Personalized mannequin and manufacturing method, system and device thereof

Abstract

The invention provides a personalized mannequin and a manufacturing method, system and device thereof, and relates to the technical field of layered mannequin, and the method mainly comprises the steps: collecting three-dimensional point cloud data of a human body sample; generating a three-dimensional model and extracting a human body sagittal plane section point cloud data text; segmenting the sagittal plane section point cloud data into a front part and a rear part, and respectively writing the front part and the rear part into corresponding contour line data texts; through a BP neural network, optimal contour lines of the front part and the rear part are obtained through fitting prediction; calculating a zero cross point through a first-order derivative, and determining feature points of the front and rear parts; segmenting the three-dimensional model to obtain a plurality of human body parts of a front part and a rear part and corresponding processing data; sending the processing data to plate cutting equipment for manufacturing; and the plates are sequentially assembled and fixed to the positioning tool, and the mannequin is obtained. According to the scheme, while personalized body shape information is reserved, the manufacturing cost can be greatly reduced, and the professional layered mannequin meeting the costume design requirement can be rapidly and accurately manufactured.

Description

Technical Field

[0001] The present invention relates to the technical field of layered mannequins, and in particular to a personalized mannequin and a manufacturing method, system and device thereof. Background Art

[0002] As people's demand for clothing fit increases, the clothing industry has gradually introduced three-dimensional measurement technology to carry out fine modeling of the human body. Existing fine human body models are mainly used in the field of virtual fitting, but are rarely used in the field of mannequins. The mannequin is a physical model made according to the usual proportions of the human body, which is convenient for clothing design and cutting and production, but it often has only fixed sizes and cannot meet the individual needs of consumers with different body shapes for tailored (Made to Measure, referred to as MTM) clothing.

[0003] At present, there are personalized mannequins made using 3D printing technology. Although the personalized mannequins can replicate the body shapes of different human bodies, the cost of this technology is relatively high, and the material is hard and rough, which is not suitable as a physical model for clothing design.

[0004] Therefore, it is urgent to develop a low-cost, high-precision personalized mannequin manufacturing method to meet the actual needs of individual consumers for tailor-made clothing and clothing design teaching. Summary of the invention

[0005] The object of the present invention is to provide a personalized mannequin and a manufacturing method, system and device thereof, so as to solve at least one of the above-mentioned technical problems existing in the prior art.

[0006] In a first aspect, in order to solve the above technical problems, the present invention provides a method for manufacturing a personalized mannequin, comprising the following steps: Step 1: The client collects 3D point cloud data of a human sample through a 3D human body scanner and sends it to the server together with the human sample information; the 3D point cloud data establishes a coordinate system based on the axes in human anatomy: the Y axis is the sagittal axis (the axis from front to back of a standing human body and perpendicular to the vertical axis), the X axis is the coronal axis (the axis from left to right of a standing human body and perpendicular to both the sagittal axis and the vertical axis), and the Z axis is the vertical axis (the axis from head to tail of a standing human body and perpendicular to the ground plane); the human sample information includes customer ID, name, age, weight, height, skin color, and collection date, etc. In a feasible implementation manner, the step 1 specifically includes: Step 11, prepare human body samples according to the requirements of 3D point cloud collection; the requirements of 3D point cloud collection include: wearing a white tights, a white swimming cap, and removing accessories; standing with feet 10-20cm apart, head held high, chest held high, parallel, and eyes looking forward; forearms hanging naturally, upper arms abducted at 20 degrees to the body, and fists clenched; breathing normally, etc.; Step 12: Perform a three-dimensional scan on the human body sample using a three-dimensional human body scanner to obtain three-dimensional point cloud data; Step 13: After encrypting and compressing the 3D point cloud data, the client sends it to the server.

[0007] Through the above steps, three-dimensional point cloud data of customer body samples can be collected by three-dimensional body scanners in physical stores, experience stores, etc., and sent to the server through the client, which is conducive to building a data foundation for personalized mannequin manufacturing; and fully protecting the privacy of customer body shape information.

[0008] Step 2: The server generates a three-dimensional surface model from the three-dimensional point cloud data and extracts the human body sagittal section point cloud data text.

[0009] In a feasible implementation manner, the step 2 specifically includes: Step 21, the server decompresses and decrypts the three-dimensional point cloud data to generate a three-dimensional surface model of the human body sample; Step 22: extract all point cloud data with an X-axis coordinate of 0 in the three-dimensional point cloud data, that is, the point cloud data on the reference plane where the left-right center line of the human body sample is located, as the sagittal section point cloud data text.

[0010] Step 3. In the sagittal section point cloud data text, take the point cloud at the minimum Z-axis coordinate as the three-dimensional origin, and take the vertical axis of the three-dimensional origin as the reference to divide the sagittal section point cloud data into two parts, namely, the two parts with Y-axis coordinate y≥0 and the two parts with Y-axis coordinate y<0, and write the corresponding contour line data texts respectively.

[0011] In a feasible implementation manner, the step 3 specifically includes: Step 31: The server loads the sagittal cross-section point cloud data text, extracts the second and third columns of data in the point cloud coordinates, and defines them as Z-axis coordinate value z and Y-axis coordinate value y, respectively; Step 32: Count the minimum value min_Z of the Z-axis coordinate value z. The specific formula can be: min_Z = min(z); Step 33: Subtract min_Z from each z to obtain the updated coordinate value z', which is used to determine the position of the three-dimensional origin. The specific formula can be: z' = z - min_Z; Step 34: Based on the vertical axis of the three-dimensional origin, the sagittal cross-section point cloud data is divided into two parts, front and back, through conditional indexing, and the corresponding contour line data texts are written respectively: The conditional indexes in the first part include z>0 and y≥0; The conditional indexing of the latter part includes z>0 and y<0; In this way, the sagittal cross-section point cloud data can be divided into two parts, front and back, so that the mannequin can be manufactured separately in the subsequent sequence; of course, the encoding and direction of the coordinate axis can also be changed according to actual needs to facilitate subsequent processing.

[0012] Step 4: Based on the contour line data text, the BP neural network is used to fit and predict the optimal contour lines of the front and back parts.

[0013] In a feasible implementation manner, the training process of the BP neural network includes: Step 41: The server loads the contour data text, performs standardization, and randomly divides it into a training set, a validation set, and a test set; Step 42: Create a BP neural network: The number of hidden layers (feedforwardnet) is set to 10; The training cycle (trainParam.epochs) is set to 1000; The training target error (trainParam.goal) is set to 1e-5; Step 43, train the BP neural network until the correlation coefficient R values ​​of the training set, the validation set, the test set and the entire set are all greater than or equal to 0.9, so that the BP neural network has a reliable fitting prediction capability.

[0014] Step 5: Based on the optimal contour line, update the three-dimensional surface model, calculate the zero crossing point through the first-order derivative, and determine the feature points of the front and back parts.

[0015] In a feasible implementation manner, the specific steps of determining the feature points in step 5 include: Step 51, calculating the first-order derivative of the optimal contour line; Step 52: taking the zero crossing point of the first-order derivative as a feature point to reflect the change of the optimal contour line; Step 53: output the number and coordinates of the feature points, and mark the first serial number and coordinate value of the feature points on the optimal contour line to achieve visualization.

[0016] Preferably, the first serial number may be a Roman numeral.

[0017] Preferably, the zero crossing point in step 52 does not include the human face area, so that the workload of post-processing can be reduced without affecting the clothing design.

[0018] Step 6, mark the feature points in the three-dimensional surface model; slice the three-dimensional surface model in parallel with the cross-sectional plane at the Z-axis coordinate value of each feature point to obtain several feature sections of the front and rear parts; calculate the plate combination for filling based on the distance between adjacent feature sections, the type of plate and the thickness of the plate; based on the plate combination, fill the plate layer between the adjacent feature sections of the front and rear parts, and combine the corresponding optimal contour lines to obtain several human body parts; based on the human body parts, output the corresponding processing data; the processing data includes the plate combination, the third serial number and the cutting contour line; the third serial number is used to distinguish the plates, and may include the second serial number and the corresponding plate number; the cutting contour line is the circumferential contour line of the plate.

[0019] In a feasible implementation manner, the step 6 specifically includes: Step 61: Mark the feature points in the three-dimensional surface model by using coordinate values; Step 62, parallel slicing the three-dimensional surface model with the cross-section at the Z-axis coordinate value of each feature point (when the three-dimensional surface model is in a vertical state, the cross-section is a horizontal plane) to obtain a plurality of feature sections of the front and rear parts; Step 63, starting from the cross-section where the three-dimensional origin is located, connecting each characteristic section in sequence; Step 64, calculating a plate combination for filling based on the distance between adjacent characteristic sections, the type of plate and the thickness of the plate; Step 65: Based on the plate combination, fill the plate layer between the adjacent characteristic sections of the front and rear parts, and combine the corresponding optimal contour lines to obtain several human body parts; based on the human body parts, output the corresponding processing data; the processing data includes the plate combination, the third serial number and the cutting contour line; set a positioning mark at the two-dimensional origin of each human body part, and the positioning mark includes a positioning symbol, a second serial number and a thickness mark; the positioning symbol is used to position the plate of the human body part when assembling the mannequin in the subsequent sequence; the second serial number is used to distinguish the human body parts, and is the first serial number of the characteristic point belonging to the characteristic section; the thickness mark is the 1st decimal place of the thickness value of the human body part, and the unit of the thickness value of the human body part is cm.

[0020] In a feasible implementation, the types of boards include foam boards and PVC chaff boards, which not only saves costs, but also has soft texture, is not easy to deform and is elastic, which is conducive to placing bead needles in three-dimensional tailoring of clothing design, and can be reused repeatedly; wherein, the foam board is suitable for large thickness cutting processing, and the PVC chaff board is suitable as a small thickness filling layer; The thickness of the plate is: The thickness of the foam board is 1 cm; The thickness of PVC Chevron sheet is 0.1cm, 0.2cm and 0.5cm; The method for calculating the plate combination specifically includes: When the distance value H between adjacent characteristic cross sections is equal to an integer multiple A of 1 cm, the board combination is A foam boards; When the distance value H between adjacent characteristic sections is greater than an integer multiple A of 1 cm, calculate the difference T1 between the two; then calculate the difference T2 between T1 and 0.5 cm: If T2 is equal to 0, the board combination is A foam boards and 1 0.5 cm thick PVC Chevron board; If T2 is not equal to 0, calculate the difference T3 between T2 and the integer multiple B of 0.2 cm. The specific formula is T3 = T2-0.2*B: If T3 is equal to 0, the board combination is A foam board, 1 0.5 cm thick PVC Chevron board and B 0.2 cm thick PVC Chevron board; If T3 is not equal to 0, and T3 is taken as an integer multiple C of 0.1 cm, the board combination is A foam boards, 1 0.5 cm thick PVC Chevron board, B 0.2 cm thick PVC Chevron boards and C 0.1 cm thick PVC Chevron boards; When the distance H between adjacent characteristic sections is less than 1 cm, the difference T2' between H and 0.5 cm is directly calculated: If T2' is equal to 0, the board combination is 1 0.5 cm thick PVC Shevron board; If T2' is not equal to 0, calculate the difference T3' between T2' and the integer multiple B of 0.2 cm. The specific formula is T3'= T2'-0.2*B: If T3' is equal to 0, the board combination is 1 0.5 cm thick PVC Chevron board and B 0.2 cm thick PVC Chevron boards; If T3' is not equal to 0, and T3' is taken as an integer multiple C of 0.1 cm, the board combination is 1 0.5 cm thick PVC Chevron board, B 0.2 cm thick PVC Chevron boards and C 0.1 cm thick PVC Chevron boards; Through the above method, an optimized combination of commonly used plates can be obtained, which facilitates the subsequent manufacture of layered mannequins with a smaller number of plates and a shorter plate cutting time, thereby improving mannequin manufacturing efficiency and reducing mannequin manufacturing costs.

[0021] Preferably, the positioning symbol includes a half cross structure, so that it is convenient to position and fix the human body part according to the positioning symbol in the subsequent sequence.

[0022] Preferably, the positioning symbol also includes a concave-convex structure on both sides of the semi-cross structure, which is used to combine the front and rear parts of the human body parts into one by inserting the concave-convex structure, thereby avoiding the use of curing methods such as glue bonding in the subsequent assembly of the mannequin, further reducing costs.

[0023] Preferably, the concave-convex structure is a T-shaped structure, that is, a T-shaped protrusion is provided on the plate of the front body part, and a corresponding T-shaped groove is provided on the plate of the rear body part; of course, the reverse arrangement can also be performed.

[0024] Preferably, the T-shaped structure also includes a method of adjusting the length dimension to achieve a fool-proof effect, so as to avoid the subsequent incorrect use of the board between different mannequins.

[0025] Step 7, the server sends the processing data to the plate cutting equipment for manufacturing; the plate cutting equipment is used to cut and manufacture the plates corresponding to the human body parts according to the shape of the optimal contour line, and cut the positioning holes on the plates according to the shape of the positioning symbols; and mark the third serial number on the plate.

[0026] Step 8: Based on the third serial number, assemble the various plates of the human body parts in layers on the positioning tooling through the positioning holes at the positioning symbols to obtain a mannequin for clothing design; mark the corresponding human body sample information on the mannequin for reference in subsequent clothing design.

[0027] In a second aspect, based on the same inventive concept, the present application also provides a manufacturing system using the above-mentioned personalized mannequin manufacturing method, including a client and a server; The client is used to collect three-dimensional point cloud data of a human body sample and send it to a server together with the human body sample information; The server comprises a data receiving module, a data processing module and a result generating module; The data receiving module is used to receive three-dimensional point cloud data and human body sample information; The data processing module includes a sagittal section point cloud data unit, a contour data text unit, a fitting prediction unit and a segmentation unit; The sagittal section point cloud data unit generates a three-dimensional surface model and extracts human body sagittal section point cloud data text from the three-dimensional point cloud data; The contour line data text unit, in the sagittal section point cloud data text, takes the point cloud at the minimum value of the Z-axis coordinate as the three-dimensional origin, takes the vertical axis of the three-dimensional origin as the reference, divides the sagittal section point cloud data into two parts, front and back, and writes the corresponding contour line data texts respectively; The fitting prediction unit stores a BP neural network; based on the contour line data text, the BP neural network is used to fit and predict the optimal contour lines of the front and rear parts respectively; The segmentation unit updates the three-dimensional surface model based on the optimal contour line, calculates the zero crossing point through the first-order derivative, and determines the characteristic points of the front and rear parts; marks the characteristic points in the three-dimensional surface model; slices the three-dimensional surface model in parallel with the cross-sectional plane at the Z-axis coordinate value of each characteristic point to obtain a number of characteristic sections of the front and rear parts; calculates a sheet material combination for filling based on the distance between adjacent characteristic sections, the type of sheet material and the thickness of sheet material; fills the sheet material layer between the adjacent characteristic sections of the front and rear parts based on the sheet material combination, and obtains a number of human body parts in combination with the corresponding optimal contour line; and outputs corresponding processing data based on the human body parts; The result generation module is used to send out the processed data and human sample information.

[0028] On the third aspect, based on the same inventive concept, the present application also provides a device for manufacturing a personalized mannequin, including a processor, a memory and a bus, wherein the memory stores instructions and data readable by the processor, and the processor is used to call the instructions and data in the memory to execute a method for manufacturing a personalized mannequin as described above, and the bus connects the functional components to transmit information.

[0029] In a feasible implementation manner, the manufacturing device further includes a three-dimensional human body scanner for performing three-dimensional scanning on a human body sample.

[0030] In a feasible implementation, the manufacturing device further includes a plate cutting device for cutting and manufacturing plates corresponding to human body parts in accordance with the shape of the optimal contour line, and cutting positioning holes on the plates in accordance with the shape of the positioning symbols.

[0031] In a feasible implementation manner, the manufacturing device further comprises a positioning tool for fixing each plate corresponding to the human body part; the positioning tool comprises a base, a rotating disk and a bracket; The center of the lower surface of the rotating disk is pivotally connected to the center of the base; a slot is arranged at the center of the upper surface of the rotating disk; the bracket is vertically fixed at the slot; and the bracket matches with the positioning hole on the plate.

[0032] Preferably, the positioning holes are of a semi-cross structure (the positioning holes of the front and rear human body component plates can form a complete cross structure correspondingly), and the bracket is of a cross structure, which facilitates horizontal orientation and fixation of the human body component.

[0033] In the fourth aspect, based on the same inventive concept, the present application also provides a personalized mannequin using the above-mentioned personalized mannequin manufacturing method, wherein the personalized mannequin is divided into two parts, front and back, along the coronal plane; each part is composed of multiple layers of plates combined along the vertical axis; and the same positioning hole is set at the two-dimensional origin of each plate.

[0034] In a feasible implementation manner, the personalized mannequin is integrated with a positioning tool, and the positioning tool comprises a base, a rotating disk and a bracket; The center of the lower surface of the rotating disk is pivotally connected to the center of the base; a slot is arranged at the center of the upper surface of the rotating disk; the bracket is vertically fixed at the slot; and the bracket matches with the positioning hole on the plate.

[0035] By adopting the above technical solution, the present invention has the following beneficial effects: The present invention provides a personalized mannequin and a manufacturing method, system and device thereof. Through three-dimensional human body scanning, three-dimensional point cloud data of a human body sample is obtained and a three-dimensional curved surface model is constructed; the three-dimensional curved surface model is segmented front and back and layered up and down to generate processing data containing information such as plate combination; through the processing data, standard plates are cut and assembled into a complete and personalized layered mannequin. Compared with the existing technology, this solution can significantly reduce the manufacturing cost of the mannequin while retaining the customer's personalized body shape information, and can quickly and accurately manufacture a professional layered mannequin that meets the requirements of clothing design. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] In order to more clearly illustrate the specific implementation methods of the present invention or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings in the following description are some implementation methods of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0037] Figure 1 A flow chart of a method for manufacturing a personalized mannequin provided by an embodiment of the present invention; Figure 2 It is an illustration of the human anatomical coordinate system in the prior art; Figure 3 A schematic diagram of a sagittal cross-section point cloud provided by an embodiment of the present invention; Figure 4 The R value diagram of each data set for the BP neural network used for predicting and fitting the contour line of the posterior part of the human body provided in the embodiment of the present invention: wherein, the diagram a represents the training set, the diagram b represents the validation set, the diagram c represents the test set, and the diagram d represents the entire set; Figure 5The R value diagram of each data set for the BP neural network used for predicting and fitting the contour line of the front part of the human body provided in the embodiment of the present invention: wherein, the diagram a represents the training set, the diagram b represents the validation set, the diagram c represents the test set, and the diagram d represents the entire set; Figure 6 The embodiment of the present invention provides Figure 3 An example of a top view of a human body part at the front part at feature point a; Figure 7 A system diagram for manufacturing a personalized mannequin provided by an embodiment of the present invention; Figure 8 A cross-sectional view of a personalized mannequin positioning tool provided in an embodiment of the present invention; Fig. 9 A three-dimensional diagram of the personalized mannequin positioning tooling provided in an embodiment of the present invention.

[0038] Reference numerals: 1-base; 2-rotating plate; 3-bracket. DETAILED DESCRIPTION

[0039] The technical solution of the present invention will be described clearly and completely below in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0040] In the description of the present invention, it should be noted that the terms "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inner", "outer", etc., indicating the orientation or positional relationship, are based on the orientation or positional relationship shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, the terms "first", "second", and "third" are used for descriptive purposes only, and cannot be understood as indicating or implying relative importance.

[0041] In the description of the present invention, it should be noted that, 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 ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0042] The present invention is further explained below in conjunction with specific implementation modes.

[0043] It should also be noted that the following specific embodiments or specific implementations are a series of optimized settings listed in the present invention to further explain the specific content of the invention, and these settings can be used in combination or in association with each other.

[0044] Embodiment 1: like Figure 1 As shown, the present embodiment provides a method for manufacturing a personalized mannequin, comprising the following steps: Step 1: The client collects the 3D point cloud data of the human body sample through a 3D human body scanner and sends it to the server together with the human body sample information; the 3D point cloud data establishes a coordinate system based on the axes in human anatomy, such as Figure 2 As shown: the Y axis is the sagittal axis (the axis from front to back of a standing human body and perpendicular to the vertical axis), the X axis is the coronal axis (the axis from left to right of a standing human body and perpendicular to both the sagittal axis and the vertical axis), and the Z axis is the vertical axis (the axis from head to tail of a standing human body and perpendicular to the ground plane); the human sample information includes customer ID, name, age, weight, height, skin color, and collection date, etc.; Furthermore, the step 1 specifically includes: Step 11, prepare human body samples according to the requirements of 3D point cloud collection; the requirements of 3D point cloud collection include: wearing a white tights, a white swimming cap, and removing accessories; standing with feet 10-20cm apart, head held high, chest held high, parallel, and eyes looking forward; forearms hanging naturally, upper arms abducted at 20 degrees to the body, and fists clenched; breathing normally, etc.; Step 12: Perform a three-dimensional scan on the human body sample by using a three-dimensional human body scanner to obtain three-dimensional point cloud data; the three-dimensional human body scanner may use existing equipment; Step 13: After encrypting and compressing the 3D point cloud data, the client sends it to the server.

[0045] Through the above steps, three-dimensional point cloud data of customer body samples can be collected by three-dimensional body scanners in physical stores, experience stores, etc., and sent to the server through the client, which is conducive to building a data foundation for personalized mannequin manufacturing; and fully protecting the privacy of customer body shape information.

[0046] Step 2: The server generates a three-dimensional surface model from the three-dimensional point cloud data and extracts the human body sagittal section point cloud data text.

[0047] Furthermore, the step 2 specifically includes: Step 21, the server decompresses and decrypts the three-dimensional point cloud data to generate a three-dimensional surface model of the human body sample; Step 22: Figure 3As shown, all point cloud data with an X-axis coordinate of 0 in the three-dimensional point cloud data, that is, the point cloud data in the reference plane where the left-right center line of the human body sample is located, are extracted as the sagittal section point cloud data text, which can be named Standard_Section_0.txt.

[0048] Step 3. In the sagittal section point cloud data text, take the point cloud at the minimum Z-axis coordinate as the three-dimensional origin, and take the vertical axis of the three-dimensional origin as the reference to divide the sagittal section point cloud data into two parts, namely, the two parts with Y-axis coordinate y≥0 and the two parts with Y-axis coordinate y<0, and write the corresponding contour line data texts respectively.

[0049] Furthermore, the step 3 specifically includes: Step 31: The server loads the sagittal cross-section point cloud data text, extracts the second and third columns of data in the point cloud coordinates, and defines them as the coordinate values ​​of the Z axis and the Y axis respectively; Step 32: Count the minimum value min_Z of the Z-axis coordinate value z. The specific formula can be: min_Z = min(z); Step 33: Subtract min_Z from each z to obtain the updated coordinate value z', which is used to determine the position of the three-dimensional origin. The specific formula can be: z' = z - min_Z; Step 34: Based on the vertical axis of the three-dimensional origin, the sagittal cross-section point cloud data is divided into two parts, front and back, through conditional indexing, and the corresponding contour line data texts are written respectively: The conditional indexes in the first part include z>0 and y≥0; The conditional indexing of the latter part includes z>0 and y<0; In this way, the sagittal cross-section point cloud data can be divided into two parts, front and back, so that the mannequin can be manufactured separately in the subsequent sequence; of course, the encoding and direction of the coordinate axis can also be changed according to actual needs to facilitate subsequent processing.

[0050] Step 4: Based on the contour line data text, the BP neural network is used to fit and predict the optimal contour lines of the front and back parts.

[0051] Furthermore, the training process of the BP neural network includes: Step 41, the server loads the contour data text, and after standardization, randomly divides it into a training set, a validation set and a test set, with a specific division ratio of 14:3:3; Step 42: Create a BP neural network: The number of hidden layers (feedforwardnet) is set to 10; The training cycle (trainParam.epochs) is set to 1000; The training target error (trainParam.goal) is set to 1e-5; Step 43, train the BP neural network until the correlation coefficient R values ​​of the training set, validation set, test set and all sets are greater than or equal to 0.9, so that the BP neural network has reliable fitting prediction capabilities: like Figure 4 As shown, Figure 4 Figure a in the middle shows that the R value of the training set of the back contour data text in the BP neural network is 0.99967; Figure 4 Figure b in the middle shows that the R value of the validation set of the back contour data text in the BP neural network is 0.9995; Figure 4 Figure c in the middle shows that the R value of the training set of the back contour data text in the BP neural network is 0.99973; Figure 4 The middle d figure shows that the R value of the training set of the back part of the contour data text in the BP neural network is 0.99965; like Figure 5 As shown, Figure 5 Figure a in the middle shows that the R value of the training set of the front part of the contour data text in the BP neural network is 0.99334; Figure 5 Figure b in the middle shows that the R value of the validation set of the front part of the contour data text in the BP neural network is 0.99557; Figure 5 Figure c in the middle shows that the R value of the training set of the front part of the contour data text in the BP neural network is 0.99899; Figure 5 The figure d in the middle shows that the R value of the training set of the front part of the contour data text in the BP neural network is 0.99422; In summary, the BP neural network meets the requirements.

[0052] Step 5: Based on the optimal contour line, update the three-dimensional surface model, calculate the zero crossing point through the first-order derivative, and determine the feature points of the front and back parts.

[0053] Furthermore, the specific steps of determining the feature points in step 5 include: Step 51, calculating the first-order derivative of the optimal contour line; Step 52: taking the zero crossing point of the first-order derivative as a feature point to reflect the change of the optimal contour line; Step 53: output the number and coordinates of the feature points, and mark the first serial number and coordinate value of the feature points on the optimal contour line to achieve visualization.

[0054] Preferably, the first serial number may be a Roman numeral.

[0055] Preferably, the zero crossing point in step 52 does not include the human face area, so that the workload of post-processing can be reduced without affecting the clothing design.

[0056] Step 6, mark the feature points in the three-dimensional surface model; slice the three-dimensional surface model in parallel with the cross-sectional plane at the Z-axis coordinate value of each feature point to obtain several feature sections of the front and rear parts; calculate the plate combination for filling based on the distance between adjacent feature sections, the type of plate and the thickness of the plate; based on the plate combination, fill the plate layer between the adjacent feature sections of the front and rear parts, and combine the corresponding optimal contour lines to obtain several human body parts; based on the human body parts, output the corresponding processing data; the processing data includes the plate combination, the third serial number and the cutting contour line; the third serial number is used to distinguish the plates, and may include the second serial number and the corresponding plate number; the cutting contour line is the circumferential contour line of the plate.

[0057] Furthermore, the step 6 specifically includes: Step 61: Mark the feature points in the three-dimensional surface model by using coordinate values; Step 62, parallel slicing the three-dimensional surface model with the cross-section at the Z-axis coordinate value of each feature point (when the three-dimensional surface model is in a vertical state, the cross-section is a horizontal plane) to obtain a plurality of feature sections of the front and rear parts; Step 63, starting from the cross-section where the three-dimensional origin is located, connecting each characteristic section in sequence; Step 64, calculating a plate combination for filling based on the distance between adjacent characteristic sections, the type of plate and the thickness of the plate; Step 65: Based on the sheet material combination, fill the sheet material layer between the adjacent characteristic sections of the front and rear parts, and combine the corresponding optimal contour lines to obtain a number of human body parts; based on the human body parts, output the corresponding processing data; the processing data includes the sheet material combination, the third serial number and the cutting contour line; set a positioning mark at the two-dimensional origin of each human body part, and the positioning mark includes a positioning symbol, a second serial number and a thickness mark; the positioning symbol is used to position the sheet material of the human body part when assembling the mannequin in the subsequent sequence; the second serial number is used to distinguish the human body parts, and is the first serial number of the characteristic point belonging to the characteristic section; the thickness mark is the decimal place of the thickness value of the human body part, and the unit of the thickness value of the human body part is cm; The types of boards include foam boards and PVC chevron boards, which not only save costs, but also have soft textures, are not easy to deform and are elastic, which is conducive to placing bead needles in three-dimensional tailoring of clothing design, and can be reused repeatedly; among them, foam boards are suitable for large-thickness cutting processing, and PVC chevron boards are suitable as small-thickness filling layers; The thickness of the plate is: The thickness of the foam board is 1 cm; The thickness of PVC Chevron sheet is 0.1cm, 0.2cm and 0.5cm; The method for calculating the plate combination specifically includes: When the distance value H between adjacent characteristic cross sections is equal to an integer multiple A of 1 cm, the board combination is A foam boards; When the distance value H between adjacent characteristic sections is greater than an integer multiple A of 1 cm, calculate the difference T1 between the two; then calculate the difference T2 between T1 and 0.5 cm: If T2 is equal to 0, the board combination is A foam boards and 1 0.5 cm thick PVC Chevron board; If T2 is not equal to 0, calculate the difference T3 between T2 and the integer multiple B of 0.2 cm. The specific formula is T3 = T2-0.2*B: If T3 is equal to 0, the board combination is A foam board, 1 0.5 cm thick PVC Chevron board and B 0.2 cm thick PVC Chevron board; If T3 is not equal to 0, and T3 is taken as an integer multiple C of 0.1 cm, the board combination is A foam boards, 1 0.5 cm thick PVC Chevron board, B 0.2 cm thick PVC Chevron boards and C 0.1 cm thick PVC Chevron boards; When the distance H between adjacent characteristic sections is less than 1 cm, the difference T2' between H and 0.5 cm is directly calculated: If T2' is equal to 0, the board combination is 1 0.5 cm thick PVC Shevron board; If T2' is not equal to 0, calculate the difference T3' between T2' and the integer multiple B of 0.2 cm. The specific formula is T3'= T2'-0.2*B: If T3' is equal to 0, the board combination is 1 0.5 cm thick PVC Chevron board and B 0.2 cm thick PVC Chevron boards; If T3' is not equal to 0, and T3' is taken as an integer multiple C of 0.1 cm, the board combination is 1 0.5 cm thick PVC Chevron board, B 0.2 cm thick PVC Chevron boards and C 0.1 cm thick PVC Chevron boards; Through the above method, an optimized plate combination can be obtained, so that it is convenient to manufacture the mannequin with a smaller number of plates and a shorter plate cutting time in the subsequent process, thereby improving the mannequin manufacturing efficiency and reducing the mannequin manufacturing cost.

[0058] Step 7, the server sends the processing data to the plate cutting equipment for manufacturing; the plate cutting equipment is used to cut and manufacture the plates corresponding to the human body parts according to the shape of the optimal contour line, and cut the positioning holes on the plates according to the shape of the positioning symbols; and mark the third serial number on the plate.

[0059] Step 8: Based on the third serial number, assemble the various plates of the human body parts in layers on the positioning tooling through the positioning holes at the positioning symbols to obtain a mannequin for clothing design; mark the corresponding human body sample information on the mannequin for reference in subsequent clothing design.

[0060] Embodiment 2: like Figure 7 As shown, this embodiment provides a manufacturing system using the above-mentioned personalized mannequin manufacturing method, including a client and a server; The client is used to collect three-dimensional point cloud data of a human body sample and send it to a server together with the human body sample information; The server comprises a data receiving module, a data processing module and a result generating module; The data receiving module is used to receive three-dimensional point cloud data and human body sample information; The data processing module includes a sagittal section point cloud data unit, a contour data text unit, a fitting prediction unit and a segmentation unit; The sagittal section point cloud data unit generates a three-dimensional surface model and extracts human body sagittal section point cloud data text from the three-dimensional point cloud data; The contour line data text unit, in the sagittal section point cloud data text, takes the point cloud at the minimum value of the Z-axis coordinate as the three-dimensional origin, takes the vertical axis of the three-dimensional origin as the reference, divides the sagittal section point cloud data into two parts, front and back, and writes the corresponding contour line data texts respectively; The fitting prediction unit stores a BP neural network; based on the contour line data text, the BP neural network is used to fit and predict the optimal contour lines of the front and rear parts respectively; The segmentation unit updates the three-dimensional surface model based on the optimal contour line, calculates the zero crossing point through the first-order derivative, and determines the characteristic points of the front and rear parts; marks the characteristic points in the three-dimensional surface model; slices the three-dimensional surface model in parallel with the cross-sectional plane at the Z-axis coordinate value of each characteristic point to obtain a number of characteristic sections of the front and rear parts; calculates a sheet material combination for filling based on the distance between adjacent characteristic sections, the type of sheet material and the thickness of sheet material; fills the sheet material layer between the adjacent characteristic sections of the front and rear parts based on the sheet material combination, and obtains a number of human body parts in combination with the corresponding optimal contour line; and outputs corresponding processing data based on the human body parts; The result generation module is used to send out the processed data and human sample information.

[0061] Embodiment three: The present embodiment provides a device for manufacturing a personalized mannequin, including a processor, a memory and a bus. The memory stores instructions and data that can be read by the processor. The processor is used to call the instructions and data in the memory to execute a method for manufacturing a personalized mannequin as described above. The bus connects the functional components to transmit information.

[0062] Furthermore, the manufacturing device also includes a three-dimensional human body scanner for performing three-dimensional scanning on the human body sample.

[0063] Furthermore, the manufacturing device also includes a plate cutting device, which can use existing CNC equipment to imitate the shape of the optimal contour line to cut and manufacture various plates corresponding to the human body parts, and imitate the shape of the positioning symbol to cut the positioning holes on the plate.

[0064] Furthermore, the manufacturing device also includes a positioning tool for fixing each plate corresponding to the human body part, such as Figure 8-9 As shown; the positioning tool comprises a base 1, a rotating disk 2 and a bracket 3; The center of the lower surface of the rotating disk 2 is pivotally connected to the center of the base 1; a slot is provided at the center of the upper surface of the rotating disk 2; the bracket 3 is vertically fixed at the slot; and the bracket 3 matches the positioning hole on the plate.

[0065] The positioning holes are of a semi-cross structure (the positioning holes of the front and rear human body component plates can form a complete cross structure correspondingly), and the bracket 3 is of a cross structure, which facilitates horizontal orientation and fixation of the human body component; based on Figure 3 The characteristic points a, b, c, d, e, f, g, h, i, j, k and the origin o vertex s are used to define the corresponding human body parts. The specific plate combination refers to the calculation method in Example 1. The results are as follows: Human body part a (about 9.6 cm) includes: 9 foam boards, 1 0.5 cm thick PVC Chevron board and 1 0.1 cm thick PVC Chevron board; Human body part b (about 7.2 cm) includes: 7 foam boards, 1 0.2 cm thick PVC Chevron board; Human body part c (about 7.5 cm) includes: 7 foam boards, 1 0.5 cm thick PVC Chevron board; Human body part d (about 5.3 cm) includes: 5 foam boards, 1 0.2 cm thick PVC Chevron board and 1 0.1 cm thick PVC Chevron board; Human body part e (about 8.1 cm) includes: 8 foam boards and 1 0.1 cm thick PVC Chevron board; Human body part f (about 31.3 cm) includes: 31 foam boards, 1 0.2 cm thick PVC Chevron board and 1 0.1 cm thick PVC Chevron board; The human body parts (about 30 cm) include: 30 foam boards (the top foam board has no positioning holes and is used for capping); Human body part g (about 13.3 cm) includes: 13 foam boards, 1 0.2 cm thick PVC Chevron board and 1 0.1 cm thick PVC Chevron board; The human body part h (about 18.6 cm) includes: 18 foam boards, 1 0.5 cm thick PVC Chevron board and 1 0.1 cm thick PVC Chevron board; Human body part I (about 23.5cm) includes: 23 foam boards and 1 0.5cm thick PVC Chevron board; Human body part j (about 23.5cm) includes: 23 foam boards and 1 0.5cm thick PVC Chevron board; Human body part k (about 10.6 cm) includes: 10 foam boards, 1 0.5 cm thick PVC Chevron board and 1 0.1 cm thick PVC Chevron board; Human body part-s (about 9.5cm) includes: 9 foam boards and 1 0.5cm thick PVC Chevron board (without positioning holes, used for capping).

[0066] In another embodiment, the present solution can be implemented by an integrated device, which may include corresponding modules for performing each or several steps in the above-mentioned embodiments. The module may be one or more hardware modules specially configured to perform the corresponding steps, or implemented by a processor configured to perform the corresponding steps, or stored in a computer-readable medium for implementation by a processor, or implemented by some combination.

[0067] The processor performs the various methods and processes described above. For example, the method implementation in the present solution can be implemented as a software program, which is tangibly contained in a machine-readable medium, such as a memory. In some embodiments, part or all of the software program can be loaded and / or installed via a memory and / or a communication interface. When the software program is loaded into the memory and executed by the processor, one or more steps in the method described above can be executed. Alternatively, in other embodiments, the processor can be configured to perform one of the above methods by any other appropriate means (for example, by means of firmware).

[0068] The device can be implemented using a bus architecture. The bus architecture can include any number of interconnecting buses and bridges, depending on the specific application and overall design constraints of the hardware. The bus connects various circuits including one or more processors, memories, and / or hardware modules together. The bus can also connect various other circuits such as peripherals, voltage regulators, power management circuits, external antennas, etc.

[0069] The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Component (EISA) bus, etc. The bus can be divided into an address bus, a data bus, a control bus, etc.

[0070] Embodiment 4: The present embodiment provides a personalized mannequin using the aforementioned personalized mannequin manufacturing method, wherein the personalized mannequin is divided into two parts, front and back, along the coronal plane; each part is composed of multiple layers of plate materials combined along a vertical axis; and the same positioning hole is provided at the two-dimensional origin of each plate material.

[0071] Furthermore, the personalized mannequin is integrated with a positioning tool, and the positioning tool comprises a base, a rotating disk and a bracket; The center of the lower surface of the rotating disk is pivotally connected to the center of the base; a slot is arranged at the center of the upper surface of the rotating disk; the bracket is vertically fixed at the slot; and the bracket matches with the positioning hole on the plate.

[0072] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for manufacturing a personalized mannequin, characterized in that: include: Step 1: The client collects the 3D point cloud data of the human body sample through a 3D human body scanner, and sends it to the server together with the human body sample information; The three-dimensional point cloud data establishes a coordinate system based on the axes in human anatomy: the Y axis is the sagittal axis, the X axis is the coronal axis, and the Z axis is the vertical axis; Step 2: The server generates a three-dimensional surface model from the three-dimensional point cloud data and extracts the human body sagittal section point cloud data text; Step 3: In the sagittal section point cloud data text, the point cloud at the minimum value of the Z-axis coordinate is used as the three-dimensional origin, and the vertical axis of the three-dimensional origin is used as the reference to divide the sagittal section point cloud data into two parts, front and back, and write the corresponding contour line data texts respectively; Step 4: Based on the contour line data text, the BP neural network is used to fit and predict the optimal contour lines of the front and back parts respectively; Step 5: Based on the optimal contour line, the three-dimensional surface model is updated, and the zero crossing point is calculated by the first-order derivative to determine the feature points of the front and rear parts; Step 6, marking the characteristic points in the three-dimensional surface model; slicing the three-dimensional surface model in parallel with the cross-section at the Z-axis coordinate value of each characteristic point to obtain a plurality of characteristic sections of the front and rear parts; Based on the distance between adjacent characteristic sections, the type of sheet and the thickness of the sheet, a sheet combination for filling is calculated; based on the sheet combination, a sheet layer is filled between adjacent characteristic sections of the front and rear parts, and a number of human body parts are obtained in combination with corresponding optimal contour lines; based on the human body parts, corresponding processing data is output; the processing data includes the sheet combination, the third sequence number and the cutting contour line; Step 7, the server sends the processing data to the plate cutting device for manufacturing; the plate cutting device is used to cut and manufacture the plates corresponding to the human body parts according to the shape of the optimal contour line, and cut the positioning holes on the plates according to the shape of the positioning symbols; and mark the third serial number on the plate; Step 8: Based on the third serial number, assemble the various plates of the human body parts in layers on the positioning tooling through the positioning holes at the positioning symbols to obtain a mannequin for clothing design; and mark the corresponding human body sample information on the mannequin.

2. The method according to claim 1, characterized in that The step 1 specifically includes: Step 11, prepare human body samples according to the requirements of 3D point cloud collection; the requirements of 3D point cloud collection include: wearing a white tights, a white swimming cap, and removing accessories; standing with feet apart, head held high, chest held high, parallel, and eyes looking forward; forearms hanging naturally, upper arms abducted, and hands clenched into fists; keep breathing normally; Step 12: Perform a three-dimensional scan on the human body sample using a three-dimensional human body scanner to obtain three-dimensional point cloud data; Step 13: After encrypting and compressing the 3D point cloud data, the client sends it to the server.

3. The method according to claim 2, characterized in that The step 2 specifically includes: Step 21, the server decompresses and decrypts the three-dimensional point cloud data to generate a three-dimensional surface model of the human body sample; Step 22: extract all point cloud data with an X-axis coordinate of 0 in the three-dimensional point cloud data as sagittal section point cloud data text.

4. The method according to claim 1, characterized in that: The step 3 specifically includes: Step 31: The server loads the sagittal cross-section point cloud data text, extracts the second and third columns of data in the point cloud coordinates, and defines them as Z-axis coordinate value z and Y-axis coordinate value y, respectively; Step 32: Calculate the minimum value min_Z of the Z-axis coordinate value z. The specific formula is: min_Z = min(z); Step 33: Subtract min_Z from each z to obtain the updated coordinate value z', which is used to determine the position of the three-dimensional origin. The specific formula is: z' = z - min_Z; Step 34: Based on the vertical axis of the three-dimensional origin, the sagittal section point cloud data is divided into two parts, front and back, through conditional indexing, and the corresponding contour line data texts are written respectively: The conditional indexes in the first part include z>0 and y≥0; The conditional indexing of the latter part includes z>0 and y<0.

5. The method according to claim 1, characterized in that: The training process of the BP neural network includes: Step 41: The server loads the contour data text, performs standardization, and randomly divides it into a training set, a validation set, and a test set; Step 42: Create a BP neural network: The number of hidden layers is set to 10; The training cycle is set to 1000; The training target error is set to 1e-5; Step 43: Train the BP neural network until the correlation coefficient R values ​​of the training set, validation set, test set and all sets are greater than or equal to 0.

9.

6. The method according to claim 1, characterized in that The specific steps of determining the feature points in step 5 include: Step 51, calculating the first-order derivative of the optimal contour line; Step 52: taking the zero crossing point of the first-order derivative as a feature point to reflect the change of the optimal contour line; Step 53: Output the number and coordinates of the feature points, and mark the first serial number and coordinate value of the feature points on the optimal contour line.

7. The method according to claim 1, characterized in that The step 6 specifically includes: Step 61: Mark the feature points in the three-dimensional surface model by using coordinate values; Step 62, parallel slicing the three-dimensional surface model with the cross-section at the Z-axis coordinate value of each characteristic point to obtain a plurality of characteristic sections of the front and rear parts; Step 63, starting from the cross-section where the three-dimensional origin is located, connecting each characteristic section in sequence; Step 64, calculating a plate combination for filling based on the distance between adjacent characteristic sections, the type of plate and the thickness of the plate; Step 65: Based on the plate combination, fill the plate layer between the adjacent characteristic sections of the front and rear parts, and combine the corresponding optimal contour lines to obtain several human body parts; based on the human body parts, output the corresponding processing data; the processing data includes the plate combination, the third serial number and the cutting contour line; set a positioning mark at the two-dimensional origin of each human body part, and the positioning mark includes a positioning symbol, a second serial number and a thickness mark; the positioning symbol is used to position the plate of the human body part when assembling the mannequin in the subsequent sequence; the second serial number is used to distinguish the human body parts, and is the first serial number of the characteristic point belonging to the characteristic section; the thickness mark is the 1st decimal place of the thickness value of the human body part, and the unit of the thickness value of the human body part is cm.

8. The method according to claim 7, characterized in that The types of boards include foam boards and PVC chevron boards; The thickness of the plate is: The thickness of the foam board is 1 cm; The thickness of PVC Chevron sheet is 0.1cm, 0.2cm and 0.5cm; The method for calculating the plate combination specifically includes: When the distance value H between adjacent characteristic cross sections is equal to an integer multiple A of 1 cm, the board combination is A foam boards; When the distance value H between adjacent characteristic sections is greater than an integer multiple A of 1 cm, calculate the difference T1 between the two; then calculate the difference T2 between T1 and 0.5 cm: If T2 is equal to 0, the board combination is A foam boards and 1 0.5 cm thick PVC Chevron board; If T2 is not equal to 0, calculate the difference T3 between T2 and the integer multiple B of 0.2 cm. The specific formula is T3 = T2-0.2*B: If T3 is equal to 0, the board combination is A foam board, 1 0.5 cm thick PVC Chevron board and B 0.2 cm thick PVC Chevron board; If T3 is not equal to 0, and T3 is taken as an integer multiple C of 0.1 cm, the board combination is A foam boards, 1 0.5 cm thick PVC Chevron board, B 0.2 cm thick PVC Chevron boards and C 0.1 cm thick PVC Chevron boards; When the distance H between adjacent characteristic sections is less than 1 cm, the difference T2' between H and 0.5 cm is directly calculated: If T2' is equal to 0, the board combination is 1 0.5 cm thick PVC Shevron board; If T2' is not equal to 0, calculate the difference T3' between T2' and the integer multiple B of 0.2 cm. The specific formula is T3'= T2'-0.2*B: If T3' is equal to 0, the board combination is 1 0.5 cm thick PVC Chevron board and B 0.2 cm thick PVC Chevron boards; If T3' is not equal to 0, and T3' is taken as an integer multiple C of 0.1 cm, the board combination is 1 0.5 cm thick PVC Chevron board, B 0.2 cm thick PVC Chevron boards and C 0.1 cm thick PVC Chevron boards.

9. A system for manufacturing a personalized mannequin using the method as claimed in any one of claims 1 to 8, characterized in that: Including client and server; The client is used to collect three-dimensional point cloud data of a human body sample and send it to a server together with the human body sample information; The server comprises a data receiving module, a data processing module and a result generating module; The data receiving module is used to receive three-dimensional point cloud data and human body sample information; The data processing module includes a sagittal section point cloud data unit, a contour data text unit, a fitting prediction unit and a segmentation unit; The sagittal section point cloud data unit generates a three-dimensional surface model and extracts human body sagittal section point cloud data text from the three-dimensional point cloud data; The contour line data text unit, in the sagittal section point cloud data text, takes the point cloud at the minimum value of the Z-axis coordinate as the three-dimensional origin, takes the vertical axis of the three-dimensional origin as the reference, divides the sagittal section point cloud data into two parts, front and back, and writes the corresponding contour line data texts respectively; The fitting prediction unit stores a BP neural network; based on the contour line data text, the BP neural network is used to fit and predict the optimal contour lines of the front and rear parts respectively; The segmentation unit updates the three-dimensional surface model based on the optimal contour line, calculates the zero crossing point through the first-order derivative, and determines the characteristic points of the front and rear parts; marks the characteristic points in the three-dimensional surface model; and slices the three-dimensional surface model in parallel with the cross-sectional plane at the Z-axis coordinate value of each characteristic point to obtain a plurality of characteristic sections of the front and rear parts; Based on the distance between adjacent characteristic sections, the type of sheet material and the thickness of the sheet material, a sheet material combination for filling is calculated; based on the sheet material combination, a sheet material layer is filled between adjacent characteristic sections of the front and rear parts, and a number of human body parts are obtained in combination with corresponding optimal contour lines; based on the human body parts, corresponding processing data is output; The result generation module is used to send out the processed data and human sample information.

10. A device for manufacturing a personalized mannequin, characterized in that: It includes a processor, a memory and a bus, wherein the memory stores instructions and data read by the processor, the processor is used to call the instructions and data in the memory to execute any method as claimed in claim 1-8, and the bus connects the functional components for transmitting information.

Images