnification to a corresponding value according to the number of areas with overlap abnormalities, and determining whether to adjust the relative angle between the single joint and the superior joint to a corresponding value according to the judgment result of the superior joint of the single joint, and adjusting the distance between the two joints to a corresponding value according to the ratio of the length of the trajectory curve to the length of the preset trajectory curve. The analysis module redetermines the coordinates of the single joint based on the adjusted distance between the two joints, and re-determines whether the operation of the single joint meets the preset standard based on the overlap according to the redetermined coordinates of the single joint.
4. The AIGC-based 3D digital human generation method according to claim 3, characterized in that: The analysis module obtains the overlapping graph of the trajectory curve and the corresponding preset trajectory curve, and marks each area with overlapping abnormalities as an abnormal area. The analysis module is provided with a plurality of rendering adjustment methods for the preset rendering magnification based on the number of abnormal areas obtained, and the adjustment range of each rendering adjustment method for the preset rendering magnification is different.
5. The AIGC-based 3D digital human generation method according to claim 4, characterized in that: When the analysis module completes the adjustment of the rendering magnification, it re-determines whether the operation of the single joint meets the preset standard based on the re-acquired single joint. If the analysis module still determines to adjust the rendering magnification, the analysis module determines to use the current rendering magnification as the generation parameter for the 3D digital human, and adjusts the preset data volume to a corresponding value according to the difference between the second preset overlap and the re-acquired overlap; The analysis module is provided with a plurality of data adjustment methods for the preset data volume based on the difference between the second preset overlap degree and the re-acquired overlap degree, and the adjustment range of each data adjustment method for the preset data volume is different.
6. The AIGC-based 3D digital human generation method according to claim 5, characterized in that: The analysis module obtains the determination result of the superior joint of the single joint; when it is determined that the operation of the superior joint meets the preset standard, the relative angle between the two joints is adjusted to a corresponding value according to the second pre-overlap difference and the lower level difference of the overlap difference; When it is determined that the operation of the upper joint does not meet the preset standard, the upper joint is processed accordingly, and the single joint is re-tested after the processing is completed, and the analysis mechanism determines whether the operation of the single joint meets the preset standard based on the re-acquired overlap; Among them, the relative angle between the two joints is the angle between the two line segments whose intersection points are the upper joints obtained by connecting the joints.
7. The AIGC-based 3D digital human generation method according to claim 6, characterized in that: The analysis module is provided with a plurality of angle adjustment methods for the relative angle based on a lower level difference between the second pre-overlap difference and the overlap difference, and each angle adjustment method has a different adjustment range for the relative angle between the two joints.
8. The AIGC-based 3D digital human generation method according to claim 7, characterized in that: The analysis module records the ratio of the length of the trajectory curve to the length of the preset trajectory curve as the length ratio. Based on the obtained length ratio, the analysis module sets several adjustment methods for the distance between the two joints, and the adjustment range of each distance adjustment method for the distance between the two joints is different.
9. The AIGC-based 3D digital human generation method according to claim 8, characterized in that: The analysis module records the difference between the distance ratio and the preset distance ratio as the distance difference. The analysis module sets several model adjustment methods for the preset rendering ratio based on the obtained distance difference, and the adjustment range of each model adjustment method for the preset rendering ratio is different.
Citation Information
Patent Citations
Virtual digital human model planning system based on AIGC algorithm
CN116630495A
Animation generation method and device, equipment and storage medium
CN116934913A
Animation creation system for face scanning and animation character generation
CN117710542A
Cited By
Cultural feature digital human generation method fusing LoRA fine tuning and cross-modal interaction
CN121414932A