Feature extraction method and system for human body three-dimensional scanning data
By preprocessing and matching feature points on the dynamic three-dimensional point cloud data on the human body surface, it integrates into a dynamic three-dimensional feature map, and solves the problem of static state and a large number of manual labels in the existing technology of human body feature extraction, achieving more accurate feature extraction.
Patent Information
- Application Number
- CN202411770302.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-12-04
- Publication Date
- 2025-04-25
AI Technical Summary
The prior art requires scanning in the static state of the human body in the extraction of human body characteristics, with high posture correlation and a large amount of manual labeling, resulting in inaccurate extraction results.
A feature extraction method for human body three-dimensional scanning data is adopted. By pre-processing the dynamic three-dimensional point cloud data on the human body surface, multiple static three-dimensional point cloud data on the human body surface are obtained, matching the feature points in the feature area map, and integrating them into a dynamic three-dimensional feature map.
Dynamic feature extraction is realized, the accuracy of the extraction results is improved, and the need for manual labeling is reduced.
Smart Images
Figure CN119888067A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of three-dimensional scanning technology, and in particular to a method and system for extracting features from three-dimensional scanning data of a human body. Background Art
[0002] With the rapid development of 3D scanning technology, its application in virtual human modeling, clothing customization, medical diagnosis and other fields is becoming more and more extensive. At present, in medical diagnosis, 3D scanning technology can be used to extract human body features, especially skeleton information.
[0003] However, in the current human feature extraction, it is necessary to scan the human body in a static state, the posture correlation is high, and a large amount of manual annotation is required, which leads to inaccurate extraction results. Summary of the invention
[0004] The purpose of the present invention is to provide a method and system for extracting features from three-dimensional human body scanning data, aiming to solve the technical problems in the prior art that in human body feature extraction, the human body needs to be scanned in a static state, the posture correlation is high, and a large amount of manual annotation is required, which leads to inaccurate extraction results.
[0005] To achieve the above object, the present invention adopts a method for extracting features of three-dimensional human body scanning data, comprising the following steps:
[0006] Scanning a human body, collecting raw data of dynamic three-dimensional points on the human body surface, and preprocessing the raw data to obtain dynamic three-dimensional point cloud data on the human body surface; wherein the dynamic three-dimensional point cloud data on the human body surface includes a plurality of static three-dimensional point cloud data on the human body surface;
[0007] Input the original feature region map, match the feature points with the static three-dimensional point cloud data of the human body surface, adjust the feature points in the feature region map, and output the matching feature region map;
[0008] The matching feature area maps of multiple static three-dimensional point cloud data of the human body surface are obtained and integrated into a dynamic three-dimensional feature map.
[0009] Among them, in the step of scanning the human body, collecting the original data of dynamic three-dimensional points on the surface of the human body, and preprocessing the original data to obtain the dynamic three-dimensional point cloud data on the surface of the human body:
[0010] Use a 3D scanner to scan the human body and obtain the original data of dynamic 3D points on the human body surface;
[0011] De-noising and smoothing the original data of dynamic three-dimensional points on the human body surface;
[0012] Obtain static 3D point cloud data of the human body surface for each frame.
[0013] Before obtaining the static three-dimensional point cloud data of the human body surface of each frame:
[0014] Divide the time nodes and capture frames according to the node positioning.
[0015] Among them, in the steps of inputting the original feature region map, matching the feature points with the static three-dimensional point cloud data of the human body surface, adjusting the feature points in the feature region map, and outputting the matching feature region map:
[0016] Input the original feature area map of the current scanned human body;
[0017] Match the positioning feature points and the target feature points, adjust the positioning feature points, and output the adjusted positioning feature points as dynamic feature points.
[0018] Among them, after the steps of matching the positioning feature points and the target feature points, adjusting the positioning feature points, and outputting the adjusted positioning feature points as dynamic feature points:
[0019] Output dynamic feature points to the feature area map, update the feature point area, and output the matching feature area map of the current frame.
[0020] Among them, in the step of obtaining matching feature area maps of multiple static three-dimensional point cloud data of human body surfaces and integrating them into dynamic three-dimensional feature maps:
[0021] According to the time nodes, the matching feature area maps are arranged, and interpolation calculations are performed between the same feature point areas in each matching feature area map to generate transition dynamic data;
[0022] Generate animation frames using transition dynamics data.
[0023] Wherein, after the step of generating animation frames using transition dynamic data:
[0024] Verify the features in the animation frames, optimize and adjust them, and then output a dynamic 3D feature map.
[0025] The present invention also provides a feature extraction system for three-dimensional human body scanning data, comprising a data acquisition module, a feature point adjustment module and a region integration module, wherein:
[0026] The data acquisition module is used to scan the human body, collect the original data of dynamic three-dimensional points on the human body surface, and pre-process the original data to obtain dynamic three-dimensional point cloud data on the human body surface;
[0027] The feature point adjustment module is used to input the original feature area map, match the feature points with the static three-dimensional point cloud data of the human body surface, adjust the feature points in the feature area map, and output the matching feature area map;
[0028] The region integration module is used to obtain matching feature region maps of multiple static three-dimensional point cloud data on the human body surface and integrate them into a dynamic three-dimensional feature map.
[0029] The present invention provides a method and system for extracting features from three-dimensional scanning data of a human body. The method and system respectively use the data acquisition module, the feature point adjustment module and the region integration module to perform the following steps: scanning the human body, collecting original data of dynamic three-dimensional points on the surface of the human body, and preprocessing the original data to obtain dynamic three-dimensional point cloud data on the surface of the human body; wherein the dynamic three-dimensional point cloud data on the surface of the human body includes a plurality of static three-dimensional point cloud data on the surface of the human body; inputting an original feature region map, matching feature points with the static three-dimensional point cloud data on the surface of the human body, adjusting feature points in the feature region map, and outputting a matching feature region map; wherein the feature points are human joint points; obtaining matching feature region maps of multiple static three-dimensional point cloud data on the surface of the human body, and integrating them into a dynamic three-dimensional feature map; achieving dynamic feature extraction and improving the accuracy of the extraction result. BRIEF DESCRIPTION OF THE DRAWINGS
[0030] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.
[0031] Figure 1 It is a flow chart of the steps of the method for extracting features from three-dimensional human body scanning data of the present invention.
[0032] Figure 2 It is a step flow chart of S100 of the present invention.
[0033] Figure 3 It is a step flow chart of S200 of the present invention.
[0034] Figure 4 It is a step flow chart of S300 of the present invention.
[0035] Figure 5 It is a schematic diagram of the original characteristic region map of the present invention.
[0036] Figure 6 It is a schematic diagram of a matching feature area map after updating the feature point area of the present invention.
[0037] Figure 7 It is a schematic diagram of the principle of the feature extraction system of the human body three-dimensional scanning data of the present invention.
[0038] Figure 8 It is a schematic diagram of the principle of the electronic device of the present invention.
[0039] 401 - data acquisition module, 402 - feature point adjustment module, 403 - regional integration module. DETAILED DESCRIPTION
[0040] Here, exemplary embodiments are described in detail, and examples thereof are shown in the accompanying drawings. When the following description refers to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The implementations described in the following exemplary embodiments do not represent all implementations consistent with the present application.
[0041] The terms used in this application are for the purpose of describing specific embodiments only and are not intended to limit this application. The singular forms of "a", "said" and "the" used in this application and the appended claims are also intended to include plural forms unless the context clearly indicates other meanings. It should also be understood that the term "and / or" used herein refers to and includes any or all possible combinations of one or more associated listed items.
[0042] It should be understood that although the terms first, second, third, etc. may be used in the present application to describe various information, these information should not be limited to these terms. These terms are only used to distinguish the same type of information from each other. For example, without departing from the scope of the present application, the first information may also be referred to as the second information, and similarly, the second information may also be referred to as the first information. Depending on the context, the word "if" as used herein may be interpreted as "at the time of" or "when" or "in response to determining".
[0043] See also Figure 1 to Figure 6 The present invention provides a method for extracting features from three-dimensional human body scanning data, comprising the following steps:
[0044] S100: Scanning a human body, collecting raw data of dynamic three-dimensional points on the surface of the human body, and preprocessing the raw data to obtain dynamic three-dimensional point cloud data on the surface of the human body;
[0045] In this embodiment, the human body is first scanned to collect original data of dynamic three-dimensional points on the human body surface, and the original data is preprocessed to obtain dynamic three-dimensional point cloud data on the human body surface; wherein the dynamic three-dimensional point cloud data on the human body surface includes multiple static three-dimensional point cloud data on the human body surface.
[0046] The specific steps are:
[0047] S101: Scanning a human body with a three-dimensional scanner to obtain original data of dynamic three-dimensional points on the surface of the human body;
[0048] S102: performing denoising and smoothing processing on the original data of dynamic three-dimensional points on the human body surface;
[0049] S103: Divide the time nodes, intercept frames according to the node positioning, and obtain static three-dimensional point cloud data of the human body surface in each frame.
[0050] The human body is scanned with a 3D scanner to obtain the original data of dynamic 3D points on the human body surface; then the original data of dynamic 3D points on the human body surface is denoised and smoothed; the time nodes are divided, and frames are captured according to node positioning to obtain static 3D point cloud data of the human body surface in each frame.
[0051] S200: inputting an original feature region map, matching feature points with static three-dimensional point cloud data of a human body surface, adjusting feature points in the feature region map, and outputting a matching feature region map;
[0052] In this embodiment, after obtaining the dynamic three-dimensional point cloud data of the human body surface, the original feature area map is input, the feature points are matched with the static three-dimensional point cloud data of the human body surface, the feature points in the feature area map are adjusted, and the matching feature area map is output.
[0053] The specific steps are:
[0054] S201: inputting the original feature region map of the current scanned human body;
[0055] S202: matching the positioning feature points with the target feature points, adjusting the positioning feature points, and outputting the adjusted positioning feature points as dynamic feature points;
[0056] S203: Output the dynamic feature points to the feature region map, update the feature point region, and output the matching feature region map of the current frame.
[0057] By inputting the original characteristic region map of the current scanned human body according to the height and weight of the current scanned human body, the original characteristic region map can be a region map obtained by scanning a normal human body, such as Figure 5 As shown, Figure 5 The filled area in is the feature point, and the feature point is the human joint point; the feature point in the original feature area map is defined as the positioning feature point, and the feature in the static 3D point cloud data of the human surface is the target feature point. Then match the positioning feature point and the target feature point, adjust the positioning feature point, and output the adjusted positioning feature point as the dynamic feature point. Then output the dynamic feature point to the feature area map, update the feature point area, and output the matching feature area map of the current frame, such as Figure 6 As shown, Figure 6 The filled area in the figure is the adjusted human joint points.
[0058] S300: Acquire matching feature area maps of multiple static three-dimensional point cloud data on the surface of a human body, and integrate them into a dynamic three-dimensional feature map.
[0059] In this embodiment, after the matching feature region map of the current frame is output, the matching feature region maps of multiple static three-dimensional point cloud data of the human body surface are obtained and integrated into a dynamic three-dimensional feature map.
[0060] The specific steps are:
[0061] S301: arranging matching feature region maps according to time nodes, performing interpolation calculations between regions with the same feature points in each matching feature region map, and generating transition dynamic data;
[0062] S302: Generate animation frames using transition dynamic data;
[0063] S303: Verify the features in the animation frame, optimize and adjust, and then output a dynamic three-dimensional feature map.
[0064] By arranging the matching feature region maps according to the time nodes, interpolation calculation is performed between the same feature point regions in each matching feature region map to generate transition dynamic data; the interpolation calculation can be performed in a linear interpolation manner, such as:
[0065] Determine two key frames (or static skeleton feature points) A and B to be interpolated, A is the first human joint point in the first frame matching feature area map, and B is the first human joint point in the second frame matching feature area map.
[0066] Determine the position of the interpolation point C on the time axis, that is, the ratio t of C between A and B, where 0≤t≤1.
[0067] Calculate the position of the interpolation point and interpolate the position of the joint point using the formula:
[0068] P_C=P_A+t*(P_B-P_A),
[0069] Where P_A and P_B are the position vectors of joint points A and B respectively, and P_C is the position vector of interpolation point C.
[0070] For the rotation of the joint points, quaternion interpolation is used:
[0071] Convert Euler angles to quaternions q_A and q_B.
[0072] Calculate the interpolated quaternion using the spherical linear interpolation formula:
[0073] q_C:q_C=q_A*(q_At);
[0074] Where t represents the tth power of the quaternion.
[0075] Convert the interpolated quaternion q_C back to Euler angles to get the interpolated rotation angle. Generate animation frames using transition dynamic data. Verify the features in the animation frames, optimize and adjust them, and output dynamic 3D feature maps.
[0076] Corresponding to the aforementioned embodiment of the method for extracting features from three-dimensional human body scanning data, the present application also provides an embodiment of a system for extracting features from three-dimensional human body scanning data.
[0077] Figure 7 is a block diagram of a system for extracting features from three-dimensional human body scan data according to an exemplary embodiment. Figure 7 The system may include: a data acquisition module 401, a feature point adjustment module 402 and a region integration module 403, wherein:
[0078] The data acquisition module 401 is used to scan the human body, collect the original data of dynamic three-dimensional points on the human body surface, and pre-process the original data to obtain dynamic three-dimensional point cloud data on the human body surface;
[0079] The feature point adjustment module 402 is used to input the original feature region map, match the feature points with the static three-dimensional point cloud data of the human body surface, adjust the feature points in the feature region map, and output the matching feature region map;
[0080] The region integration module 403 is used to obtain matching feature region maps of multiple static three-dimensional point cloud data on the human body surface and integrate them into a dynamic three-dimensional feature map.
[0081] In this embodiment, the data acquisition module 401 scans the human body, collects the original data of dynamic three-dimensional points on the human body surface, and preprocesses the original data to obtain dynamic three-dimensional point cloud data on the human body surface; the feature point adjustment module 402 inputs the original feature area map, matches the feature points with the static three-dimensional point cloud data on the human body surface, adjusts the feature points in the feature area map, and outputs the matching feature area map; the area integration module 403 obtains the matching feature area maps of multiple static three-dimensional point cloud data on the human body surface, and integrates them into a dynamic three-dimensional feature map, thereby realizing dynamic feature extraction and improving the accuracy of the extraction results.
[0082] Regarding the system in the above embodiment, the specific manner in which each module performs operations has been described in detail in the embodiment of the method, and will not be elaborated here.
[0083] For the system embodiment, since it basically corresponds to the method embodiment, the relevant parts can refer to the partial description of the method embodiment. The device embodiment described above is only schematic, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the present application. A person of ordinary skill in the art can understand and implement it without creative work.
[0084] Accordingly, the present application also provides an electronic device, comprising: one or more processors; a memory for storing one or more programs; when the one or more programs are executed by the one or more processors, the one or more processors implement the above-mentioned method for extracting features from three-dimensional human body scanning data. Figure 8 As shown in FIG. 1 , a hardware structure diagram of a human body three-dimensional scanning data feature extraction system provided by an embodiment of the present invention is provided in any device with data processing capability, except Figure 8 In addition to the processor, memory and network interface shown, any device with data processing capability in which the apparatus in the embodiment is located may also include other hardware, generally based on the actual functions of the device with data processing capability, which will not be described in detail.
[0085] Accordingly, the present application also provides a computer-readable storage medium on which computer instructions are stored, and when the instructions are executed by a processor, the feature extraction method of the three-dimensional scanning data of the human body as described above is implemented. The computer-readable storage medium can be an internal storage unit of any device with data processing capabilities described in any of the aforementioned embodiments, such as a hard disk or a memory. The computer-readable storage medium can also be an external storage device, such as a plug-in hard disk, a smart memory card (Smart Media Card, SMC), an SD card, a flash card (Flash Card), etc. equipped on the device. Furthermore, the computer-readable storage medium can also include both an internal storage unit of any device with data processing capabilities and an external storage device. The computer-readable storage medium is used to store the computer program and other programs and data required by any device with data processing capabilities, and can also be used to temporarily store data that has been output or is to be output.
[0086] Those skilled in the art will readily appreciate other embodiments of the present application after considering the specification and practicing the contents disclosed herein. The present application is intended to cover any variations, uses or adaptations of the present application, which follow the general principles of the present application and include common knowledge or customary technical means in the art that are not disclosed in the present application.
[0087] It should be understood that the present application is not limited to the exact construction that has been described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof.
Claims
1. A method for extracting features from three-dimensional human body scanning data, characterized in that: The steps include: Scanning a human body, collecting raw data of dynamic three-dimensional points on the surface of the human body, and preprocessing the raw data to obtain dynamic three-dimensional point cloud data on the surface of the human body; wherein the dynamic three-dimensional point cloud data on the surface of the human body includes a plurality of static three-dimensional point cloud data on the surface of the human body; Input the original feature region map, match the feature points with the static three-dimensional point cloud data of the human body surface, adjust the feature points in the feature region map, and output the matching feature region map; The matching feature area maps of multiple static three-dimensional point cloud data of the human body surface are obtained and integrated into a dynamic three-dimensional feature map.
2. The method for extracting features from three-dimensional human body scan data according to claim 1, characterized in that: In the steps of scanning a human body, collecting raw data of dynamic three-dimensional points on the surface of the human body, and preprocessing the raw data to obtain dynamic three-dimensional point cloud data on the surface of the human body: Use a 3D scanner to scan the human body and obtain the original data of dynamic 3D points on the human body surface; De-noising and smoothing the original data of dynamic three-dimensional points on the human body surface; Obtain static 3D point cloud data of the human body surface for each frame.
3. The method for extracting features from three-dimensional human body scan data according to claim 2, characterized in that: Before the step of obtaining static 3D point cloud data of the human body surface for each frame: Divide the time nodes and capture frames according to the node positioning.
4. The method for extracting features from three-dimensional human body scan data according to claim 1, characterized in that: In the steps of inputting the original feature region map, matching the feature points with the static 3D point cloud data of the human body surface, adjusting the feature points in the feature region map, and outputting the matching feature region map: Input the original feature area map of the current scanned human body; Match the positioning feature points and the target feature points, adjust the positioning feature points, and output the adjusted positioning feature points as dynamic feature points.
5. The method for extracting features from three-dimensional human body scan data according to claim 4, characterized in that: After matching the positioning feature points and the target feature points, adjusting the positioning feature points, and outputting the adjusted positioning feature points as dynamic feature points: Output dynamic feature points to the feature area map, update the feature point area, and output the matching feature area map of the current frame.
6. The method for extracting features from three-dimensional human body scan data as claimed in claim 3, characterized in that: In the step of obtaining matching feature area maps of multiple static three-dimensional point cloud data of human body surfaces and integrating them into dynamic three-dimensional feature maps: According to the time nodes, the matching feature area maps are arranged, and interpolation calculations are performed between the same feature point areas in each matching feature area map to generate transition dynamic data; Generate animation frames using transition dynamics data.
7. The method for extracting features from three-dimensional human body scan data according to claim 6, characterized in that: After the step of generating animation frames using transition dynamic data: Verify the features in the animation frames, optimize and adjust them, and then output a dynamic 3D feature map.
8. A system for extracting features from three-dimensional human body scan data, applied to the method for extracting features from three-dimensional human body scan data as claimed in claim 1, characterized in that: It includes data acquisition module, feature point adjustment module and regional integration module, among which: The data acquisition module is used to scan the human body, collect the original data of dynamic three-dimensional points on the human body surface, and pre-process the original data to obtain dynamic three-dimensional point cloud data on the human body surface; The feature point adjustment module is used to input the original feature area map, match the feature points with the static three-dimensional point cloud data of the human body surface, adjust the feature points in the feature area map, and output the matching feature area map; The region integration module is used to obtain matching feature region maps of multiple static three-dimensional point cloud data on the human body surface and integrate them into a dynamic three-dimensional feature map.