CN108648053A - A kind of imaging method for virtual fitting - Google Patents
A kind of imaging method for virtual fitting Download PDFInfo
- Publication number
- CN108648053A CN108648053A CN201810441208.5A CN201810441208A CN108648053A CN 108648053 A CN108648053 A CN 108648053A CN 201810441208 A CN201810441208 A CN 201810441208A CN 108648053 A CN108648053 A CN 108648053A
- Authority
- CN
- China
- Prior art keywords
- picture
- clothes
- human body
- customer
- data
- 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.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/06—Buying, selling or leasing transactions
- G06Q30/0601—Electronic shopping [e-shopping]
- G06Q30/0641—Shopping interfaces
- G06Q30/0643—Graphical representation of items or shoppers
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/30—Determination of transform parameters for the alignment of images, i.e. image registration
- G06T7/33—Determination of transform parameters for the alignment of images, i.e. image registration using feature-based methods
- G06T7/344—Determination of transform parameters for the alignment of images, i.e. image registration using feature-based methods involving models
Abstract
The present invention relates to a kind of imaging methods for virtual fitting, including：The markup information of one utilization clothing rule establishes human body AI models and is based on Mask RCNN training to model.Two：Garment data picture is obtained according to the markup information of amount clothing rule and is stored into garment data.Three obtain customer's somatic data picture according to the markup information of amount clothing rule.Four data analyses and the matching of people's clothing；The human body AI models that the customer's somatic data picture obtained in step 3 is brought into step 1 are trained；Human region position to the output of AI models and key point, that is, markup information, and estimate the size in each region of customer's human body.The clothes of suitable dimension are retrieved according to the dimension information in each region of customer's human body estimated and are fitted；Clothes are fitted successful result with human body to be shown by image engine.This method fitting efficiency improves；The effect that can not only be shown build and be tried on a dress, can also really show whether the clothes and the colour of skin, head dummy, makings arrange in pairs or groups.
Description
Technical field
The present invention relates to virtual fitting fields, and in particular to a kind of imaging method for virtual fitting.
Background technology
E-commerce becomes a kind of trend, and present all trades and professions are all to e-commerce development.Clothes are people's lives
Requirement, to clothes it is aesthetic, to whether meeting the individual character of buyer, be through on one's body whether it is suitable the problems such as in net purchase
Be required for solving in actual life, if these problems do not have effective processing method that will cause certain damage to net purchase
It loses.After being chosen above the net purchase platform now with many net consumers, have been found that not to be oneself imagination after waiting clothes in one's hands
In effect, so as to cause return goods to the distrust of seller and the distrust of shopping platform.
Already existing three-dimensional modeling emulation is to be previously entered height, weight by user, then carry out body in the prior art
Type feature selects, and shows clothes effect from the point of view of the manikin of an emulation, and no hardware device does data source support, fits
A substantially virtual model similar with oneself stature fit and the effect of oneself true man difference it is too big.In the prior art
In there is also pass through plane mirror AR imaging to realize virtual fitting.Compared to Dummy modeling, this mode can be very good to solve fitting
The problem of validity, but clothes effect is planar, cannot 360 ° see the effect worn the clothes.Therefore need one kind can
Accurately, the comprehensive virtual fit method for showing clothes effect.
Invention content
1, technical problem to be solved：
The present invention provides a kind of virtual fit method, is adopted at manufacturer end using the human body pictorial information for acquiring customer at customer end
The pictorial information for collecting clothes three-dimensionally shows the effect of customer's fitting by image procossing and machine learning techniques.
2, technical solution：
A kind of imaging method for virtual fitting includes the following steps：
Step 1：Human body AI models are established using markup information；Wherein markup information is the people for needing to mark according to amount clothing rule
Each region of body and key point；The data for inputting the human body markup information of various human body statures trained based on Mask-RCNN, are obtained
The human body AI models of stature must be corresponded to.
Step 2：It obtains garment data picture and is stored into garment data；It is obtained by picture collection device same
Each bearing images of dress, and each bearing images are pre-processed, and identify the data of the markup information in the clothes；
The information storage of pretreated garment data picture and the clothes being manually entered is entered into garment data.
Step 3：Obtain customer's somatic data picture；User shows preset common fitting posture, each to obtain human body
Bearing images；The image of acquisition is subjected to pretreatment and obtains customer's somatic data picture.
Step 4：Data analysis and the matching of people's clothing；Bring the customer's somatic data picture obtained in step 3 into step 1
In human body AI models be trained；Human region position to the output of AI models and key point, that is, markup information, and calculate
Go out the size in each region of customer's human body.
The clothes of suitable dimension are retrieved according to the dimension information in calculated each region of customer's human body, and pass through Texture
Clothes key point and human body key point are fitted by mapping；If human dimension exceeds the preset range in region of clothes,
Flexible clothes carry out deformation matching in corresponding key point position.
Clothes and human body are fitted successful result to be shown by image engine, shown in effect picture also wrap
Include face, head, arm and other positions not covered by clothes except human body.
Further, further include step 5：According to user to the feedback of bandwagon effect；If human region position and key
The magnitude estimation of point is inaccurate, and the image data of the customer is inputted human body AI models, that is, uses the data further to train mould
Type increases training set.
Further, the markup information is to need each region of human body in mark image, including neck according to amount clothing rule
Portion, shoulder, chest, abdomen, buttocks, crotch, crotch, the key point in four limbs region and each region.
Further, it is to the pretreated detailed process of the picture of the clothes of acquisition progress in the step 2：
S21 acquires the static background figure in each orientation in place, utilizes Pinhole Camera in specifically acquisition place
Model demarcates the bulk and relative position in place.
S22 receives the image of clothing in each orientation, and adjusts picture according to the bulk and relative position in calibration place
Size；And the dual-shoulder position for adjusting clothes in image is maintained at predeterminated level height；After image load, Mean is utilized
Filter realizes front and back scape picture separation；Wherein foreground picture is the picture in each orientation of clothes, and background is background picture.
The picture that S23 obtains S22 carries out being based on Pulse Coupled Neural Network（PCNN）, small echo is set again in conjunction with double
（DTCWT）With Laplce's component definitely sum（SML）Processing, foreground data in step S22 is synthesized with preset background to be had
The picture format of Alpha channel informations；Obtain garment data picture.
S24 administrators input the label of the clothes；The content of label includes the corresponding size of the clothes, style information；It will
Label is stored into garment data after being associated with garment data picture.
Further, the detailed process of step 3 acquisition customer's somatic data picture includes：
S31：In specifically acquisition place, the static background figure in each orientation in place is acquired, Pinhole Camera are utilized
Model demarcates the bulk and relative position in place.
S32：The custom image in each orientation is received, and picture is adjusted according to the bulk and relative position in calibration place
Size；After custom image load, realize that front and back scape picture detaches using Mean filter；Wherein foreground picture is customer's
The picture in each orientation, background are background picture.
S33：The picture that S32 is obtained carries out being based on Pulse Coupled Neural Network（PCNN）, small echo is set again in conjunction with double
（DTCWT）With Laplce's component definitely sum（SML）Processing, foreground data in step S32 is synthesized with self-defined background to be had
The picture format of Alpha channel informations is customer's somatic data picture.
3, advantageous effect：
Human body AI models are first preset in the present invention, and the human body image data of customer is brought into AI models and trains customer correspondence
Human body AI models, the clothes for being suitble to customer's build that can wear is selected by corresponding human body AI models.Customer can basis
The label of clothes selects the type oneself liked to fit, and the build after its fitting is shown by way of stereo-picture.
Only the build of customer is fitted in the design sketch of displaying, other positions or true picture.Its examination of this method
Clothing efficiency improves；The effect that customer can not only see build and be tried on a dress, can also really find out the clothes and the colour of skin, head
Whether type, makings arrange in pairs or groups.The time of customer's fitting can be saved by carrying out virtual fitting by this method simultaneously, and customer can try
Clothing style and quantity greatly increase；And realize a secondary amounts body, it is repeatedly in and buys clothing.
Specific implementation mode
Embodiments of the present invention are specifically described.
This method is divided into manufacturer end, customer end and high in the clouds.Manufacturer end utilizes garment data harvester by different garment root
Comprehensive acquisition picture is carried out according to different sizes.The picture of acquisition is gone out to need the human body marked according to amount clothing rule analysis
The data of each region clothes corresponding with key point.For example, to acquire arm according in amount clothing rule characteristics of human body's value
Length, i.e. shoulder to the distance of wrist, the key point is corresponded in garment image be exactly clothes sleeve upper and lower side it is corresponding
Coordinate value.These markup informations are included in the garment data picture of the clothes.Administrator is manually entered the label of the clothes,
And it will be in the garment data picture associated storage to clothes database of label and corresponding clothes.
Customer end is typically the characteristics of human body using the type collection customer of fitting room.The method of this method acquisition is to use
Camera it is comprehensive acquisition customer's different gestures when picture.And picture is handled, i.e., by the picture of acquisition according to amount
Clothing rule analysis goes out to need each region of human body marked and key point.Show that customer's somatic data picture, these pictures can lead to
The mobile phone that transmitting device is transmitted to user is crossed, user can store.When customer wants to try on a dress, directly in hand
The display diagram of my fitting can be found out on machine.And platform can be recorded according to the fitting history and conclusion of the business of user, be recommended
The favorite clothes new product of user.
The database in high in the clouds ceaselessly collects the picture of customer, and brings somatic data picture into human body AI models and instruct
Practice machine learning, to keep human body AI models more and more accurate.When high in the clouds receives each bearing images of human body that customer end is sent
When, high in the clouds is brought into human body AI models after being pre-processed to picture, export customer's somatic data picture of the customer, and should
Information is transmitted on the mobile terminal or preset presentation device of user.When high in the clouds receives the request of a certain customer need fitting
When, high in the clouds recalls the corresponding human body AI models of the customer, and the clothes of the suitable model are selected according to the model.Customer can root
It fits according to the label for selecting clothes is needed.What is be replaced in customer's pictorial information in this process only has with clothes
The human body parts of pass.
Method includes step in detail below：
A kind of imaging method for virtual fitting includes the following steps：
Step 1：Human body AI models are established using markup information；Wherein markup information is the people for needing to mark according to amount clothing rule
Each region of body and key point；The data for inputting the human body markup information of various human body statures trained based on Mask-RCNN, are obtained
The human body AI models of stature must be corresponded to.
Step 2：It obtains garment data picture and is stored into garment data；It is obtained by picture collection device same
Each bearing images of dress, and each bearing images are pre-processed, and identify the data of the markup information in the clothes；
The information storage of pretreated garment data picture and the clothes being manually entered is entered into garment data.
Step 3：Obtain customer's somatic data picture；User shows preset common fitting posture, each to obtain human body
Bearing images；The image of acquisition is subjected to pretreatment and obtains customer's somatic data picture.
Step 4：Data analysis and the matching of people's clothing；Bring the customer's somatic data picture obtained in step 3 into step 1
In human body AI models be trained；Human region position to the output of AI models and key point, that is, markup information, and calculate
Go out the size in each region of customer's human body.
The clothes of suitable dimension are retrieved according to the dimension information in calculated each region of customer's human body, and pass through Texture
Clothes key point and human body key point are fitted by mapping；If human dimension exceeds the preset range in region of clothes,
Flexible clothes carry out deformation matching in corresponding key point position.
Clothes and human body are fitted successful result to be shown by image engine, shown in effect picture also wrap
Include face, head, arm and other positions not covered by clothes except human body.
Further, further include step 5：According to user to the feedback of bandwagon effect；If human region position and key
The magnitude estimation of point is inaccurate, and the image data of the customer is inputted human body AI models, that is, uses the data further to train mould
Type increases training set.
Further, the markup information is to need each region of human body in mark image, including neck according to amount clothing rule
Portion, shoulder, chest, abdomen, buttocks, crotch, crotch, the key point in four limbs region and each region.
Further, it is to the pretreated detailed process of the picture of the clothes of acquisition progress in the step 2：
S21：In specifically acquisition place, the static background figure in each orientation in place is acquired, Pinhole Camera are utilized
Model demarcates the bulk and relative position in place.
S22：The image of clothing in each orientation is received, and picture is adjusted according to the bulk and relative position in calibration place
Size；And the dual-shoulder position for adjusting clothes in image is maintained at predeterminated level height；After image load, Mean is utilized
Filter realizes front and back scape picture separation；Wherein foreground picture is the picture in each orientation of clothes, and background is background picture.
S23：The picture that S22 is obtained carries out being based on Pulse Coupled Neural Network（PCNN）, small echo is set again in conjunction with double
（DTCWT）With Laplce's component definitely sum（SML）Processing, foreground data in step S22 is synthesized with self-defined background to be had
The picture format of Alpha channel informations；Obtain garment data picture.
S24：Administrator inputs the label of the clothes；The content of label includes the corresponding size of the clothes, style information；It will
Label is stored into garment data after being associated with garment data picture.
Further, the detailed process of step 3 acquisition customer's somatic data picture includes：
S31：In specifically acquisition place, the static background figure in each orientation in place is acquired, Pinhole Camera are utilized
Model demarcates the bulk and relative position in place.
S32：The custom image in each orientation is received, and picture is adjusted according to the bulk and relative position in calibration place
Size；After custom image load, realize that front and back scape picture detaches using Mean filter；Wherein foreground picture is customer's
The picture in each orientation, background are background picture.
S33：The picture that S32 is obtained carries out being based on Pulse Coupled Neural Network（PCNN）, small echo is set again in conjunction with double
（DTCWT）With Laplce's component definitely sum（SML）Processing, foreground data in step S32 is synthesized with default background to be had
The picture format of Alpha channel informations is customer's somatic data picture.Preset background therein is according to being
The background that system automatically selects or client selects as needed.
During above-mentioned, in order to control the size of each picture, to make background match with foreground, therefore select
The bulk and relative position in identical calibration place.Therefore, adjustment space scale is needed during picture is handled
With relative position.Because this content belongs to this field conventional technical means, just it is not unfolded to illustrate at this.
Although the present invention has been described by way of example and in terms of the preferred embodiments, they be not it is for the purpose of limiting the invention, it is any ripe
This those skilled in the art is practised, without departing from the spirit and scope of the invention, can make various changes or retouch from working as, therefore the guarantor of the present invention
Shield range should be subject to what claims hereof protection domain was defined.
Claims (5)
1. a kind of imaging method for virtual fitting includes the following steps：
Step 1：Human body AI models are established using markup information；Wherein markup information is the people for needing to mark according to amount clothing rule
Each region of body and key point；The data for inputting the human body markup information of various human body statures trained based on Mask-RCNN, are obtained
The human body AI models of stature must be corresponded to；
Step 2：It obtains garment data picture and is stored into garment data；Same part is obtained by picture collection device
Each bearing images of clothes, and each bearing images are pre-processed, and identify the data of the markup information in the clothes；It will be pre-
The information storage of treated garment data picture and the clothes being manually entered enters garment data；
Step 3：Obtain customer's somatic data picture；User shows preset common fitting posture, to obtain each orientation of human body
Image；The image of acquisition is subjected to pretreatment and obtains customer's somatic data picture；
Step 4：Data analysis and the matching of people's clothing；The customer's somatic data picture obtained in step 3 is brought into step 1
Human body AI models are trained；Human region position to the output of AI models and key point, that is, markup information, and calculate this
The size in each region of customer's human body；
The clothes of suitable dimension are retrieved according to the dimension information in calculated each region of customer's human body, and pass through Texture
Clothes key point and human body key point are fitted by mapping；If human dimension exceeds the preset range in region of clothes,
Flexible clothes carry out deformation matching in corresponding key point position；
Clothes and human body are fitted successful result to be shown by image engine, shown in effect picture further include removing
Face, head, arm and other positions not covered by clothes of human body.
2. a kind of imaging method for virtual fitting according to claim 1, it is characterised in that：It further include step 5：
According to user to the feedback of bandwagon effect；If the magnitude estimation of human region position and key point is inaccurate, by the customer's
Image data inputs human body AI models, that is, uses the further training pattern of the data, increases training set.
3. a kind of imaging method for virtual fitting according to claim 1, it is characterised in that：The markup information is
Each region of human body in mark image, including neck, shoulder, chest, abdomen, buttocks, crotch, crotch are needed according to amount clothing rule
The key point in portion, four limbs region and each region.
4. a kind of imaging method for virtual fitting according to claim 1, it is characterised in that：It is right in the step 2
The pretreated detailed process that the pictures of the clothes of acquisition carries out is：
S21 acquires the static background figure in each orientation in place, utilizes Pinhole Camera in specifically acquisition place
Model demarcates the bulk and relative position in place；
S22：The image of clothing in each orientation is received, and the big of picture is adjusted according to the bulk and relative position in calibration place
It is small；And the dual-shoulder position for adjusting clothes in image is maintained at predeterminated level height；After image load, Mean filter are utilized
Realize front and back scape picture separation；Wherein foreground picture is the picture in each orientation of clothes, and background is background picture；
S23：The picture that S22 is obtained carries out being based on Pulse Coupled Neural Network（PCNN）, small echo is set again in conjunction with double（DTCWT）With
Laplce's component definitely sum（SML）Processing, foreground data in step S22 is synthesized with preset background with the channels Alpha
The picture format of information；Obtain garment data picture；
S24：Administrator inputs the label of the clothes；The content of label includes the corresponding size of the clothes, style information；By label
It is stored into garment data after being associated with garment data picture.
5. a kind of imaging method for virtual fitting according to claim 1, it is characterised in that：The step 3 obtains
The detailed process of customer's somatic data picture includes：
S31：In specifically acquisition place, the static background figure in each orientation in place is acquired, Pinhole Camera are utilized
Model demarcates the bulk and relative position in place；
S32：The custom image in each orientation is received, and the big of picture is adjusted according to the bulk and relative position in calibration place
It is small；After custom image load, realize that front and back scape picture detaches using Mean filter；Wherein foreground picture is each of customer
The picture in orientation, background are background picture；
S33：The picture that S32 is obtained carries out being based on Pulse Coupled Neural Network（PCNN）, small echo is set again in conjunction with double（DTCWT）With
Laplce's component definitely sum（SML）Processing, foreground data in step S32 is synthesized with self-defined background with the channels Alpha
The picture format of information is customer's somatic data picture.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810441208.5A CN108648053A (en) | 2018-05-10 | 2018-05-10 | A kind of imaging method for virtual fitting |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810441208.5A CN108648053A (en) | 2018-05-10 | 2018-05-10 | A kind of imaging method for virtual fitting |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108648053A true CN108648053A (en) | 2018-10-12 |
Family
ID=63753989
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810441208.5A Pending CN108648053A (en) | 2018-05-10 | 2018-05-10 | A kind of imaging method for virtual fitting |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108648053A (en) |
Cited By (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109493330A (en) * | 2018-11-06 | 2019-03-19 | 电子科技大学 | A kind of nucleus example dividing method based on multi-task learning |
| CN109523593A (en) * | 2018-10-19 | 2019-03-26 | 北京陌上花科技有限公司 | Data processing method and device for image of clothing |
| CN109816492A (en) * | 2019-01-29 | 2019-05-28 | 许广明 | Implementation method, terminal and medium between virtual fitting |
| CN109934613A (en) * | 2019-01-16 | 2019-06-25 | 中德（珠海）人工智能研究院有限公司 | A kind of virtual costume system for trying |
| CN110059690A (en) * | 2019-03-28 | 2019-07-26 | 广州智方信息科技有限公司 | Floor plan semanteme automatic analysis method and system based on depth convolutional neural networks |
| CN110377824A (en) * | 2019-07-15 | 2019-10-25 | 贝壳技术有限公司 | Information-pushing method, device, computer readable storage medium and electronic equipment |
| CN110706359A (en) * | 2019-09-30 | 2020-01-17 | 恒信东方文化股份有限公司 | Image fitting correction method and system |
| CN110852941A (en) * | 2019-11-05 | 2020-02-28 | 中山大学 | Two-dimensional virtual fitting method based on neural network |
| CN111145207A (en) * | 2018-10-17 | 2020-05-12 | 深圳市衣锦未来科技有限公司 | On-line customization method for making clothes through photo measurement |
| CN111695525A (en) * | 2020-06-15 | 2020-09-22 | 恒信东方文化股份有限公司 | 360-degree clothes fitting display method and device |
| CN112001413A (en) * | 2020-07-13 | 2020-11-27 | 中国标准化研究院 | Human body shape data prediction system based on human body size database |
| CN112200717A (en) * | 2020-10-26 | 2021-01-08 | 广州紫为云科技有限公司 | Complex garment virtual fitting method and device based on neural network and storage medium |
| CN113012303A (en) * | 2021-03-10 | 2021-06-22 | 浙江大学 | Multi-variable-scale virtual fitting method capable of keeping clothes texture characteristics |
| CN113711269A (en) * | 2019-05-31 | 2021-11-26 | 预尺寸有限公司 | Method and system for determining body metrics and providing garment size recommendations |
| WO2022166805A1 (en) * | 2021-02-07 | 2022-08-11 | 上海英立视电子有限公司 | Virtual fitting method |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103678836A (en) * | 2012-08-30 | 2014-03-26 | 北京三星通信技术研究有限公司 | Virtual fit system and method |
| CN104517219A (en) * | 2013-09-28 | 2015-04-15 | 南京专创知识产权服务有限公司 | Taobao virtual fitting system |
| CN107924579A (en) * | 2015-08-14 | 2018-04-17 | 麦特尔有限公司 | The method for generating personalization 3D head models or 3D body models |
| CN108010134A (en) * | 2017-11-29 | 2018-05-08 | 湘潭大学 | A kind of real-time three-dimensional virtual fit method based on mobile terminal |
| CN108010122A (en) * | 2017-11-14 | 2018-05-08 | 深圳市云之梦科技有限公司 | A kind of human 3d model rebuilds the method and system with measurement |
-
2018
- 2018-05-10 CN CN201810441208.5A patent/CN108648053A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103678836A (en) * | 2012-08-30 | 2014-03-26 | 北京三星通信技术研究有限公司 | Virtual fit system and method |
| CN104517219A (en) * | 2013-09-28 | 2015-04-15 | 南京专创知识产权服务有限公司 | Taobao virtual fitting system |
| CN107924579A (en) * | 2015-08-14 | 2018-04-17 | 麦特尔有限公司 | The method for generating personalization 3D head models or 3D body models |
| CN108010122A (en) * | 2017-11-14 | 2018-05-08 | 深圳市云之梦科技有限公司 | A kind of human 3d model rebuilds the method and system with measurement |
| CN108010134A (en) * | 2017-11-29 | 2018-05-08 | 湘潭大学 | A kind of real-time three-dimensional virtual fit method based on mobile terminal |
Non-Patent Citations (1)
| Title |
|---|
| KAIMING HE等: "Mask R-CNN", 《COMPUTER VISION AND PATTERN RECOGNITION》 * |
Cited By (21)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111145207A (en) * | 2018-10-17 | 2020-05-12 | 深圳市衣锦未来科技有限公司 | On-line customization method for making clothes through photo measurement |
| CN109523593A (en) * | 2018-10-19 | 2019-03-26 | 北京陌上花科技有限公司 | Data processing method and device for image of clothing |
| CN109493330A (en) * | 2018-11-06 | 2019-03-19 | 电子科技大学 | A kind of nucleus example dividing method based on multi-task learning |
| CN109934613A (en) * | 2019-01-16 | 2019-06-25 | 中德（珠海）人工智能研究院有限公司 | A kind of virtual costume system for trying |
| CN109816492A (en) * | 2019-01-29 | 2019-05-28 | 许广明 | Implementation method, terminal and medium between virtual fitting |
| CN109816492B (en) * | 2019-01-29 | 2023-01-03 | 许广明 | Method, terminal and medium for realizing virtual fitting room |
| CN110059690A (en) * | 2019-03-28 | 2019-07-26 | 广州智方信息科技有限公司 | Floor plan semanteme automatic analysis method and system based on depth convolutional neural networks |
| CN113711269A (en) * | 2019-05-31 | 2021-11-26 | 预尺寸有限公司 | Method and system for determining body metrics and providing garment size recommendations |
| CN110377824B (en) * | 2019-07-15 | 2020-06-05 | 贝壳找房(北京)科技有限公司 | Information pushing method and device, computer readable storage medium and electronic equipment |
| CN110377824A (en) * | 2019-07-15 | 2019-10-25 | 贝壳技术有限公司 | Information-pushing method, device, computer readable storage medium and electronic equipment |
| CN110706359A (en) * | 2019-09-30 | 2020-01-17 | 恒信东方文化股份有限公司 | Image fitting correction method and system |
| CN110852941A (en) * | 2019-11-05 | 2020-02-28 | 中山大学 | Two-dimensional virtual fitting method based on neural network |
| CN110852941B (en) * | 2019-11-05 | 2023-08-01 | 中山大学 | Neural network-based two-dimensional virtual fitting method |
| CN111695525A (en) * | 2020-06-15 | 2020-09-22 | 恒信东方文化股份有限公司 | 360-degree clothes fitting display method and device |
| CN111695525B (en) * | 2020-06-15 | 2023-05-16 | 恒信东方文化股份有限公司 | 360-degree clothing fitting display method and device |
| CN112001413B (en) * | 2020-07-13 | 2023-06-06 | 中国标准化研究院 | Human body appearance data prediction system based on human body size database |
| CN112001413A (en) * | 2020-07-13 | 2020-11-27 | 中国标准化研究院 | Human body shape data prediction system based on human body size database |
| CN112200717A (en) * | 2020-10-26 | 2021-01-08 | 广州紫为云科技有限公司 | Complex garment virtual fitting method and device based on neural network and storage medium |
| WO2022166805A1 (en) * | 2021-02-07 | 2022-08-11 | 上海英立视电子有限公司 | Virtual fitting method |
| CN113012303A (en) * | 2021-03-10 | 2021-06-22 | 浙江大学 | Multi-variable-scale virtual fitting method capable of keeping clothes texture characteristics |
| CN113012303B (en) * | 2021-03-10 | 2022-04-08 | 浙江大学 | Multi-variable-scale virtual fitting method capable of keeping clothes texture characteristics |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108648053A (en) | A kind of imaging method for virtual fitting | |
| CN105637512B (en) | For creating the method and system of customed product | |
| US8571698B2 (en) | Simple techniques for three-dimensional modeling | |
| US8976230B1 (en) | User interface and methods to adapt images for approximating torso dimensions to simulate the appearance of various states of dress | |
| EP3745352B1 (en) | Methods and systems for determining body measurements and providing clothing size recommendations | |
| CN110175897A (en) | A kind of 3D synthesis fitting method and system | |
| US20130173226A1 (en) | Garment modeling simulation system and process | |
| US20130170715A1 (en) | Garment modeling simulation system and process | |
| CN109801380A (en) | A kind of method, apparatus of virtual fitting, storage medium and computer equipment | |
| CN107111833A (en) | Quick 3D model adaptations and anthropological measuring | |
| JP2013526735A (en) | How to determine clothing dimensions remotely | |
| WO2020203656A1 (en) | Information processing device, information processing method, and program | |
| US20080201641A1 (en) | Method And The Associated Mechanism For 3-D Simulation Stored-Image Database-Driven Spectacle Frame Fitting Services Over Public Network | |
| CN105069837B (en) | A kind of clothes trying analogy method and device | |
| US20020093515A1 (en) | Method of measuring and sizing objects from an image of a human face using iris size | |
| CN105374057A (en) | Virtual try-on apparatus and virtual try-on method | |
| CN108346156B (en) | User stature parameter acquisition method and device | |
| CN105374058A (en) | Virtual try-on apparatus, virtual try-on system and virtual try-on method | |
| CN108664884A (en) | A kind of virtually examination cosmetic method and device | |
| CN109871589A (en) | Intelligent clothing system and method based on Stereo face recognition | |
| CN108960985A (en) | Body parameter generation method and online shopping item recommendation method based on image or video | |
| CN103530782A (en) | WEB based online fitting-room system | |
| CN106773050A (en) | A kind of intelligent AR glasses virtually integrated based on two dimensional image | |
| CN103793832A (en) | Three-dimensional modeling method applied to network fitting and based on image collecting | |
| WO2020156627A1 (en) | The virtual caliper: rapid creation of metrically accurate avatars from 3d measurements |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20181012 |
|
| RJ01 | Rejection of invention patent application after publication |