CN109003298A - virtual glasses matching method and system - Google Patents

Abstract

The invention relates to a virtual glasses matching method and a system, which provide a user to use a computer device or a handheld device, screen out a plurality of alternative glasses frames by using the self images of the user obtained by shooting or scanning through a glasses matching application program, and then perform virtual try-on the alternative glasses frames in sequence to select a suitable glasses frame. The virtual glasses matching method comprises the following steps: shooting or obtaining a digital image of the face of the user; analyzing or measuring the digital image to obtain any one of facial features of interpupillary distance, face width, nose bridge height, nose bridge width and eye-ear distance; screening a plurality of alternative picture frames from a picture frame database based on the facial feature numerical data; and enabling the user to perform virtual try-on the plurality of optional glasses frames in sequence so as to select a suitable glasses frame.

Description

Method and system for matching virtual glasses

technical field

The present invention relates to a technology for providing users with selecting glasses, in particular to a method and system for providing users with selecting and matching virtual glasses in plane images or stereoscopic images by using software methods through image processing.

Background technique

Generally, eyewear companies in the market are the first choice for consumers who choose to wear glasses. Consumers will obtain optometry information from eyewear companies or hospitals, and then go to eyewear companies or eyewear stores that cooperate with hospitals to choose frames and lenses. It is common for employees of the company to introduce various frames and let consumers try them on in front of the mirror, and then choose the lens style they need. Then the company will make glasses according to some wearing data of consumers, such as the degree of glasses, and the number of pupils. The distance between the eyes, the position of the eyes, the wearing height on the bridge of the nose, the curvature of the mirror surface and the inclination angle of the frame, etc. Most of the data is to allow consumers to inform the service provider based on their self-wearing feeling when receiving the glasses, so that they can adjust in real time.

Contents of the invention

In order to overcome the above-mentioned deficiencies in the prior art, the present invention discloses a method for matching virtual glasses, which provides users with a computer device or a handheld device to shoot or scan their own images, and then choose suitable glasses according to their own face or head features. glasses, and can instantly display the image of wearing the virtual glasses on the display screen of the computer device, achieving the effect of real-time preview like looking in a mirror, and allowing the user to flexibly choose the virtual experience of the glasses style.

The virtual glasses matching method of the present invention includes an image acquisition step: taking or obtaining a digital image of the user's face; a facial feature value measurement step: analyzing or measuring the digital image to obtain the user's pupillary distance and face width , nose bridge height, nose bridge width, and facial features of eye-ear distance; frame screening step: based on the numerical data of the facial features, and using equation 1: P=-1.036X ² +23.365X-2.089Y +Z±5 (wherein, P is the frame width (mm); X is the user's nose height (mm); Y is the user's nose width (mm); Z is the user's eye-ear distance (mm)) from A plurality of candidate frames are screened out from the frame database; and a frame decision step: the user of the glasses application program conducts virtual try-ons on the plurality of candidate frames in turn to select a suitable frame; wherein the glasses application The program at least includes an image transmission interface, a frame screening interface, and a virtual frame try-on interface; wherein the image transmission interface is used to receive and transmit the user's facial image; the frame screening interface is used to accept the user's frame screening conditions and select a plurality of candidate frames from the frame database based on the frame screening conditions; the frame virtual try-on interface allows users to perform virtual try-ons on the plurality of candidate frames in turn.

The present invention further proposes a method for matching virtual glasses, which further includes wearing the identification object on the user's face, and photographing the user to create a digital image; identifying the user's two eyes in the stereoscopic image according to the position of the identification object , the position of the nose and two ears to obtain the facial feature information step; based on the facial feature screening value data, and using equation 1: P=-1.036X ² +23.365X-2.089Y+Z±5 (wherein, P is the width of the frame (mm); X is the height of the bridge of the user's nose (mm); Y is the width of the bridge of the user's nose (mm); Z is the distance between the eyes and ears of the user (mm); or use Equation 3: R=Z +13+F (where, R represents the length of temples (mm); Z represents the distance between eyes and ears (mm); F represents the upper and lower limits (mm), generally ranging from 0 to 40mm) selected from the frame database A plurality of candidate frames; the user of the above-mentioned glasses matching application program sequentially performs virtual try-ons on the plurality of candidate frames to select a suitable frame.

According to a specific embodiment, the present invention can also provide a method for matching virtual glasses, wherein the digital image can be a 2D image, a 3D image, or an augmented reality image.

According to a specific embodiment, the present invention can also provide a method for matching virtual glasses, wherein the frame screening step further includes selecting a plurality of candidate frames based on at least two facial features ; for example, using Equation 2: M=180-Q=180-Tan ^-1 (X/(Y/2)) (where M is the angle of the nose pad; Y is the width of the bridge of the nose; X is the height of the bridge of the nose: Q is the tilt of the bridge of the nose degrees.); or for example, use equation 3: R=Z+13+F (where R represents the length of the temples (mm); Z represents the distance between the eyes and ears (mm); F represents the upper and lower limits (mm), generally between 0 ~ 40mm range).

According to a specific embodiment, the present invention may also provide a method for matching virtual glasses, and the present invention may also provide a method for matching virtual glasses, wherein the frame determining step further includes displaying the The digital image is combined with the candidate frame for the user to decide whether the frame is suitable.

According to a specific embodiment, the present invention can also provide a method for matching virtual glasses, wherein after the frame screening step, a frame selection step is further included: the user manually inputs filter conditions, and filters glasses based on the filter conditions. Come up with multiple alternative frames.

According to a specific embodiment, the present invention can also provide a method for matching virtual glasses, wherein the screening conditions can be the width of the frame, the inclination of the bridge of the nose, the width of the nose pad distance, the length of the nose pad distance, the length of the temples, Any one or more of shape and color.

According to a specific embodiment, the present invention can also provide a method for matching virtual glasses, wherein the width of the face conforms to the following formula:

Face width T＝-1.036X ² +23.365X-2.089Y+Z

(Wherein, T is the width of the user's face, X is the height of the bridge of the user's nose, Y is the width of the bridge of the user's nose, and Z is the distance between the user's eyes and ears).

Also, because the screening range P of the frame width is approximately equal to the face width plus or minus one upper and lower threshold. The upper and lower limit thresholds, for example, may be 5mm or other numerical values. Therefore, when the upper and lower limit thresholds are set to 5 mm, the screening range P of the frame width satisfies the following equation 1:

P＝-1.036X ² +23.365X-2.089Y+Z±5-------(equation 1)

According to a specific embodiment, the present invention can also provide a method for matching virtual glasses, wherein the frame determination step further includes a physical wearing step: taking out the frame entity selected by the user from the warehouse for the user Wear it and confirm whether it meets your needs.

According to a specific embodiment, the present invention can also provide a method for matching virtual glasses, wherein the frame selection step, the frame selection step, and the frame determination step can be implemented repeatedly.

In addition to the aforementioned implementation aspects, the present invention further proposes a virtual glasses matching system, which includes a frame database and a computer device. The cloud server is used to store multiple glasses frame models; the computer device has a processing unit, a camera unit, a display unit, a communication unit and a storage unit, and the processing unit is electrically connected to the camera unit, display unit, communication unit and storage unit, and the communication unit network The road is connected to the cloud server, and the storage unit stores the matching application program. The camera unit is used to take digital images of the user, and the processing unit is used to execute the matching application program to obtain the first facial features and the second facial features of the user; the matching application program includes at least an image transmission interface, a frame screening interface, and a virtual frame try-on interface; wherein the image transmission interface is used to receive and transmit the user's facial image; the frame screening interface is used to accept the user's frame screening conditions and select the frames based on the frame screening conditions A plurality of candidate frames are selected from the database; the frame virtual try-on interface is for the user to perform virtual try-on for the plurality of candidate frames in turn;

According to a specific embodiment, the present invention can also provide a virtual glasses matching system, which includes: a frame database for storing a plurality of secondary frames; and a computer device, the computer device has a processing unit, A camera unit, a display unit, a communication unit and a storage unit, the processing unit is electrically connected to the camera unit, a display unit, the communication unit and the storage unit, and the communication unit is connected to the network In the frame database, the storage unit stores a matching application program, and the processing unit executes the matching application program; wherein the matching application program may include: for receiving and transmitting the user's facial image an image transmission interface, a frame screening interface for accepting the user's frame screening conditions and screening out a plurality of candidate frames from the frame database based on the frame screening conditions, and performing virtual tests on the multiple candidate frames in sequence The virtual try-on interface of the frame to be worn; wherein the frame screening is based on the facial feature numerical data, and using equation 1: P=-1.036X ² +23.365X-2.089Y+Z±5 (wherein, P is the frame Width (mm); X is the height of the bridge of the user's nose (mm); Y is the width of the bridge of the user's nose (mm); Z is the distance between the user's eyes and ears (mm)) Select multiple options from the frame database picture frame.

According to a specific embodiment, the present invention can also provide a virtual eyewear matching system, wherein the digital image can be obtained by taking a picture through the camera unit, or transmitting it to the processing unit through the communication unit.

The virtual glasses matching method and system provided by the present invention provide users with the ability to select frame types and frame-related parameters through a graphical user interface, and the glasses matching system of the present invention can provide glasses frame models according to pupillary distance conditions. The matching system can draw the selected spectacle frame model on the user's digital image in real time, and automatically adjust the size of the frame, the position of the nose pad or the length of the temples, etc. according to the user's face shape.

Regarding the features and implementation of the present invention, the preferred embodiments are described in detail below in conjunction with the drawings.

Description of drawings

The present invention will be further described in detail below in conjunction with the accompanying drawings and specific embodiments.

FIG. 1 is a schematic diagram of the hardware architecture of the present invention.

Fig. 2 is a schematic flow chart of an embodiment of the present invention.

Fig. 3 is a schematic diagram of eye height in the present invention.

FIG. 4 is a schematic diagram of a three-dimensional digital image of a user's head according to the present invention.

FIG. 5 is a schematic diagram of a digital image taken of a side face of a user according to the present invention.

FIG. 6 is a schematic diagram of image uploading in an embodiment of the present invention.

Fig. 7 is a schematic diagram of frame screening in an embodiment of the present invention.

Fig. 8 is a schematic diagram of a virtual try-on in an embodiment of the present invention.

FIG. 9 is a schematic diagram of shooting stereo head images in another embodiment of the present invention.

Description of main component symbols: 110: frame database; 111: spectacle frame model; 112: basic frame data; 120: computer device; 121: processing unit; 122: camera unit; 123: display unit; 124: communication unit; 125: Storage unit; 131: digital image; 132: glasses app; 30: face image; 301: eye height line; 510: object recognition; 520: user; 601, 602: ear; 611, 612: ear features Description point; 603: nose; 631, 632, 633: nose feature description points; 613: schematic line of browbone curvature; d: distance; 701: schematic line connecting ear and head; 702: schematic line on eyeball surface.

Detailed ways

Please refer to FIG. 1 , which is a schematic diagram of the architecture of the virtual glasses matching system of the present invention. The virtual glasses matching system of the present invention includes a frame database 110 and a computer device 120 . The frame database 110 is connected to the computer device 120 through the network. The frame database 110 stores multiple eyeglass frame models 111 and frame basic data 112 . The spectacle frame model 111 of the present invention can also be a three-dimensional spectacle frame image in addition to the planar spectacle frame figure. Each group of spectacle frame models 111 has its own preset frame basic data 112, which at least includes lens material, temple length, nose pad angle, nose pad width, frame width, and frame color. The computer device 120 of the present invention selects several candidate frames by comparing the frame basic data 112 with the frame screening range specified in the frame screening step, and then identifies the selected frames from the candidate frames according to the user's designation. The specified frame is automatically worn on the user's face image and adjusted accordingly, so that the user can evaluate and decide whether it meets his needs.

The computer device 120 has a processing unit 121 , a camera unit 122 , a display unit 123 , a communication unit 124 and a storage unit 125 . The processing unit 121 is electrically connected to the camera unit 122 , the display unit 123 , the communication unit 124 and the storage unit 125 . The type of the computer device 120 is not limited to a computer, and may also be an electronic device with the computing capability, such as a mobile phone or a tablet. The camera unit 122 is used for shooting a digital image 131 (digital image). The digital image 131 described in the present invention may be a plane two-dimensional image or a three-dimensional image of the user captured or recorded by the camera unit 122; images or stereoscopic images.

In addition to displaying the captured digital image 131 , the display unit 123 can also display an Augmented Reality (AR) image and manipulate objects in the Augmented Reality image. For the convenience of description, the display unit 123 of the present invention allows the user to view the screen and input operation commands through a touch display screen. For example, when the computer device 120 shoots a stereoscopic image, the user can view the object in the digital image 131 through the display unit 123 , or rotate the object in the digital image 131 . In the case of other applications, the computer device 120 can perform related operations through a human-machine interface (HID). The communication unit 124 is used to connect to the frame database 110 , and the communication unit 124 can be connected to the frame database 110 through a network or a cable.

The storage unit 125 stores the matching record history of the matching application program 132 and the user. The glasses application program 132 includes numerical data that can be used to identify the position of each facial feature such as eyes, nose, and ears in the user's digital image 131 and measure the distance, angle, etc. of the above-mentioned various facial features; . In addition to the functions of face recognition and automatic tracking, the glasses application program 132 also provides relevant interfaces for the user to select the spectacle frame model 111; including: an image transmission interface for receiving and transmitting the user's facial image, for accepting user A frame screening interface for selecting a plurality of candidate frames from the frame database based on the frame screening conditions, and a frame virtual try-on interface for sequentially performing virtual try-ons on the plurality of candidate frames. In order to further illustrate the recognition process of the facial features of various digital images 131 and the pairing process of the glasses frame model 111, please refer to FIG. 2, the virtual glasses matching method of the present invention includes the following steps:

Step S210: taking or obtaining a digital image of the user's face;

Step S220: Analyzing or measuring the digital image to obtain facial features;

Step S230: Screen out a plurality of candidate frames based on the facial features;

Step S240: The user performs virtual try-on on the plurality of candidate frames in order to select a suitable frame.

In this embodiment, through step S210, the plane image or stereoscopic image of the user's face can be further analyzed to obtain facial image feature data. The facial image feature data includes the interpupillary distance between the two eyes in the face, Any of facial features such as pupillary distance, face width, nose bridge height, nose bridge width, and eye-ear distance. First, for example, when a digital image of the user is acquired through imaging, the computer device 120 captures a digital image 131 of the front of the user. Then, the processing unit 121 runs the glasses application 132 , and loads the digital image 131 and displays the image on the display unit 123 . If the user has not recorded in the computer device 120 , the glasses matching application program 132 can be activated to identify the image features of the digital image 131 . Next, through step 220, the digital image of the user's face can be further analyzed or measured; in other words, the glasses application 132 can automatically locate the pupil, face width, and bridge of the nose from the digital image 131 or the user can choose Height, nose width, and eye-ear distance feature positions of each face, so as to obtain the data of each face feature of pupil distance, face width, nose height, nose width, and eye-ear distance.

In addition, another input interface will pop up on the display unit 123 for inputting the user's frame selection range. In the frame screening step, the glasses matching application program 132 can use at least two or more facial features included in the digital image 131 as a frame screening range to filter out a plurality of candidate frames.

For example, the optician application 132 may use, for example, Equation 2: M=180-Q=180-Tan ^-1 (X/(Y/2)) (where M is the nose pad angle (°); Y represents the bridge of the nose Width (mm); X represents nose bridge height (mm): Q represents nose bridge slope (°).); or for example using equation 3: R=Z+13+F (where R represents temple length (mm); Z Indicates the eye-ear distance (mm); F indicates the upper and lower limits (mm), generally between 0 and 40mm) to screen out a plurality of alternative frames.

In addition, the user can open the glasses frame model selection interface from the computer device 120, and specify the filter condition as pupil distance, face width, eye-ear distance, or nose bridge width. The processing unit 130 screens out frames meeting the filter conditions from the frame database 110 based on the filter conditions specified by the user and displays them on the selection interface. The user can view various spectacle frame models 111 from the selection interface and select a preferred style for virtual try-on. When the user selects any spectacle frame model 111, the processing unit 130 combines the spectacle frame model 111 with the digital image 131, and displays the combined picture on the virtual try-on interface for the user to view the try-on picture.

In this embodiment, the digital image 131 captured by the user is the user's face image 30 . In step S240 of virtual try-on, the processing unit 121 may further include performing a function of judging and determining whether the spectacle frame model 111 listed as an alternative is suitable according to the position of the bridge of the user's nose and the position of the nose pad of the spectacle frame model 111 . . In addition, the processing unit 121 may further draw corresponding eye height schematic lines according to the user's eyes, as shown in FIG. 3 . The processing unit 121 can also further judge and determine whether the spectacle frame model 111 listed as an alternative is suitable according to the aforementioned nose pad and eye height schematic line through the glasses application program 132 . Generally speaking, the optimal position of the spectacle frame model for the user's face is "a horizontal line tangent to the horizontal reference line of the spectacle frame with the lower eyelid of the user". This horizontal line can also be defined as the eye height schematic line 301 . In addition, the user can also move the position of the spectacle frame model 111 on the user's face, so as to observe and judge which one of the plurality of spectacle frame models 111 listed as options is the most suitable.

In addition to the aforementioned embodiments, the present invention can also be applied to the combination of the augmented reality between the user and the spectacle frame model 111 . The virtual glasses matching system in this embodiment includes a frame database 110 and a computer device 120 , please refer to FIG. 1 for the hardware structure and composition. The hardware architecture and component functions of this embodiment are the same as those in FIG. 1 , and the description of the hardware connection will not be repeated.

In addition to shooting a single image, the camera unit 122 of this embodiment can also continuously shoot a digital image 131 , so as to combine the user's stereoscopic image with the image of the spectacle frame model 111 . This implementation aspect includes wearing the recognition object on the user's face, and taking pictures of the user to create a digital image; recognizing the positions of the user's eyes, nose and ears in the stereoscopic image according to the position of the recognition object to obtain the face feature information; screening a plurality of candidate frames based on the facial features; and the user virtual try-on to select a suitable frame.

In this embodiment, firstly, the identification object 510 may be worn on the face of the user 520 . The identification object 510 can be glasses, a card or a detector. Generally speaking, the wearing position of the identification object 510 can be further set to different positions according to the developer. For example, the identification object 510 can be placed on the forehead, both sides of the cheeks or nose of the user 520 . Assume that the recognition object 510 is placed on the forehead of the user 520 . The glasses application 132 can determine the image scaling ratio of the identified object 510 in the digital image 131 according to the actual size of the identified object 510 . The processing unit 121 may further have a function of automatically recognizing the positions of the eyes, the nose and the ears of the user 520 according to the zoom ratio of the image. The glasses matching application 132 can select a plurality of candidate frames from the database according to the first facial features of the positions of the eyes, nose and ears as the frame screening range. In addition, the first facial feature may be the straight line distance (face width) formed by the connection between the two ears and the eyes in the front image of the user 520 .

In addition, in this embodiment, the computer device 120 can also use the computer device 120 to capture images obtained from other positions of the face of the user 520, and then obtain the distance and height of the forehead, both sides of the cheeks, or the tip of the nose from the images. Data such as angles are used as the first facial feature. For example, images of the left and right side faces or face shapes of the user 520 taken from a top view can be used. The present invention defines respective facial features according to shooting different facial positions.

In addition, in this embodiment, the second facial feature, the third facial feature, the fourth facial feature, or other features can also be used. Specifically, for example, the second facial feature may be the distance from the sectional plane of the eyes on the same side of the user 520 to the ears (ie, the eye-to-ear distance) when facing the front. The third facial feature may be the nose width of the inner distance between the two eyes when the user 520 is facing the front, and the nose height of the nose when the digital image 131 is on the side of the user 520 . The fourth facial feature may be a brow bone curvature of the eyes when the digital image 131 looks down on the user 520 .

FIG. 4 is a stereoscopic digital image 131 of the head of a user 520 . The stereoscopic digital image is obtained by scanning the head of the user 520 through a stereo camera to obtain the digital image 131 of the head of the user 520 . As shown in FIG. 4 , the user's head image 60 includes images of ears 601 , 602 and a nose 603 . In an embodiment, the ear feature description points 611, 612 can be used as the second facial features; the nose feature description points 631, 632, 633 can also be drawn according to the position of the nose 603 as the third facial features; The curvature radius or radian of the browbone line 613 is used as the fourth facial feature.

In addition, as shown in FIG. 5 , the distance between an ear-head connection schematic line 701 and the eyeball surface schematic line 702 in the digital image 131 obtained by shooting the side face of the user 520, that is, the distance between the ear and the eyeball surface can also be used. The distance d is used as the second facial feature. The second facial feature, the third facial feature, and the fourth facial feature used above are only for example, and can be changed and modified at will without departing from the spirit of the present invention. It is understood that these are also included within the scope of the present invention.

As shown in FIG. 4 , the computer device 120 captures the image features of eyes, nose, ears and head shape according to the digital image 131 obtained by the glasses application program 132. For example, when the face width and length are set to T mm and the nose bridge height is set to When X mm, the nose width is set to Y mm, and the vertical distance from the ear to the nose (eyeball surface) is set to Z mm; a large amount of facial feature data is obtained by the above-mentioned method, and after big data analysis is performed by mathematical fitting , the present inventors have found that there is a certain functional proportional relationship as shown in the following formula 1 between the frame width (P), nose bridge height (X), nose bridge width (Y), and eye-ear distance (Z):

P=T±5=-1.036X2 ⁺ 23.365X-2.089Y+Z±5...(Equation 1)

(Wherein, P is the frame width (mm);

T is the user's face width (mm);

X is the height of the bridge of the nose of the user (mm);

Y is the width of the bridge of the user's nose (mm);

Z is the user's eye-ear distance (mm). )

In addition, the inventor further measured a large number of spectacle frames and used mathematical fitting to carry out large data analysis results, and found that the face width (T), the inclination of the bridge of the nose (Q), the length of the temples (R), the width of the spectacle frame (P), nose pad angle (M), nose bridge height (X), nose bridge width (Y), and eye-ear distance (Z) have more than a certain degree of correlation. For example, there is a relationship between nose pad angle M and nose bridge inclination (Q) and face width as shown in the following formula 2:

Nose pad angle M=180-Q=180-Tan ^-1 (X/(Y/2))...(Equation 2)

(Wherein, M represents the nose pad angle (°)

Y represents the width of the bridge of the nose (mm);

X represents the height of the bridge of the nose (mm):

Q represents the inclination of the bridge of the nose (°). )

In addition, the inventors further measured a large number of spectacle frames and used mathematical fitting to carry out large data analysis results, and found that there is a relationship between the temple length (R) and the eye-ear distance (Z) as follows: The relationship shown:

R=Z+13+F...(Equation 3)

(Wherein, Z represents the eye-ear distance (mm);

R indicates the length of the temple (mm)

F represents the upper and lower limits (mm), generally in the range of 0 to 40mm. )

Then, the user can open the frame screening interface from the computer device 120, and specify the filter condition as pupil distance, face width, eye-ear distance, nose bridge width, or nose bridge height. The processing unit 130 screens out frames meeting the filter conditions from the frame database 110 based on the filter conditions specified by the user and displays them on the frame filter interface. The user can view various spectacle frame models 111 from the frame screening interface and select a preferred style for virtual try-on. When the user selects any spectacle frame model 111, the processing unit 130 combines the spectacle frame model 111 with the digital image 131, and displays the combined picture on the virtual try-on interface for the user to watch the try-on picture.

In addition, in another embodiment of the present invention, the user can further manually input filter conditions, including frame width, nose pad angle, temple length, color, or shape, so as to adjust the number of candidate frames.

When the user further selects the width of the picture frame as the filter condition, the face width plus an upper and lower limit range can be used as the correction value of the filter condition, for example, the face width ± 5mm is used as the filter condition of the picture frame; then, the processing unit 121 will A plurality of candidate frames are screened out according to the frame filter conditions set by the user.

The modification method of the box screening conditions illustrated above is only for example, and can be changed and modified at will without departing from the spirit of the present invention. Those who are familiar with this technology should understand that these are also included in the scope of the present invention. .

When the user inputs the angle of the nose pad as the filter condition, the processing unit 121 will filter out the corresponding frame according to Table 1 . In addition, the nose pad angle M can also be converted through the following formula 2.

M=180-Q=180-Tan ^-1 (X/(Y/2))...(Equation 2)

(Wherein, M represents the nose pad angle (°)

Y represents the width of the bridge of the nose (mm);

X represents the height of the bridge of the nose (mm):

Q represents the inclination of the bridge of the nose (°). )

Table 1

Remarks: Q indicates the inclination of the bridge of the nose (degrees); Y indicates the width of the bridge of the nose (mm); X indicates the height of the bridge of the nose

(mm); W means the width of the nose pads (mm); S means the length of the nose pads (mm)

In addition, when the user selects the length of the temples as the filter condition, the correction value of the length of the temples can be input; then, the processing unit 121 will filter out the mirror frames that meet the user's settings according to the following procedure 3:

R=Z+13+F...(Equation 3)

(Wherein, Z represents the eye-ear distance (mm);

R indicates the length of the temple (mm);

F represents the upper and lower limits (mm), generally in the range of 0 to 40mm. )

In the above implementations, the technology for performing facial recognition when shooting or scanning the facial image of the user 520 can refer to the inventor's previous filing: U.S. Patent Application No. 14/447652 (application date: July 31, 2014); and The technology of calculating the ratio of real pupil distance to virtual pupil distance through a known size reference object can refer to the patent application of the same inventor: US patent application No. 14/469569 (application date: August 26, 2014) and No. 14/447652. In addition, the technology of scanning fine facial images through the 3D lens of the stereo camera can refer to 3D maps and face algorithms, such as the inventor of this case US Patent Application No. 13/649170, Chinese Patent Application No. CN201310117970.5 etc. After the system matches the spectacle frame model 111 in the database, the user can directly use the computer device 120 to automatically obtain a real spectacle frame that fits his size and looks good.

In order to clearly illustrate the selection process of the present invention for users with different face shapes, the following examples 1-3 are used to illustrate the virtual glasses matching method of the present invention in detail.

Example 1

First of all, start the glasses application program of the present invention. Next, please refer to Figure 6, which is a schematic diagram of user image upload. User A can identify the source of the selected image in the target source: including 3D camera, 2D camera, and uploaded camera; or choose to import face information to use the previously analyzed face Import the internal information file into the program; in addition, you can also choose to input facial information to manually input the values of each feature of the face directly. In this embodiment, user A clicks the option of uploading video from the image transmission interface, then selects the path of the digital image file from the target source selection interface, and uploads the digital image file of the front and side of the face to the glasses In the system, the camera identification status in the image transmission interface shows success, and it can be seen that the optical matching application program of the present invention has obtained the facial feature data of the user as shown in Table 2 below.

Next, use the above-mentioned equation 1, equation 2, and equation 3 to calculate the frame width (mm), nose pad angle (degrees), temple length (mm), etc., and use them as the filter conditions A1 and A3 of the frame.

Table 2

P＝-1.036X2 ⁺ 23.365X-2.089Y+Z±5...(Equation 1)

M=(180-Q)=180-Tan ^-1 (X/(Y/2))…………(Equation 2)

R=Z+13+F(mm)…………………(Equation 3)

Next, please refer to FIG. 7 , which is a schematic diagram of frame screening; the face feature information of user A is displayed in the frame screening interface, including face width, nose bridge height, nose bridge width, and eye-ear distance. User A specifies the intersection of the above-mentioned frame filter conditions A1 and A3 in the identification information interface. For example, take the A1 frame width of 138.7mm as the filter condition of the frame, and then the application will automatically filter from the frame database, and a total of 318 pairs of candidate frames that meet the above conditions will be screened out; and 387 pairs will be screened out based on the length of the temples. . Spectacle frames whose temple length falls between 131mm and 141mm; the angle of the nose pad is between 47.82 degrees and 55.68 degrees, with an average of 53.26 degrees. It is confirmed to adjust the angle suitable for consumers when shipping. The number of intersections in A1 and A3 is 184, and consumers can try on the selected frames.

Next, please refer to FIG. 8 , which is a schematic diagram of a virtual try-on. User A selects a preferred frame style from the available frame options, and then the application program automatically combines the selected frame with user A's facial image. And the combination screen is displayed in the virtual try-on interface, so that user A can view the effect of the try-on from the virtual try-on interface.

User A repeatedly selects alternative frames and checks the try-on screen, then selects the frame whose appearance meets expectations, and selects the option "Yes" in the filter frame style on the right side of the virtual try-on interface, indicating that the frame has been selected ; Then the store will take out a matching frame entity from the warehouse according to the frame information selected by user A for user A to wear and confirm whether it meets the demand.

Example 2

First of all, start the glasses application program of the present invention. Similar to Embodiment 1, user B can identify the source of the selected image in the target source, including 3D camera, 2D camera, uploaded camera; or choose to import facial information to import the previously analyzed facial information file into the program ;In addition, you can also choose to input face information to manually input the values of each feature of the face directly. In this embodiment, user B selects 3D camera as the image source; please refer to FIG. After the stereoscopic head image of B is directly transmitted to the optical dispensing system via the wireless network, the camera identification status in the image transmission interface shows success, and it can be seen that the optical dispensing application program of the present invention has obtained user B as shown in the following table facial feature data.

Next, use the above-mentioned equation 1, equation 2, and equation 3 to calculate the frame width (mm), nose pad angle (degrees), temple length (mm), etc. and use them as the filter conditions B1 and B3 of the frame.

table 3

Next, specify the intersection of the above-mentioned frame filtering conditions B1 and B3 in the frame screening interface. For example, take B1 frame width (133.5mm) and B3 temple length (138mm) as frame filtering conditions, and then the application will automatically filter from the frame database, and screen out a total of 58 pairs of candidate frames that meet the above conditions. The frame width of the 58 pairs of glasses falls between 128.5mm and 138.5mm, with an average of 133.5mm; the length of the temples falls between 133mm and 143m, with an average of 138mm; the width of the nose pads falls between 11.5 and 15.5mm Between, the average value is 13.9mm.

Then, user B selects the preferred frame style from the available frame options, and then the application program automatically combines the selected frame with user B's head image, and displays the combined screen in the virtual try-on interface, so that User B checks the effect of the try-on from the virtual try-on interface.

User B repeatedly selects alternative frames and checks the try-on screen, selects a frame whose appearance meets expectations, and selects the option "Yes" in the filter frame style on the right side of the virtual try-on interface, indicating that the frame has been selected ; Then, the store will take out a matching frame entity from the warehouse according to the frame information selected by user B for user B to wear and confirm whether it meets the demand.

Example 3

Firstly, start the glasses application program of the present invention installed on the mobile device. Next, point the camera lens of the mobile device directly at user C, so that the head image of user C appears on the display screen of the mobile device. By turning the head of the user C in front of the camera lens, the following table 4 can be obtained. The facial feature data of user C.

Next, use the above-mentioned equation 1, equation 2, and equation 3 to calculate the frame width (mm), nose pad angle (degrees), temple length (mm), etc., and use them as the filter conditions C1 and C3 of the frame.

Table 4

Next, specify the intersection of the above-mentioned frame filter conditions C1 and C3 in the frame filter interface. For example, take C1 frame width (125.5mm), C3 temple length (125mm) or nose bridge width as the frame filter conditions, and then the application will automatically filter from the frame database, and screen out a total of 45 pairs of candidates that meet the above conditions picture frame. The frame width of the 45 pairs of glasses falls between 120.2 and 130.2mm, with an average of 125.2mm; the length of the temples falls between 120 and 130mm, with an average of 125mm; the inclination of the bridge of the nose falls between 27.4 and 33.8 degrees , with an average of 31.4 degrees.

Then, user C selects the preferred frame style from the available frame options, and then the application program automatically displays the selected frame in the virtual try-on interface, and user C can view his head through the camera lens of the mobile device. The images are synchronously displayed on the virtual try-on interface, and the position of the head image in the virtual try-on interface is adjusted by moving the head, so as to check the effect of the try-on.

User C repeatedly selects alternative frames and checks the try-on screen, then selects the frame whose appearance meets expectations, and selects the option "Yes" in the filter frame style on the right side of the virtual try-on interface, indicating that the frame has been selected ; Then the store will take out a matching frame entity from the warehouse according to the frame information selected by user C for user C to wear and confirm whether it meets the demand.

To sum up, according to the virtual glasses selection method and system described in the above embodiments, in addition to using the plane image to determine the pupil distance of the glasses frame model, the three-dimensional glasses simulation technology is also used. Select the frame type and frame-related parameters, and the system can provide the spectacle frame model according to the interpupillary distance. However, it can further obtain the distance from the ear to the surface of the eyeball according to the position of the eyeball and ear, and the height and width of the bridge of the nose according to the characteristics of the nose. Bone curvature allows users to automatically find glasses that match their size and preferences through computer devices, such as mobile devices, to achieve the purpose of choosing virtual glasses as if they were going to an optical company to get glasses in person.

Although the present invention is disclosed above with the aforementioned preferred embodiments, it is not intended to limit the present invention. Any person familiar with the similar art can make some changes and modifications without departing from the spirit and scope of the present invention. Therefore, The scope of patent protection of the present invention must be defined by the scope of patent application attached to this specification.

Claims (8)

1. a kind of virtual glasses matching method, which is characterized in that

Image obtaining step: shooting or the digital image for obtaining user face；

Face feature numerical value measurement step: parsing or measure the digital image and obtain the bridge of the nose height of user, nose bridge widths, And eye-ear distance from face feature data information；

Frame screening step: data information is sieved based on the face feature and is selected from frame data bank using following equation 1 A plurality of alternative frames out:

P=-1.036X²+ 23.365X-2.089Y+Z ± 5 (equation 1)

Wherein, P is frame width (mm)；

X is the bridge of the nose height (mm) of user；

Y is the nose bridge widths (mm) of user；

Z is the eye-ear distance of user from (mm)；And

Frame deciding step: user successively virtually tries a plurality of alternative frames on and selects suitable mirror Frame.

2. virtual glasses matching method according to claim 1, which is characterized in that further wrapped in frame screening step It includes using following equation 2 and selects a plurality of alternative frames from frame data bank:

M=180-Q=180-Tan^-1(X/ (Y/2)) (equation 2)

Wherein, M indicates nose support angle (°)；

Y indicates nose bridge widths (mm)；

X indicates bridge of the nose height (mm):

Q indicates bridge of the nose gradient (°).

3. virtual glasses matching method according to claim 1, which is characterized in that further wrapped in frame screening step It includes using following equation 3 and selects a plurality of alternative frames from frame data bank:

R=Z+13+F (mm) (equation 3)

Wherein, R indicates mirror pin length (mm)；

Z indicates eye-ear distance from (mm)；

F indicates upper lower limit value (mm), is typically in the range of the range of 0~40mm.

4. virtual glasses matching method according to claim 1, which is characterized in that work as user in frame deciding step Then include using following equation 2 in the rear of frame deciding step when failing to select the frame for being suitble to wearing, further to Selected a plurality of alternative frames are screened in the frame screening step:

M=180-Q=180-Tan^-1(X/ (Y/2)) (equation 2)

Wherein, M indicates nose support angle (°)；

Y indicates nose bridge widths (mm)；

X indicates bridge of the nose height (mm):

Q indicates bridge of the nose gradient (°).

5. virtual glasses matching method according to claim 1, which is characterized in that work as user in frame deciding step Then include using following equation 3 in the rear of frame deciding step when failing to select the frame for being suitble to wearing, further to Selected a plurality of alternative frames are screened in the frame screening step:

R=Z+13+F (equation 3)

Wherein, R indicates mirror pin length (mm)；

Z indicates eye-ear distance from (mm)；

F indicates upper lower limit value (mm), is typically in the range of the range of 0~40mm.

6. virtual glasses matching method according to claim 5, which is characterized in that the face feature data information is into one Step includes wide interpupillary distance, face, nose support distance degree, mirror pin length, at least any one in shape and color.

7. a kind of virtual glasses match system, which is characterized in that

Frame data bank, for storing plural pair of glasses frame model；And

Computer apparatus, the computer apparatus have processing unit, camera unit, display unit, communication unit and storage element, institute It states processing unit and is electrically connected at the camera unit, display unit, the communication unit and the storage element, the communication Unit network is connected to the frame data bank, and mirror application is matched in the storage element storage one, and the processing unit executes It is described to match mirror application；Wherein it is described be with mirror application include: image of face to receive and transmit user Image transmission interface, the frame screening conditions to receive user are simultaneously based on the frame screening conditions from frame data bank Select the mirror that the frame of a plurality of alternative frames screens interface and successively virtually tried on for a plurality of alternative frames Frame virtually tries interface on；The processing unit is based on the face feature and filters out a plurality of alternative mirrors from frame data bank Frame；The display unit shows the alternative frame and in conjunction with the digital image of the user, for user's selection and Determine suitable frame.

8. virtual glasses according to claim 7 match system, which is characterized in that the digital image is single via camera shooting Member shooting is sent to the processing unit via the communication unit and obtains.