CN212345491U - Fitting equipment based on deep learning - Google Patents
Fitting equipment based on deep learning Download PDFInfo
- Publication number
- CN212345491U CN212345491U CN202020076253.8U CN202020076253U CN212345491U CN 212345491 U CN212345491 U CN 212345491U CN 202020076253 U CN202020076253 U CN 202020076253U CN 212345491 U CN212345491 U CN 212345491U
- Authority
- CN
- China
- Prior art keywords
- human body
- fitting
- garment
- deep learning
- yarn sensor
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Abstract
The utility model provides a fitting equipment based on degree of depth study relates to intelligence wearing equipment and virtual fitting system field. The fitting device comprises a close-fitting garment body; the flexible yarn sensor comprises a weft yarn sensor and a warp yarn sensor and is used for measuring the girth information and the motion information of the human body; the signal processor is used for converting the signal output by the flexible yarn sensor into a resistance value; the deep learning module is used for acquiring predicted human body parameters through the mapping relation between the resistance values stored in the deep learning module and the actual human body parameters; the signal display is used for constructing and displaying a human body three-dimensional model based on the predicted human body parameters. The fitting degree of the fitting equipment and the human body is high, the structure is simple, the sensitivity is high, and a foundation can be provided for subsequent intelligent fitting.
Description
Technical Field
The utility model relates to an intelligence wearing equipment field, and in particular to fitting equipment based on degree of depth study.
Background
At present, the human body circumference information required to be tested by the intelligent virtual fitting system is mainly obtained according to the soft size, the obtained body circumference data only can reflect the human body circumference information at the measuring moment, and the virtual fitting system usually needs to scan a real human body by means of an optical method so as to reconstruct a three-dimensional model of the human body. The method for obtaining the human body circumference data and reconstructing the three-dimensional model enables consumers to have the worry of uncomfortable feeling and personal information safety hidden danger. Other methods exist, such as CN 110135078A, which can make the user answer a series of set general stature describing questions to estimate the circumference value of the user's body. However, the subjective factors introduced in the method are too large, and a large error exists between the obtained circumference value and the circumference value of the real human body. And the human body circumference information obtained by the method is measured as the static circumference.
Therefore, if the intelligent clothes capable of detecting the human body form information and the motion information in real time can be invented, more comprehensive guidance can be provided for the intelligent fitting system.
SUMMERY OF THE UTILITY MODEL
An object of the utility model is to provide a fitting equipment based on degree of depth study, this fitting equipment is through utilizing the girth of the human each position of flexible yarn sensor real-time measurement, the arm length, and can take notes the motion custom and the dynamic form information of wearer, and its accessible degree of depth study module establishes the resistance value that flexible yarn sensing was surveyed and the mapping relation between the actual human parameter, can obtain fitting personnel's prediction human parameter fast, be used for human three-dimensional model's the establishment with this prediction human body, the method is quick effective, and can compensate in the clothing height can not consider its motion custom to adjust the not enough of clothing volume of releasing.
The technical problem of the utility model is solved and following technical scheme is adopted to realize.
The utility model provides a fitting equipment based on degree of depth study, include:
the close-fitting garment body comprises a garment front side and a garment back side, which respectively correspond to one side of the chest cavity and one side of the back of a human body;
the flexible yarn sensor comprises a weft yarn sensor and a warp yarn sensor; the weft yarn sensors are fixed on the parts of the garment body corresponding to the chest, waist, neck, arms and wrists of a human body, are arranged along the circumferential direction of the human body and are used for measuring the circumference information of the human body; the warp yarn sensors are fixed on the garment body corresponding to the arms, shoulders and elbows of the human body, are arranged in the axial direction of the human body and are used for measuring the motion information of the human body; the weft yarn sensor and the warp yarn sensor both comprise an elastic inner core, conductive yarns spirally wound on the surface of the elastic inner core according to an S shape and electrode wires fixedly connected to two ends of the conductive yarns;
the signal processor is fixed on the garment body, comprises a signal acquisition module, a signal processing module, a wireless signal transmission module and a power supply module which are sequentially connected, and is used for converting the signals output by the flexible yarn sensor into resistance values; the signal acquisition module is connected with the flexible yarn sensor through an electrode wire;
the deep learning module is connected with the wireless signal transmission module and used for acquiring predicted human body parameters through the mapping relation between the resistance value stored in the deep learning module and the actual human body parameters;
and the signal display is connected with the deep learning module and used for constructing and displaying a human body three-dimensional model based on the predicted human body parameters.
As a further optimization, the diameter of the flexible yarn sensor is 0.1 mm-20 mm.
As further optimization, the electrode wire is a metal conductive wire or is formed by a base yarn and a conductive material layer wrapped on the surface of the base yarn; the conductive material layer is made of one selected from metal nanoparticles, metal nanowires, carbon materials or conductive polymer materials, and the carbon materials are one selected from amorphous carbon, graphite, carbon nanotubes and graphene; the conductive polymer material is selected from one of polyaniline, polythiophene, polypyrrole, polyaniline derivatives, polythiophene derivatives and polypyrrole derivatives.
As further optimization, the weft yarn sensors fixed at the parts corresponding to the chest, waist, arms and wrists of the human body are positioned on the front side of the garment; the weft yarn sensor fixed on the part corresponding to the neck of the human body is positioned on the back of the garment.
As further optimization, the warp yarn sensors fixed on the corresponding arms and shoulders of the human body are positioned on the front side of the garment; the warp yarn sensors fixed at the corresponding elbow parts of the human body are positioned on the back of the garment.
As further optimization, still include the holder, the holder includes the clamping piece and sets up the peripheral viscose layer of clamping piece, the viscose layer with the clothing body bonds in order to fix signal processor between the clamping piece with the clothing body.
The utility model provides a fitting equipment based on degree of depth study and fitting method's beneficial effect is:
compare with traditional virtual fitting system, the utility model provides a fitting equipment based on degree of depth study and fitting method thereof need not the fitting personnel and obtains the information that individual prepared through optical scanning, has reduced fitting personnel discomfort and the sense of tension at the measurand in-process. The utility model discloses in adopt the clothing that disposes flexible yarn resistance-type sensor as human waistline, chest circumference, shoulder circumference, arm circumference, elbow circumference, the long measuring tool of sleeve, adopt the method of degree of depth study to establish the corresponding relation between human body circumference and each position yarn sensor resistance, the person of awaiting measuring only needs to be on the person the utility model provides a resistance value of the just human each health position of clothing of fitting equipment, the resistance value in the rethread degree of depth study module and actual human parameter's mapping relation find the prediction human parameter to according to predicting human parameter and establishing human three-dimensional model. The method can obtain the static circumference information of the human body, can record the dynamic circumference information and the motion information of the human body in real time, and is simple, easy and efficient in test method.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings that are required to be used in the embodiments will be briefly described below, it should be understood that the following drawings only illustrate some embodiments of the present invention, and therefore should not be considered as limiting the scope, and for those skilled in the art, other related drawings can be obtained according to the drawings without inventive efforts.
Fig. 1 is a schematic view of a front structure of a close-fitting garment body of a fitting device based on deep learning according to a first embodiment of the present invention;
fig. 2 is a schematic view of a back side structure of a close-fitting garment body of a fitting device based on deep learning according to a first embodiment of the present invention;
fig. 3 is a block diagram of a fitting apparatus based on deep learning according to a first embodiment of the present invention;
fig. 4 is a front view and a side view of a three-dimensional model of a human body reconstructed after an eight-bit consumer wears the fitting apparatus based on deep learning according to the first embodiment of the present invention.
Icon: 1-a garment body; 11-garment front; 12-back of garment; 2-a flexible yarn sensor; 21-weft yarn sensor; 22-warp yarn sensor; 3-a signal processor; 31-a signal acquisition module; 32-a signal processing module; 321-a signal amplification unit; 322-a filtering unit; 323-a signal analysis unit; 33-a wireless signal transmission module; 4-a deep learning module; 5-signal display.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below. The examples, in which specific conditions are not specified, were conducted under conventional conditions or conditions recommended by the manufacturer. The reagents or instruments used are not indicated by the manufacturer, and are all conventional products available commercially.
The features and properties of the present invention are described in further detail below with reference to examples.
As shown in fig. 1-2, a first embodiment of the present invention provides a fitting device based on deep learning, which includes a close-fitting garment body 1, a flexible yarn sensor 2, a signal processor 3, a deep learning module 4 and a signal display 5.
The close-fitting garment body 1 comprises a garment front 11 and a garment back 12, which respectively correspond to one side of the chest cavity and one side of the back of a human body. The garment body 1 can be a close-fitting garment for women and men, and the garment body 1 can be tightly attached to a human body as long as the garment body has elasticity. The garment body 1 of the embodiment has different sizes, including seven models from XXS-XXL, and the proper garment body 1 can be selected for testing according to the height and the weight of a measured person.
The flexible yarn sensor 2 comprises a weft yarn sensor 21 and a warp yarn sensor 22, wherein the weft yarn sensor 21 is fixed on the corresponding human chest, waist, neck, arm and wrist parts of the garment body 1, arranged in the circumferential direction of the human body and used for measuring the girth information of the human body. The warp yarn sensors 22 are fixed on the garment body 1 corresponding to the arms, shoulders and elbows of the human body, are arranged in the axial direction of the human body, and are used for measuring the motion information of the human body. The weft yarn sensor 21 and the warp yarn sensor 22 both include an elastic inner core, conductive yarns spirally wound on the surface of the elastic inner core in an "S" shape, and electrode wires fixedly connected to both ends of the conductive yarns.
Specifically, in the present embodiment, as shown in fig. 1 and 2, the flexible yarn sensors 2 are distributed on the garment front surface 11 and the garment back surface 12, and are respectively fixed at the positions of the garment body 1 corresponding to the human body to be measured. Wherein, the parts to be measured are the chest, waist, neck, arms, wrists, shoulders and elbows of the human body respectively. In this embodiment, the weft yarn sensor 21 can be used to test the electrical information corresponding to these body parts, and then the electrical information is converted into the body circumference information and the arm length information in the following steps, for example: the human body waistline, chest circumference, shoulder circumference, arm circumference, elbow circumference, arm length and the like are beneficial to the construction of a human body three-dimensional model; the motion information of the elbows and the shoulders of the human body can be tested through the warp yarn sensors 22, the motion habits of the human body can be effectively analyzed, and more suitable clothes can be selected for testers.
The flexible yarn sensor 2 can be drawn according to the fact that the flexible yarn sensor 2 is drawn when a fitting person wears the clothes body 1, and therefore the change of the resistance of the flexible yarn sensor is caused to obtain the girth and the motion information of a human body. The weft sensor 21 has a certain resistance when it is in an unstretched state. When the conductive yarn is stretched, the conductive yarn wound on the elastic inner core is stretched by the elastic inner core, so that the surrounding gap formed by the conductive yarn on the surface of the elastic inner core is increased, and the resistance is increased. Therefore, when the chest, waist, neck, arm, wrist, shoulder and elbow parts of the human body are larger, the resistance is larger, different length information can be correspondingly obtained according to different resistances, and then the circumference information of each part of the human body can be obtained by acquiring and processing the change of resistance signals of the weft yarn sensor 21 through the signal processor 3. The warp yarn sensor 22 is arranged at the position where the human body needs to move, when the positions move, the elastic inner core bends, and the conductive yarns wound on the surface of the elastic inner core bend due to traction, so that the surrounding gap of the outer side of the conductive yarns far away from the joint is increased, and the resistance is increased, therefore, when the movement angle of the positions is larger, the resistance is larger, and the clothes which accord with the movement range of a fitting person can be selected. The signal processor 3 can collect and process the change of the resistance signal of the warp yarn sensor 22 to obtain the motion information of the elbow and the shoulder of the human body. And the angle range that this warp direction yarn sensor 22 can accurate measurement is 0 ~ 180, and sensitivity is high, and the pliability is good.
As a further optimization, in the preferred embodiment of the present invention, as shown in fig. 1 and 2, the weft yarn sensor 21 fixed at the corresponding position of the chest, waist, arm and wrist of the human body is located at the front 11 of the garment, and the weft yarn sensor 21 fixed at the corresponding position of the neck of the human body is located at the back 12 of the garment. The characteristics of the chest and the waist of a human body are obvious, the fixation on the front 11 of the garment is beneficial to accurately measuring the circumference data of the human body, and the characteristics of the neck of the human body are obvious on the back and need to be fixed on the back. The reason why the weft yarn sensor 21 is fixed only to any one side of the garment body 1 is that the size of the circumference of the whole human body can be estimated from the resistance change rate of the half-stretching after the garment body 1 is worn, which is more material-saving. The warp yarn sensor 22 fixed on the corresponding arm and shoulder of the human body is positioned on the front side 11 of the garment; the warp yarn sensors 22 fixed at the corresponding elbow parts of the human body are positioned on the back surface 12 of the garment. The range of motion of the arms and shoulders of the human body is mainly in the front side, and the elbow joint is positioned on the back side of the human body, so that the arrangement of the warp yarn sensors 22 aiming at each part is more beneficial to acquiring the motion information of the human body.
As a further optimization, the diameter of the flexible yarn sensor 2 is 0.1mm to 20 mm. More preferably, the diameter of the flexible yarn sensor 2 is 0.7 mm-10 mm, and the flexible yarn sensor 2 with the specification has small diameter and good flexibility, can be attached to a human body, and can measure more accurate human body information. The diameter of the flexible yarn sensor 2 referred to above refers to the diameter of the body part not including the electrode wires.
As a further optimization, in this embodiment, the elastic inner core is spandex yarn, and the conductive yarn is silver-plated nylon yarn.
Specifically, in this embodiment, the elastic core is made of spandex yarn. When the flexible yarn sensor 2 is attached to a garment and worn by a fitter, it will stretch to different degrees due to the different sizes of the fitter, separating the surrounding gaps of the conductive yarn. Therefore, the elastic inner core prepared from the spandex material has higher tensile property and is more suitable for being used as a supporting inner core. The conductive yarn can be silver-plated nylon yarn sold in markets, such as silver-plated nylon yarn sold by Qingdao Zhihang functional fabric GmbH. The silver-plated nylon yarn is a yarn material with a conductive function, has stretching and bending properties, can change resistance along with the change of a contact area, and can be used as a conductive yarn for the flexible yarn sensor 2.
Preferably, the elastic inner core is spandex yarn of 10D-1800D, and the conductive yarn is silver-plated nylon yarn of 30D-300D. More preferably, the elastic inner core is spandex yarn with the thickness of 450D-700D, and the conductive yarn is silver-plated nylon yarn with the thickness of 65D-245D. The elastic inner core and the conductive yarn with proper diameters are selected, so that the attaching degree of the flexible yarn sensor 2 and the skin can be further improved, and accurate measurement of data is facilitated.
Furthermore, the flexible yarn sensor 2 is fixed on the clothing body 1 through sewing threads, so that the flexible yarn sensor 2 cannot fall off from the clothing body 1 in the wearing and moving processes.
As further optimization, the electrode wire is a metal conductive wire or is formed by a base yarn and a conductive material layer wrapped on the surface of the base yarn; the conductive material layer is made of one of metal nano particles, metal nano wires, carbon materials or conductive high polymer materials, and the carbon materials are made of one of amorphous carbon, graphite, carbon nano tubes and graphene; the conductive polymer material is selected from one of polyaniline, polythiophene, polypyrrole, polyaniline derivatives, polythiophene derivatives and polypyrrole derivatives.
Specifically, the base yarn is artificial fiber or natural fiber, such as terylene, chinlon, silk fibroin fiber, and the like. It can be understood that the material of the metal conductive wire may be gold, silver, copper or other conductive metals, and the conductive material layer may be coated on the surface of the base yarn by means of dipping or coating. In addition, the electrode wires can also be made of silver-plated nylon yarns, so that the electrode wires are softer in texture, have no foreign body sensation and are woven into the garment body 1 in a changed manner. The electrode wires can be respectively and fixedly connected with the two ends of the conductive yarn through conductive adhesive. The conductive adhesive can be one of silver conductive adhesive, gold conductive adhesive, copper conductive adhesive and carbon conductive adhesive. In the preferred embodiment of the present invention, the conductive silver paste in the silver conductive adhesive is selected as the conductive adhesive.
The signal processor 3 is fixed on the garment body 1, comprises a signal acquisition module 31, a signal processing module 32, a wireless signal transmission module 33 and a power supply module which are connected in sequence, and is used for converting signals output by the flexible yarn sensor 2 into resistance values; wherein, the signal acquisition module 31 is connected with the flexible yarn sensor 2 through an electrode wire.
Specifically, the electrode wires of the flexible yarn sensor 2, the signal acquisition module 31, the signal processing module 32, the wireless signal transmission module 33, and the power module are welded with the connecting wires on the circuit board to form the close-fitting garment body 1 with the human body information detection function. The deep learning module 4 is wirelessly connected to the wireless signal transmission module 33. As shown in fig. 3, the power module provides power for the whole circuit, when the fitting person wears the garment body 1, the resistance of the flexible yarn sensor 2 changes, and the signal acquisition module 31 connected with the flexible yarn sensor 2 acquires the resistance change signal of the sensor and transmits the resistance change signal to the signal processing module 32. The signal processing module 32 includes a signal amplifying unit 321, a filtering unit 322, and a signal analyzing unit 323. Since the resistance signal is small, the resistance signal needs to be amplified by the signal amplifying unit 321 before entering the filtering unit 322. The filtering unit 322 may remove an interference signal from the resistance variation signal, and then analyze the resistance variation signal through the signal analyzing unit 323 to obtain the resistance value of each body part. The resistance value is then wirelessly transmitted to the deep learning module 4 through the wireless signal transmission module 33.
It can be understood that the power module may adopt a 2000mAh lithium battery, the wireless signal transmission module 33 may adopt a CC2541 main control bluetooth chip, which has the function of the signal analysis unit 323 in the signal processing module 32, and may also connect the resistance value transmission deep learning module 4 of the tested human body information through bluetooth. The signal acquisition module 31 and the signal processing module 32 are in the prior art, and are not described herein.
As a further optimization, in the preferred embodiment of the present invention, the fitting device further comprises a clamping member, the clamping member comprises a clamping piece and a viscose layer arranged on the periphery of the clamping piece, and the viscose layer is bonded with the clothing body 1 to fix the signal processor 3 between the clamping piece and the clothing body 1. The clamping piece is made of cloth materials, the adhesive layer at the periphery is a thermoplastic adhesive lining, the periphery of the clamping piece is adhered to the garment body 1 through a hot pressing method, the circuit board can be placed in the accommodating space formed by the clamping piece and the garment, and when the circuit board or the sensor is damaged, the circuit board can be maintained through disassembling the clamping piece or a new sensor is welded on the circuit board again.
And the deep learning module 4 is connected with the wireless signal transmission module 33 and is used for acquiring the predicted human body parameters through the mapping relation between the resistance value stored in the deep learning module 4 and the actual human body parameters. The deep learning module 4 can be realized by an FPGA chip, the resistance value obtained by measuring the flexible yarn sensor 2 and processing the resistance value by the signal processor 3 and the corresponding actual human body parameter are input into the FPGA chip for training by means of a deep learning technology, the mapping relation between the resistance value and the actual human body parameter is obtained and stored, and the predicted human body parameter can be obtained after the resistance value obtained by subsequent testing is input conveniently. In addition to the above method, the deep learning module 4 may also be implemented by constructing a deep learning model, such as an LSTM model.
And the signal display 5 is connected with the deep learning module 4 and used for constructing and displaying a human body three-dimensional model based on the predicted human body parameters. The signal display 5 receives the predicted human body parameters sent by the deep learning module 4, constructs a human body three-dimensional model according to the human body parameters and displays the human body three-dimensional model, and the signal display 5 can be a personal computer or a tablet and the like. It should be noted that, the method of the human body three-dimensional model is the prior art, and is not described herein again. The signal display 5 stores a large amount of garment information for fitting on the constructed three-dimensional model of the human body. Wherein, the deep learning module 4 can be installed in the signal display 5 for connection and processing.
Compare with traditional virtual fitting system, the utility model provides a fitting equipment based on degree of depth study and fitting method thereof need not the fitting personnel and obtains the information that individual prepared through optical scanning, has reduced fitting personnel discomfort and the sense of tension at the measurand in-process. The utility model discloses in adopt the clothing that disposes flexible yarn resistance-type sensor as human waistline, chest circumference, shoulder circumference, arm circumference, elbow circumference, the measuring tool of sleeve length, adopt the method of degree of depth study to establish the corresponding relation between human body circumference and each position yarn sensor resistance, the person of awaiting measuring only needs to be on the person the utility model provides a resistance value of the just human each health position of clothing of fitting equipment, the resistance value in the rethread degree of depth study module 4 and the mapping relation of actual human parameter find the prediction human parameter to according to predicting human parameter and establishing human three-dimensional model. The method can obtain the static circumference information of the human body, can record the dynamic circumference information and the motion information of the human body in real time, and is simple, easy and efficient in test method.
The second embodiment of the utility model provides a fitting method based on deep learning uses the first embodiment to provide fitting equipment based on deep learning to realize fitting, and it can be carried out by signal display 5, and wherein, deep learning module 4 is the deep learning model that founds at signal display 5, including following step:
s10, acquiring the resistance value of each body part of the fitting person based on fitting equipment; wherein the body part comprises chest cavity, waist, neck, arm, wrist, shoulder and elbow.
In the embodiment, before the test, the fitting person provides height and weight of the fitting person, and the size of the close-fitting garment body 1 is determined according to the ratio of the height to the weight. After the garment body 1 is worn, the resistance of the flexible yarn sensor 2 fixed at each part of the garment body 1 changes due to stretching, and the resistance is acquired and processed by the signal processor 3, so that the resistance value of each body part is obtained.
S20, acquiring predicted human body parameters based on the resistance value and a pre-established deep learning model; the deep learning model is obtained by training the mapping relation between the resistance values obtained by arranging the flexible yarn sensor 2 on each body part and the actual human body parameters; the predicted human body parameters and the actual human body parameters comprise the chest circumference, the waist circumference, the neck circumference, the arm circumference, the wrist circumference, the shoulder circumference, the elbow circumference and the arm length of the human body.
In this embodiment, the deep learning model is obtained by training a mapping relationship between the resistance value of the flexible yarn sensor 2 and an actual human body parameter. Therefore, after the deep learning model receives the resistance value, the predicted human parameter information is obtained through the mapping relation between the resistance value and the actual human parameter, and the information comprises the information of the chest circumference, the waist circumference, the neck circumference, the arm circumference, the wrist circumference, the shoulder circumference, the elbow circumference and the arm length of the human body, so that the subsequent construction of the three-dimensional model of the human body is facilitated. Table 1 shows the predicted human body parameters obtained by the deep learning model after eight fitting persons try on the garment body 1.
TABLE 1 waist, chest, shoulder, arm, elbow and sleeve length data measured by fitting device based on deep learning
| Tester | Waist/cm | Circumference of chest/cm | Shoulder circumference/cm | Wrist circumference/cm | Elbow circumference/cm | Length/cm of |
| 1 | 767 | 851 | 1001 | 156 | 242 | 58 |
| 2 | 782 | 856 | 987 | 152 | 249 | 57 |
| 3 | 824 | 891 | 1049 | 182 | 257 | 56 |
| 4 | 736 | 843 | 1006 | 161 | 233 | 55 |
| 5 | 829 | 863 | 1025 | 155 | 243 | 58 |
| 6 | 791 | 869 | 965 | 150 | 244 | 57 |
| 7 | 759 | 854 | 998 | 156 | 241 | 56 |
| 8 | 804 | 897 | 1005 | 185 | 260 | 56 |
And S30, constructing a human body three-dimensional model based on the predicted human body parameters, and performing virtual fitting on the human body three-dimensional model.
In the embodiment, the signal processor 3 is provided with software for constructing the three-dimensional model of the human body, the three-dimensional model of the human body can be constructed when the predicted human body parameters transmitted by the deep learning model are received, and the signal processor 3 stores a large amount of clothes information, so that the fitting can be directly performed on the three-dimensional model of the human body, and the fitting is simple and quick. Figure 4 is eight fitting personnel's human three-dimensional model, can discover the fitting personnel's of every size three-dimensional model homoenergetic from the three-dimensional model on the picture by fine construction, has proved the utility model provides a fitting equipment's feasibility.
As a further optimization, on the basis of the above embodiment, the building step of the deep learning model includes:
acquiring actual human body parameters of a tester;
based on the flexible yarn sensor 2, obtaining the resistance value of each body part of the tester;
based on the deep learning technology, the actual human body parameters and the resistance value are initially trained to obtain a deep learning model so as to obtain the mapping relation between the resistance value of the flexible yarn sensor 2 and the actual human body parameters.
Specifically, in this embodiment, the actual circumference and the sleeve length of the tester can be measured by the tape for a plurality of times, so as to obtain the actual human body parameters. Then, arranging flexible yarn sensors 2 at all parts of the human body of the same tester, and recording the resistance values of the sensors at all parts in real time; and then the resistance values of all body parts of the tested person are obtained after the resistance values are processed by the signal processor 3. After the resistance values of all parts of the human body correspond to the actual human body parameters, the actual human body parameters and the resistance values are subjected to initialization training based on a deep learning technology to obtain a deep learning model, so that the mapping relation between the resistance values of the flexible yarn sensor 2 and the actual human body parameters can be obtained. The deep learning model may be an LSTM (Long Short-Term Memory, LSTM) model, or may be a deep neural network model, which is not limited herein. The method comprises the steps that at least 20 dressing resistance data of testers are collected by each garment body 1 of different types, the dressing resistance data are correlated with data of waistline, chest circumference, shoulder circumference, arm circumference, elbow circumference and sleeve length of a human body obtained by measuring with a tape, deep learning is conducted, and the numerical relation between the resistance of the flexible yarn sensor 2 and the waistline, chest circumference, shoulder circumference, arm circumference, elbow circumference and sleeve length of the human body is established, so that the accuracy of training data is guaranteed.
As a further optimization, in the preferred embodiment of the present invention, based on the predicted human body parameters, the method further includes the following steps after the human body three-dimensional model is constructed:
determining the release amount of the garment pieces of the fitting personnel according to the motion information of the fitting personnel;
and acquiring the clothes models conforming to the fitting personnel based on the release amount of the cut pieces so as to screen clothes suitable for the fitting personnel for carrying out virtual fitting on the human body three-dimensional model.
Specifically, in this embodiment, the warp yarn sensors 22 can test the motion curves of the fitting person's elbow, arm and shoulder, which are the motion information of the fitting person. The garment piece loosening amount of the garment is determined through the motion information, so that the clothes screening method is beneficial to helping the fitting personnel to screen the garments, the screened garments are guaranteed to have enough space for the fitting personnel to move, the condition of tightening hands and feet cannot occur, and the types of the garments and the residual inventory of the corresponding garment shop can be clear at a glance. The clothes can be customized through the motion information, and the application range of the fitting equipment is further expanded.
More preferably, in another implementation of the present invention, the fitting method further includes:
constructing a corresponding relation between the resistance value and the predicted human body parameter;
and training and updating the deep learning model based on the corresponding relation.
The data tested each time can be used as training data to train, the sample data of the deep learning model is enlarged, and the accuracy of the output predicted human body parameters is improved.
The embodiments described above are some, but not all embodiments of the present invention. The detailed description of the embodiments of the present invention is not intended to limit the scope of the invention as claimed, but is merely representative of selected embodiments of the invention. Based on the embodiments in the present invention, all other embodiments obtained by a person skilled in the art without creative efforts belong to the protection scope of the present invention.
Claims (6)
1. A fitting apparatus based on deep learning, comprising:
the close-fitting garment body comprises a garment front side and a garment back side, which respectively correspond to one side of the chest cavity and one side of the back of a human body;
the flexible yarn sensor comprises a weft yarn sensor and a warp yarn sensor; the weft yarn sensors are fixed on the parts of the garment body corresponding to the chest, waist, neck, arms and wrists of a human body, are arranged along the circumferential direction of the human body and are used for measuring the circumference information of the human body; the warp yarn sensors are fixed on the garment body corresponding to the arms, shoulders and elbows of the human body, are arranged in the axial direction of the human body and are used for measuring the motion information of the human body; the weft yarn sensor and the warp yarn sensor both comprise an elastic inner core, conductive yarns spirally wound on the surface of the elastic inner core according to an S shape and electrode wires fixedly connected to two ends of the conductive yarns;
the signal processor is fixed on the garment body, comprises a signal acquisition module, a signal processing module, a wireless signal transmission module and a power supply module which are sequentially connected, and is used for converting the signals output by the flexible yarn sensor into resistance values; the signal acquisition module is connected with the flexible yarn sensor through an electrode wire;
the deep learning module is connected with the wireless signal transmission module and used for acquiring predicted human body parameters through the mapping relation between the resistance value stored in the deep learning module and the actual human body parameters;
and the signal display is connected with the deep learning module and used for constructing and displaying a human body three-dimensional model based on the predicted human body parameters.
2. Fitting equipment based on deep learning according to claim 1, characterized in that the diameter of the flexible yarn sensor is 0.1-20 mm.
3. The fitting equipment based on deep learning of claim 1, wherein the electrode wire is a metal conductive wire or is formed by a base yarn and a conductive material layer wrapped on the surface of the base yarn; the conductive material layer is made of one selected from metal nanoparticles, metal nanowires, carbon materials or conductive polymer materials, and the carbon materials are one selected from amorphous carbon, graphite, carbon nanotubes and graphene; the conductive polymer material is selected from one of polyaniline, polythiophene, polypyrrole, polyaniline derivatives, polythiophene derivatives and polypyrrole derivatives.
4. The fitting apparatus based on deep learning of claim 1, wherein the weft yarn sensors fixed at the corresponding parts of the chest, waist, arms and wrists of the human body are positioned at the front of the garment; the weft yarn sensor fixed on the part corresponding to the neck of the human body is positioned on the back of the garment.
5. The deep learning based fitting apparatus according to claim 1, wherein the warp yarn sensors fixed to corresponding arms and shoulders of a human body are located on the front side of the garment; the warp yarn sensors fixed at the corresponding elbow parts of the human body are positioned on the back of the garment.
6. The deep learning-based fitting apparatus according to claim 1, further comprising a clamping member, wherein the clamping member comprises a clamping piece and an adhesive layer arranged on a periphery of the clamping piece, and the adhesive layer is bonded to the garment body to fix the signal processor between the clamping piece and the garment body.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202020076253.8U CN212345491U (en) | 2020-01-14 | 2020-01-14 | Fitting equipment based on deep learning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202020076253.8U CN212345491U (en) | 2020-01-14 | 2020-01-14 | Fitting equipment based on deep learning |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN212345491U true CN212345491U (en) | 2021-01-15 |
Family
ID=74137554
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202020076253.8U Active CN212345491U (en) | 2020-01-14 | 2020-01-14 | Fitting equipment based on deep learning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN212345491U (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111109736A (en) * | 2020-01-14 | 2020-05-08 | 厦门大学 | Fitting equipment based on deep learning and fitting method thereof |
-
2020
- 2020-01-14 CN CN202020076253.8U patent/CN212345491U/en active Active
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111109736A (en) * | 2020-01-14 | 2020-05-08 | 厦门大学 | Fitting equipment based on deep learning and fitting method thereof |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Farringdon et al. | Wearable sensor badge and sensor jacket for context awareness | |
| CN105266817B (en) | A kind of upper extremity function motion monitoring system and method based on fabric sensor | |
| US7750790B2 (en) | Fabric-based strain gauge | |
| US11253203B2 (en) | Object, method, and system for detecting heartbeat or whether or not electrodes are in proper contact | |
| Yan et al. | Weaved piezoresistive triboelectric nanogenerator for human motion monitoring and gesture recognition | |
| Yao et al. | A pressure mapping imaging device based on electrical impedance tomography of conductive fabrics | |
| CN104582517A (en) | Sensor | |
| CN103324288A (en) | System and method for human body movement identification of combined sensor | |
| TW201835526A (en) | Size measurement device and size measurement system | |
| CN212345491U (en) | Fitting equipment based on deep learning | |
| CN105982380A (en) | Intelligent garment achieving body data measurement through tension sensors | |
| JP2011036524A (en) | Electrode for living body | |
| CN106539172A (en) | A kind of intelligent clothing for measuring body dimension | |
| Esfahani et al. | Novel printed body worn sensor for measuring the human movement orientation | |
| CN111109736A (en) | Fitting equipment based on deep learning and fitting method thereof | |
| Zang et al. | A facile, precise radial artery pulse sensor based on stretchable graphene-coated fiber | |
| KR20100063595A (en) | Apparatus and method for non-intrusive wearable movement measuring using electro-conductive fiber | |
| Bystricky et al. | A comparison of embroidered and woven textile electrodes for continuous measurement of ECG | |
| CN107708561A (en) | Index equipment leading out, wearable device and portable set | |
| Metcalf et al. | Fabric-based strain sensors for measuring movement in wearable telemonitoring applications | |
| CN115808569A (en) | Method for measuring resistance of nano composite material piezoresistive sensor based on multivibrator | |
| CN101017114A (en) | Cloth-based strain gauge | |
| CN209916006U (en) | Motion gesture monitor | |
| CN107468254A (en) | Intelligent rehabilitation device, intelligent rehabilitation gloves and its signal processing method | |
| CN112535473A (en) | Wearable nano sensor for measuring clothes and measuring system thereof |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| GR01 | Patent grant | ||
| GR01 | Patent grant |