CN1447955A - Appts. and method for generating synthetic face image based on shape information about face image - Google Patents
Appts. and method for generating synthetic face image based on shape information about face image Download PDFInfo
- Publication number
- CN1447955A CN1447955A CN01814412A CN01814412A CN1447955A CN 1447955 A CN1447955 A CN 1447955A CN 01814412 A CN01814412 A CN 01814412A CN 01814412 A CN01814412 A CN 01814412A CN 1447955 A CN1447955 A CN 1447955A
- Authority
- CN
- China
- Prior art keywords
- image
- information
- texture
- shape information
- face
- 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
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T11/00—2D [Two Dimensional] image generation
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Image Processing (AREA)
- Processing Or Creating Images (AREA)
- Processing Of Color Television Signals (AREA)
Abstract
An apparatus and method for generating a synthetic face image based on shape information about an input face image ransmitted from a user interfacing device are provided. The apparatus extracts shape information, which is represented by a deformation field with respect to a predetermined reference image, and texture information, which is color or brightness information about an input image mapped on the reference image, about the input face image from face image information transmitted from a user interfacing device, and transforms a variety of face images, which are stored in an image database in advance and have the same shape as that of the reference image, using the shape information about the input face image according to a user control command, thereby synthesizing face images reflecting the shape information about the input face image. Accordingly, natural new images reflecting the shape of the input face image and having high quality regardless of the state of the input face image can be synthesized by using images, which have the same shape information as that about the reference image but have different types of texture information, and the extracted shape information about the input face image.
Description
Technical field
The present invention is relevant with a kind of apparatus and method that are used to generate synthetic face-image, and especially, the apparatus and method that generate synthetic face-image with a kind of shape information according to input face image are relevant.
Face-image is often used as a kind of intermediary that can fully show individual character and allow unsophisticated interchange.This face-image can be applicable to, such as gateway, criminal's search/picture synthesis system, computer interface, cartoon and the recreation of control/security system.Adopting typical case's application of synthetic face-image technology is the generation and the Cosmetic Design of individual character image.
The caricature of face-image as a kind of individual character image, is that the face feature according to a people generates.Therefore, the caricature of a face-image not only can be used for the manufacturing of cartoon or entertainment, represents a people's symbol or button, and a kind of peculiar symbol of PC communication or E-mail also can be used as the incarnation that embodies a user in the virtual reality.
Background technology
Traditionally, professional artist adopts the method for manual drawing caricature, perhaps uses digital screening sequence to handle automatically, carries out the caricature creation.Wherein, the method that adopts digital screening sequence to handle image produces a kind of effect of complete manual initiative caricature to input picture, this effect is to adopt screening sequence that input picture is produced suitable effect by combination, and realizes that watercolor is described or the effect of charcoal drawing.
When professional artist's manual drawing one width of cloth caricature, can obtain width of cloth nature and very perfect caricature, but this method to spend long time and be hard to keep consistent effect.Therefore, this method is confined to specific condition.The treatment technology of numeral screening sequence can be used for the image acquisition under light and the background condition of limited, like this for the simple two-dimensional image, wherein object is difficult to come out from background separation, and the picture quality of output depends on the light in the environment or the change of other conditions.Also do not propose to implement the suitable method of this technology at present.In addition, because the shape information about object in traditional caricature method for making is not independent the generation, carry out a kind of correction operation, for example enlarge the facial characteristics of the caricature that a width of cloth generated, perhaps change a kind of facial expression, be complexity very, and want to reduce a breadth portion image or the caricature that will generate expands to three-dimensional form and almost can not accomplish.
In Cosmetic Design, the user determines a style indirectly according to the model of having made up traditionally.Recently, introduced a kind of cosmetic method that utilizes computing machine.This prescribing method allows the user that product is applied to various sample pattern images in every way, but the natural effect of directly user's face-image being made up can not be provided.In other words, even adopt identical cosmetics, color also may be with various and complicated condition, such as the shade and the reflection ray that are thrown light on and the characteristics of facial shape influence, and change, therefore can not from the dressing effect of a sample pattern image, obtain the dressing effect of user's face-image of nature.
Summary of the invention
For addressing the above problem, the present invention aims to provide the apparatus and method that a kind of shape information according to face-image generates synthetic face-image, wherein information is by extracting in the face-image of importing about the form of face, and synthetic face is to generate according to the shape information that is extracted, the user can obtain a nature and exquisite cartoon image like this, utilize his/her face-image to watch the Cosmetic Design image in advance, be easy to add various annex to composograph, be easy to change composograph, and the effect of annex that Real Time Observation adds and change composograph.
For reaching target of the present invention, the invention provides the device that a kind of face-image according to an input synthesizes a new face-image.This device comprises a user's interface device, is used to receive facial image information and user's control commands, these command transfer is received to image processing apparatus, from image processing apparatus synthesize facial image information and export or store synthetic facial image information according to user control command; With an image processing apparatus, the facial image information that is used for transmitting from user's interface device is extracted shape information and the texture information about input face image, its shape information is expressed as a picture about predetermined reference anamorphose, its texture information is about the light that is plotted in the input picture on the reference picture and the information of color, and utilize the shape information of input face image according to user command, the width of cloth texture image that conversion is selected from the texture image, in the image data base that this image is stored in and with the homomorphosis of reference picture, perhaps conversion is weighted the image that summation generates by the texture image to selected texture image and the reaction texture information of extracting, thereby generates a synthetic face-image.
The present invention also provides a kind of method of synthesizing a new face-image according to the shape information of the face-image of importing.This method comprises the information following steps: (a) extraction from the facial image information of input is about the shape information and the texture information of input face image, its shape information is expressed as a picture about predetermined reference anamorphose, and its texture information is about the light that is plotted in the input picture on the reference picture and the information of color; (b) utilize the shape information of input face image according to user command, the width of cloth texture image that conversion is selected from the texture image, in the image data base that this image is stored in and with the homomorphosis of reference picture, perhaps conversion is weighted the image that summation generates by the texture image to selected texture image and the reaction texture information of extracting, thereby generates a synthetic face-image.
Description of drawings:
Figure 1A is depicted as the functional configuration block diagram of synthetic face-image generating apparatus first embodiment of the present invention;
Figure 1B is depicted as the functional configuration block diagram of synthetic face-image generating apparatus second embodiment of the present invention;
Figure 2 shows that the mechanical arrangements block diagram of the computer system of implementing the present invention first and second embodiment;
Figure 3 shows that the basic flow sheet of the synthetic face-image process of generation of the present invention;
Figure 4 shows that the particular flow sheet of facial information leaching process as shown in Figure 3;
Figure 5 shows that the face-image synthetic method that is adopted according to the synthetic face-image device of generation of the present invention generates the process flow diagram of a width of cloth cartoon image process;
The sample image replacement method that device adopted according to the synthetic face-image of generation of the present invention that Figure 6 shows that generates the process flow diagram of a width of cloth cartoon image process;
Figure 7 shows that the present invention generates the process flow diagram of the Cosmetic Design process that synthetic face-image device adopted.
Embodiment:
Below with reference to accompanying drawing embodiments of the invention have been carried out specific description.
Figure 1A and Figure 1B are respectively that the shape information according to face-image of the present invention synthesizes first embodiment 1 of face-image generating apparatus and the block diagram of second embodiment 40.First embodiment as shown in Figure 1, comprises one or more user's interface device 10a and 10b, and therefore a communication network 20, and an image processing apparatus 30 operate in the network environment.Therefore second embodiment 40 as shown in Figure 2, comprises a user's interface device 50 and an image processing apparatus 60, operates in one independently in the computer system.
User's interface device 10a and the 10b of first embodiment 1 and second embodiment 40, be embodied as a computer system 70 with image processing apparatus 30, this system comprises a computing machine 72, as shown in Figure 2, have at least one CPU (central processing unit) 74 and an internal storage location 73, an input media 75, and an output unit 76.The contact that the primary clustering of computer system 70 is realized each other by one or more bus structure 77.
CPU74 comprises an ALU (ALU) 741 carrying out an arithmetical operation and a logical operation, and a registers group 742 is in order to temporary storaging data and order, and control module 743 is with the operation of control computer system 70.CPU74 of the present invention is not limited to by special company's manufacturing specific device, and can be the processor with any money of above basic structure.
Second example 40 shown in first embodiment shown in Figure 1A and Figure 1B carries out having identical functions the communication processor 14 and 31 of data transmission except that first embodiment 1 further comprises by communication network.Therefore, the present invention will be embodied as the basis with first and describe.
With reference to Figure 1A, each user's interface device 10a and 10b receive facial image information and the user control command from the user, receive the composograph corresponding to user control command, and revise, store or the output composograph.Each user's interface device 10a and 10b comprise 12, one i/o controllers 13 of 11, one user command input blocks of an image information input block, a communication processor 14,15, one image storage unit 16 in an image correction unit, and an output unit 17.
Image information input block 11 is received from user's a facial image information, and can realize by the mode such as scanner or digital camera.In addition, image information input block 11 can comprise some cameras, to receive image and camera servicing unit of being gathered with different angles, such as a light regulator.Because image information input block 11 is as an assembly of the present invention, should consider from functional perspective, so image information input block 11 should be interpreted as not only comprising input media 75 as shown in Figure 2 widely, and comprise that storing the auxilliary of facial image information as shown in Figure 2 in advance deposits 732.
User command input block 12 is received from user's user control command, (for example, user profile, the synthetic control signal of facial information, and a facial Correction and Control signal).User command input block 12 can be implemented as one and specifically installs, and such as keyboard, mouse or touch screen, the user can import an order or information by it.
I/o controller 13 control facial image information are by the input of image information input block 11, also control the input of user command, and control transfers to image processing apparatus 30 by communication processor 14 with user command by user command input block 12.I/o controller 13 also receives the new image information that image processing apparatus 30 is synthesized according to user control command, and the correction of control new image information, storage or output.
Communication processor 14 links to each other with i/o controller 13, and transmits data to or receive data from image processing apparatus 30 by communication network 20.For example, communication processor 14 can be implemented as a device, this device comprises a serial/parallel port, USB (universal serial bus) (USB) port or an IEEE1394 port that connects Internet, and an Ethernet network interface card comes to comprise data of image information by Internet transmission and reception.
Image correction unit 15 links to each other with i/o controller 13, and according to the user control command of being imported by user command input block 12, synthetic or image transmitted information is revised angle, size or the texture of image according to image processing apparatus 30.
Image storage unit 16 deposits 732 corresponding to as shown in Figure 2 auxilliary, and storage is synthetic by image processing apparatus 30 and the new image information of transmission, perhaps stores the image information of being revised under i/o controller 13 controls by image correction unit 15.
Output unit 17 is corresponding with output unit 76 as shown in Figure 2.Under the control of i/o controller, output unit 17 shows when image processing apparatus 30 synthesizes a new image desired, be used to receive the user interface display message of user control command, and show or print new by image processing apparatus 30 synthesize with or institute's image transmitted information, the perhaps image information of being revised by image correction unit 15.
The first embodiment of the present invention shown in Figure 1A, communication network 20 are to transmit data between user's interface device 10a and 10b and the image processing apparatus 30 at one or more.Communication network 20 can be implemented as various types of networks, wire/wireless Internet for example, LAN and private line.
In the first embodiment of the present invention 1 as shown in Figure 1, image processing apparatus 30 is handled from one or more user's interface device 10a and 10b institute image transmitted information, synthesize a new images corresponding to user control command according to image information, and synthetic image is transmitted the user's interface device of a correspondence.Image processing apparatus 30 comprises a communication processor 31, image processor 32, and an image data base (DB) 33.
Communication processor 31 transfers data to and receives data from one or more user's interface device 10a and 10b.Communication processor 14 corresponding to each user's interface device 10a and 10b, communication processor 31 can be implemented as a device, comprise an Ethernet network interface card, a serial/parallel capable port, a USB (universal serial bus) (USB) port, perhaps an IEEE1394 port that connects Internet comprises data of image information by Internet transmission and reception.
Image processor 32 is from the facial image information that user's interface device 10a and 10b are transmitted, extraction is about the shape information and the texture information of an input face image, its shape information shows as the anamorphic image to reference picture, its texture information, promptly be plotted in the color or the monochrome information of the input face image on the reference picture, analysis from the user control command of user's interface device 10a and 10b with the interpreting user request, according to the customer requirements of being explained, the shape information that utilization is extracted, the texture information of extracting and be stored in various images among the image DB32, the synthetic new face-image of a width of cloth.Image processor 32 comprises 322, one parts of images substituting units 323 of 321, one face-image compositors of a facial information extraction apparatus, and an annex map picture increases unit 324.
Facial information extraction apparatus 321 is from the facial image information that user's interface device 10a and 10b are transmitted, extraction is about the shape information and the texture information of an input face image, its shape information shows as the anamorphic image to reference picture, its texture information promptly utilizes shape information to be plotted in the color or the monochrome information of the input face image on the reference picture.
Face-image compositor 322 is according to user control command, the shape information that utilizes facial information extraction apparatus 321 to be extracted about input picture, a selected width of cloth texture image will carry out conversion in the texture image among the image DB33 by being stored in, perhaps carry out image transformation, this image is that the image by the reaction texture information texture of being extracted to the texture image selected with by information extractor 321 is weighted summation and generates, and synthesizes a new face-image.
Parts of images is replaced the new face-image that unit 323 is replaced partly or completely, and this new face-image is synthetic by face-image compositor 322, and its sample image has and the highest similarity of sample image that is stored among the image DB35.
The face-images that annex map is synthesized by face-image compositor 322 as 324 pairs of adding devices add the image of an annex, and this image is to be to choose in the image appended drawings picture that is stored among the image DB32 according to user control command.
Image DB33 in advance in image processor 32, the image information that the stores processor input picture is required.Image DB33 comprises a mask DB331, and an annex map is as DB332, a sample image DB333, and a cosmetic image DB334, and an annex map is as DB335.
Mask DB331 stores various types of information (form mean value, texture mean value, morphological feature vector, and texture proper vector, these information obtain with reference to various face-image models), this information is used for according to a reference picture, the input face image in a facial information extraction apparatus 321, extract shape information and texture information.The various types of information that are stored among the mask DB331 will give more careful explanation when key diagram 4.
Appended drawings is as the information of DB332 storage about cartoon image, and this cartoon image has identical form with reference picture, but different type texture information, such as animation style, literary sketch style and watercolor print style.
The information of the relevant various caricature sample images of sample image DB333 storage, this information comprises the change of the form and the facial expression at each concrete position on the face-image.
The information that cosmetic image DB334 stores relevant Cosmetic Design image, this cosmetic image has the form identical with reference picture, and expresses with the texture information of representing various samples cosmetics.
Annex map is stored about being added to the information on the synthetic face-image, such as glasses, hair style, cap, ear and body image as DB335.
In first embodiment of the present invention shown in Figure 1A, one or more user's interface device 10a and 10b are connected on the image processing apparatus 30 by communication processor 14 and 31 and communication network 20.Perhaps, the second embodiment of the present invention shown in Figure 1B, user's interface device 50 and image processing apparatus 60 can be combined as a whole and operate in the independent computer system 70.
Shape information according to face-image of the present invention generates the device 1 of synthetic face-image and 40 basic operation will be illustrated in conjunction with Fig. 3.
In step S10, image processor 32 or 62 facial information extraction apparatus 321 or 621, reception is from user's interface device 10a, 10b or an input face image of 50, and in step S11, extract the shape information and the texture information that is plotted in the input face image on the reference picture, i.e. color or monochrome information of input face image according to the predetermined reference image.
In step S12, image processor 32 or 62 face-image compositor 322 or 622, according to a user control command that is received (a synthetic control signal of face-image) from user's interface device 10a, 10b or 50, utilization by the facial information extraction apparatus extracted based on the texture information of reference picture and the shape information of input face image, synthetic new face-image.In other words, the form of the face-image that the shape information that face-image compositor 322 or 622 utilizations are extracted from input face image recovers to import, and the form that the texture information that utilization has been extracted is twisted the input face image that is recovered, thereby a synthetic user's face-image.Wherein, by the texture information of suitable change or replacement reference picture, this texture information can be synthesized the new face-image with different input face image forms in order to synthetic face-image.
Synthetic face-image is transferred to user's interface device 10a, 10b or 50, and show by output unit 17 or 57.In step S13, user's interface device 10a, 10b or 50 user command input block 12 or 52 receive a user control command from the user, and this order is used to represent whether to change the shape information of the face-image that has shown.
When the user control command indication changes shape information, the shape information of input face image according to user control command (as, the control signal of an indicating section conversion, such as by pull the concrete part that this part enlarges or reduce the face-image of a demonstration with mouse, perhaps indicate the control signal that thoroughly changes for one, such as utilizing slide block to enlarge entire face) and change, and operation is got back to step S12 with synthetic new face-image.
When the user does not want to change shape information in the S13 step, in step S14, annex map as adding device 324 or 624 according to a user control command that adds, the face-image that is synthesized in step S12 is added the various annex map pictures that are stored in image DB33 or 63, perhaps parts of images is replaced unit 323 or 623, utilization is stored in the various sample images in image DB33 or 63, the special part of the face-image that replacement is synthesized in step S12, thereby the face-image that is synthesized added various extra effects.
Therefore, can be transferred to user's interface device 10a, 10b or 50 and be shown to the user by image processing apparatus 30 or 60 face-images that synthesized.At step S15, image correction unit 15 or 55 by user command input block 12 or 52 control commands that received (an image correction control signal), is revised synthetic face-image according to the user at last.The synthetic face-image of being revised by image correction unit 15 or 55 is stored in image storage unit 16 or 56, perhaps at step S16 by unit 17 or 57 and show or print.
Facial information extraction step S11 is as shown in Figure 3 being reduced acquisition shape information S from the input face image of known mask
InWith texture information T
InA program.
Among the present invention, the shape information of face-image is expressed as the anamorphic image with respect to reference picture, and the texture information representation of face-image is to be plotted in the color of the input picture on the reference picture or monochrome information.In other words, the shape information S of face-image is defined as every bit p on the reference picture
i(i=1 ..., n; N is the predetermined number of points on the reference picture) and face-image on distance between each corresponding point.The texture information T of face-image may be defined as every bit on the input picture with respect to each corresponding point p on the reference picture
i(i=1 ..., color n) or brightness value.When a composograph that obtains when utilizing form mean value and texture mean value is used to the reference picture of one embodiment of the present of invention, can be used for reference picture of the present invention and be not limited to above embodiment.All can be used as reference picture one of in any pre-prepd m image.
The mask that is stored in mask DB331 or 631 obtains in advance according to following program.At first, shape information S
j(j=1 ..., m) and texture information T
j(j=1 ..., be from pre-prepd m face-image, to extract to obtain m) based on reference picture.Thus, can be by every bit p
i(i=1 ..., shape information S n)
j(j=1 ..., the form mean value S that mean value constituted m), and by every bit p
i(=1 ..., n) the texture mean value T that constituted of the texture mean value on.Can obtain poor morphology like this
Covariance, and texture is poor
Covariance.
Do principal component (leading indicator) analysis by the numerical value that these are obtained, and obtain the morphological feature vector S of the covariance of m mask
i(i=1 ..., m-1), and texture proper vector t
i(i=1 ..., m-1).Therefore, according to morphological feature vector S
i(i=1 ..., m-1), and texture proper vector t
i(i=1 ..., m-1) face-image can be expressed with equation (1)
(wherein,
And m is the mask number)
By this program, form mean value S, texture mean value T, morphological feature vector S
i(i=1 ..., m-1), and texture proper vector t
i(i=1 ..., m-1) be stored in mask DB331 or 631, and be used to import the facial shape information and the extraction of texture information.
Facial information extraction step S11 can obtain more careful explanation with reference to figure 4 as shown in Figure 3.In step S111, as shown in Figure 4, input face image is normalized.More specifically, predetermined unique point (for example, the central point of the central point of each pupil and a lip) from an input face image, extracts, input face image is moved by upper and lower, left and right then, and regulate its size, so that the position of the extract minutiae of input face image is on the unique point of corresponding reference picture.This image normalization step can be carried out by predetermined software, perhaps can be according to manually being carried out by a control command that the user imported.Because its concrete analysis program has exceeded scope of the present invention, omits its explanation.
Step S112 has estimated shape information.More particularly, carry out and reference picture (perhaps synthetic texture estimation image T go on foot resulting input picture at S111 with form identical with the synthetic texture estimation of reference picture image
In_k) use a kind of layering, gradient optical flow algorithm (Lucas and Kanade) is handled, thus according to reference picture estimation shape information S
In_pre_l(put on normalized input face image and the corresponding reference picture between distance and position difference).By adopting layering, the correspondence between gradient optical flow algorithm, two similar images can utilize a kind of light stream of two image intensities to express.
According to layering, the gradient optical flow algorithm is at the resulting shape information S of step S112
In_pre_lAn error amount that comprises the light of input picture or shade and cause.Therefore, in step S113, to shape information S
In_pre_lCompensate.More specifically, by thereupon to morphological feature vector S
i(i=1 ..., m-1) carry out that linearity is decomposed and to shape information S
In_pre_lCarry out linear superposition and realize compensation, thereby obtain error compensation shape information S error amount
In_correct_lWherein, for increasing horizontal degree of freedom, the estimation shape information S that preferentially step S112 is obtained
In_pre_lAnd the compensation shape information S that obtains of step S113
In_correct_lWeighted sum S
In_lAs the result of shape information, as its computing method shown in equation (2):
S
in_l＝W·S
in_pre_l+(1-W)·S
in_correct_l(2)
(wherein, 0≤W≤1)
In step S114, carried out inverse transformation.Especially especially, utilize the shape information S that in step S113, obtains
In_l, input face image is transformed into reference picture.This process is called " inverse transformation ".
At step S115 according to texture proper vector t
i(i=1 ..., m-1), the texture information of reciprocal transformation image having carried out linear decomposition, the result that linearity is decomposed has carried out linear superposition then, thereby obtains the texture information T of input face image
In_l
Thus, the resulting normalized input face image of step S112, the input picture that is converted to reference picture by step S114 is replaced, and reference picture is replaced by the texture information that has a same modality with reference picture.Then, repeat S112 and go on foot the step, thereby obtain S to S116
In_k(k=1 ...).In other words, when repeating for the k time, the S among step S112 and the S113
In_l, S
In_pre_l, and S
In_correct_lRespectively by S
In_k, S
In_pre_k, and S
In_correct_kReplace, and the S among the step S114
In_lQuilt
Replace.In addition, the T of S115 in the step
In_lBy T
In_kReplace.The T that last circulation obtains
In_lBecome the final texture information T of input facial information
InThis circulation repeats, up to S
In_kVectorial mould ‖ S
In_kTherefore ‖ perhaps is equal to, or greater than predetermined number up to repeat number less than the predetermined lower limit value, obtains the shape information S of the input image information that calculates according to reference picture in the step at S117
In
Therefore, can utilize according to shape information S with reference to the input face image that then calculates
InAnd texture information T
In, and the storage input face image.In other words, the texture information T of the input face image that calculates according to reference picture
In, according to the shape information S of the input face image that calculates by reference picture
InAnd conversion, thereby synthetic input face image.
The exemplary embodiments of the characteristics of this composograph is that the caricature of face-image generates and Cosmetic Design.The caricature generation method of face-image can be divided into face-image synthetic method and sample image replacement method.
Adopted corresponding to the present invention synthetic face- image generating apparatus 1 and 40, utilize the face-image synthetic method, the program that generates cartoon image according to the shape information of face-image will give in explanation with reference to figure 5.
In the S20 step, facial processor 32 or 62 facial information extraction apparatus 321 or 621 are from user's interface device 10a, and 10b or 50 places receive an input face image, and extract input face image shape information S at step S21 according to the predetermined reference image
InWith texture information T
In
Thereby, face-image compositor 322 or 622 is by user's interface device 10a, 10b or 50, to the user present be stored in the different-style in image DB332 or 632 cartoon image (for example, an animation style, a kind of literary sketch style, a kind of watercolor print style), the user can go on foot the caricature style that selection is liked at S22 like this.Wherein, be stored in appended drawings and have the form identical with reference picture as the cartoon image of the different-style in DB332 or 632.
In step S23, face-image compositor 322 or 622 utilizes the shape information S of input face image
In, cartoon image or an image of synthetic selected style, this image are by the texture information T to selected cartoon image and reaction input face image
InImage be weighted summation and generate, thereby generate the cartoon image of reaction user facial shape information.
In step S24, synthetic cartoon image is transferred to user's interface device 10a, 10b or 50 and show by output unit 17 or 57, and user's interface device 10a, 10b or 50 user command input block 12 or 52 receive a user control command, and this order is used to indicate whether to change the shape information of shown user's cartoon image.
When the user control command indication changes shape information, shape information S
InAccording to user control command (for example,, the control signal that changes of indicating section, as by pull the special part that this part enlarges or reduce the face-image as shown in mouse, perhaps whole change of control signal indication, enlarge entire face such as utilizing slide block) and change, and operation is got back to the S22 step with synthetic new cartoon image.
In step, corresponding with an extra user control command, annex map is as adding device 324 or 624 at S25, various annex map picture (glasses, hair style, the caps of retrieve stored in image DB335 or 635, ear and body), and these annexes are added on the cartoon image.When annex map as adding device 324 or 624 add an annex map as the time, utilize the shape information S that extracts in the S21 step
In, automatically perform size or position adjustments and obtain more natural result.In addition, parts of images is replaced the sample image that unit 323 or 623 utilizes retrieval in sample image DB333 or 633, replaces a concrete part in the cartoon image, thereby makes cartoon image give expression to happiness, sad or angry expression.In addition, by adopting animated frame to express the effect that facial change can reach mobile image.
Thereby, by image processing apparatus 30 or 60 and synthetic animated image is transmitted user's interface device 10a, 10b or 50 and be shown to the user.In the S15 step, image correction unit 15 or 55, final according to control input block 12 or 52 user control commands that receive (an image check and correction control signal) by the user, revise synthetic cartoon image.In S16 step, by image correction unit 15 or 55 and the synthetic cartoon image revised is stored in image storage unit 16 or 56, perhaps by output unit 17 or 57 and show or print.
The cartoon image that obtains can be applied to special purpose immediately by this method.Perhaps, cartoon image can be used as a sketch when an artist makes a width of cloth caricature, thereby has increased the throughput rate of manual process.
The synthetic face-image generating apparatus 1 of the shape information based on facial information of the present invention and 40 employings utilize sample replacement method to generate the program of caricature, will be illustrated with reference to figure 6.
Fig. 6 further comprises step S35 and S36 except step as shown in Figure 5.Therefore, the explanation to S34 and S37 to S39 will be omitted for step S30.In other words, method as shown in Figure 6 is a sample image that utilization is stored in sample image DB333 or 633, and replaces according to as shown in Figure 5 method and the part of synthetic cartoon image or the method for integral part.
The sample image that is stored in sample image DB333 or 633 is to calculate by the result that different facial shape informations are analyzed.Sample image information formation method can be divided into two kinds of situations: a kind of situation is that the distortion of its sample information allows, and another kind of situation is that the distortion of its sample information is unallowed.
When the sample information method of formationing of the distortion situation that adopt to allow its sample information, formed a normalized sample image, and the size of this sample image and form are according to going on foot the shape information S that extracts at S31 from input face image
InAnd distortion.The advantage of this method is cause reacting the form and the needs sample image in a small amount of input face image, but its shortcoming to be to have twisted sample information by distortion, and to have worsened the quality of whole picture.
When adopting the sample image formation method that does not allow under its sample information distortion situation, the partly or completely part of cartoon image is only replaced by a sample image of making in advance, and synthetic new cartoon image.Therefore, the advantage of this method is to reach high resolution (sharpness), but shortcoming is to be difficult to the form of input picture that induces reaction, and need prepare sample image as much as possible in advance.
The method of the measurement difference D that step S35 adopted as shown in Figure 6 can be expressed with equation (3).
Wherein, 0≤W≤1, C
Si(i=1 ..., n) be the shape information of input picture, C
Ri(i=1 ..., n) be the shape information of sample image, C
Qi(i=1 ..., n) be the poor T of the texture information and the reference picture texture information of input picture
Ref, and C
Ti(i=1 ..., n) be the poor T of the texture information and the reference picture texture information of sample image
RefNot immediately shape information and texture information to be applied to C
Si, C
Ri, C
Qi, and C
Ti, but adopted by form and texture information being carried out the coefficient of the proper vector that linear analysis obtains, shown in equation (1).Wherein, each system has (m-1) dimension.
Partial information replacement unit 323 or 623 measure the poor of input pictures and each sample image at S35 in step, and adopt the sample image of lowest difference to replace cartoon image partly or completely in the step at S36.
When cartoon image is when generating according to sample image replacement method, it or not the cartoon image that is generated, but be used to replace the coding of the sample image of some or all of cartoon image, be compressed and in the communication environment of low speed, be transmitted, thereby significantly increase compressibility.
Illustrated what synthetic face-image generating apparatus 1 of the present invention and 40 was adopted in conjunction with Fig. 7, designed the program of cosmetic according to the shape information of facial information.
Remove Fig. 7 and comprise step S42, rather than beyond the S22 of Fig. 5 went on foot to S24, Fig. 7 was identical with Fig. 5 to S45.Therefore, omit to step S40 the explanation of S41 and S46.
In S42 goes on foot, face-image compositor 322 or 622 has presented the various sample cosmetic images that are stored in cosmetic image DB334 or 634 by user's interface device 10a, 10b or 50 to the user, and the user can select a sample cosmetic image of liking like this.Wherein, sample cosmetic image has identical form with reference picture.
In step S43, the shape information S that face-image compositor 322 or 622 utilizes S41 to extract from input face image information
In, with selected sample cosmetic image or an anamorphose, this image be by to selected sample cosmetic image and in S41 step from reaction texture information T that input face image extracted
InImage be weighted summation and generate, thereby synthetic face-image, wherein the face-image to the user has used selected sample cosmetic image.
In step S44, the synthetic face-image of reaction sample cosmetic image is transferred to user's interface device 10a, 10b or 50 and show by output unit 17 or 57, and the user command input block 12 or 52 of user's interface device 10a, 10b or 50, receive a user control command, this order indication is revised shown user's face-image.Face-image compositor 322 or 622, according to the face-image of user control command correction reaction sample combination image, and the face-image of revising is transferred to user's interface device 10a, 10b or 50, and shows by output unit 17 or 57.
Therefore, in step S45, user's interface device 10a, 10b, perhaps 50 user command input block 12 or 52 receptions are from user's user control commands, and whether this order indication user is satisfied with to shown face-image.Wherein, when user control command was represented " satisfaction ", program proceeded to the step S46 that is used to add annex.Otherwise when user control command was represented " being unsatisfied with ", program was got back to step S42.
Owing to the present invention carries out fully open and explanation in conjunction with its preferred embodiment, those skilled in the art should be clear and definite, under the situation that does not deviate from the spirit and scope of the present invention, and can be to the change of the present invention to various forms and details.More than the embodiment that is adopted in the explanation is in order to illustrate rather than to limit.Therefore, scope of the present invention will limit by the claim of being added rather than by above explanation, should be understood that any improvement within the scope of the invention all is included in the present invention.
Industrial applicibility
Corresponding with the present invention, the first, the shape information of input face image, it is right that this information table is shown An anamorphic image in reference picture extracts from input face image, and with reference Image has same modality but has different texture information, and adopted with input face image Extract the image of shape information, thereby the form of synthetic reaction input face image, and ignoring In the situation of the state of input face image, has high-quality natural new images. Therefore, originally Invention can be applied in various fields effectively, such as the generation of individual character, and virtual Cosmetic Design, crime Violate the Image Mosaics of search and make, and animation and game making.
The second, when generating cartoon image, comprise that the different caricature of user's Morphological Features is passable Immediately generation, and the part of the caricature that generates or integral body can extended or conversion. In addition, By utilizing the information of user's facial shape, complicated image correction process can be simplified and Can automatically carry out, thereby improve the productivity ratio that generates individual character.
The 3rd, when Cosmetic Design, the user can easily and check one about user's face The Cosmetic Design of image, and can carry out easily part or whole its Cosmetic Design of correction.
Per four, can immediately check the effect after a synthetic face-image adds various annexes.In addition, the present invention can be easy to various application, such as the incarnation that realizes according to shape information in the virtual reality, and the recovery of three dimensional face image, and the Video chat that needs face-image.
Claims (20)
1, a kind of shape information according to input face image synthesizes the device of a new face-image, and this device comprises:
A user's interface device, be used to receive facial image information and user's control commands, these command transfer to image processing apparatus, are received synthetic facial image information from image processing apparatus, and according to user control command output or the synthetic facial image information of storage; With
An image processing apparatus, the facial image information that is used for transmitting from user's interface device is extracted shape information and the texture information about input face image, its shape information is expressed as a picture about predetermined reference anamorphose, its texture information is about the light that is plotted in the input picture on the reference picture and the information of color, and be used for utilizing the shape information of input face image according to user command, the width of cloth texture image that conversion is selected from the texture image, in the image data base that this image is stored in and with the homomorphosis of reference picture, perhaps conversion is weighted the image that summation generates by the texture image to selected texture image and the reaction texture information of extracting, thereby generates a synthetic face-image.
2, device as claimed in claim 1 is characterized in that, user's interface device and image processing apparatus can be realized in an independent computer system.
3, device as claimed in claim 1, it is characterized in that, user's interface device and image processing apparatus are independently, be separately implemented in two or more computer systems, this device further comprises a communication network, with the data between transmission and reception user's interface device and image processing apparatus.
4, as claim 2 or 3 described devices, it is characterized in that image processing apparatus comprises:
A facial information extraction apparatus, in order to from the facial image information that user's interface device transmitted, extraction is about the shape information and the texture information of an input face image, its shape information shows as the anamorphic image to reference picture, its texture information promptly utilizes shape information to be plotted in the color or the monochrome information of the input face image on the reference picture;
A face-image compositor, in order to according to user control command, utilize the shape information of the input picture that the facial information extraction apparatus extracted, a selected width of cloth texture image will carry out conversion in the texture image in the image data base by being stored in, perhaps conversion piece image, this image is that the texture image by the reaction texture information of extracting to the texture image selected with by information extractor is weighted summation and generates, thereby synthesizes a new face-image; And
An image data base has the texture information of identical shaped various images in order to the information of stored reference image with reference picture.
5, device as claimed in claim 4 is characterized in that, image data base comprises:
A mask database, in order to storage form mean value, texture mean value, morphological feature vector, and texture proper vector, this proper vector value is to obtain by the covariance of form mean value, texture mean value, poor morphology covariance and texture difference is carried out principal component analysis, and its covariance obtains by the reference picture based on form of being extracted from some face-images and texture information;
An additional image data base, in order to store the information of relevant different-style cartoon image, this cartoon image has identical form with reference picture;
A cosmetic image data base, in order to store the information of relevant various Cosmetic Design images, this cosmetic image has the form identical with reference picture.
6, device as claimed in claim 5 is characterized in that, its face-image extraction apparatus comprises:
A normalization module is in order to carry out normalization according to reference picture to input face image;
A shape information estimation block, in order to according to shape information, the processing by normalized input face image and reference picture being carried out a kind of layering, gradient optical flow algorithm and reference picture;
A shape information compensating module, by basis being stored in the morphological feature vector in the mask database, the morphological feature vector of being estimated is carried out linearity decompose and carry out linear superposition, and realize the compensation of shape information error amount that the shape information estimation block is estimated;
An inverse transform module is utilized the shape information that compensates, and the shape information of input face image is transformed into the shape information of reference picture;
A texture information determination module by according to the texture proper vector that is stored in the mask database, has carried out linear decomposition and linear superposition to the texture information of reciprocal transformation image, thereby determines the texture information of input face image;
A circulation module is satisfied up to predetermined condition with upper module with repetitive cycling, thereby generates the shape information of input face image according to reference picture.
7, device as claimed in claim 5, it is characterized in that, the face image data storehouse further comprises a sample image database, all has the information of the caricature sample image of different facial expressions in order to each special part of storing relevant face-image, and image processing apparatus comprises that further parts of images replaces the unit, part or integral body that the sample image that utilizes of being stored in the sample image database and the face-image that synthesizes to have maximum similarity degree is replaced the synthetic face-image that is generated by the face-image compositor.
8, device as claimed in claim 5, it is characterized in that, the face data storehouse further comprises an annex image data base, in order to the information of storage about various annex map pictures, this annex map picture can be added on the synthetic face-image, and image processing apparatus further comprises an accessory unit, in order to add an annex to the synthetic face-image that is generated by the face-image compositor, this annex is one that selects from the annex map picture that is stored in the attachment data storehouse according to user control command.
9, a kind of method of synthesizing a new face-image according to the shape information of the face-image of importing, this method comprises the steps:
(a) from the facial image information of input, extract shape information and texture information about input face image, its shape information is expressed as a picture about predetermined reference anamorphose, and its texture information is about the light that is plotted in the input picture on the reference picture and the information of color; With
(b) utilize the shape information of input face image according to user command, the width of cloth texture image that conversion is selected from the texture image, this image be stored in the image data base and with the homomorphosis of reference picture, perhaps conversion is weighted the image that summation generates by the texture image to selected texture image and the reaction texture information of extracting, thereby generates a synthetic face-image.
10, method as claimed in claim 9 is characterized in that step (a) comprises following substep:
(a1) according to reference picture input face image is carried out normalization;
(a2), estimate reference picture by utilizing layering, gradient optical flow algorithm to handle to normalized input face image and reference picture according to shape information;
(a3) according to being stored in morphological feature vector in the mask database, the morphological feature vector of being estimated is carried out linearity decompose and carry out linear superposition, and the compensation of the shape information error amount that realization is estimated step (a2);
(a4) utilize the shape information that compensates in step (a4), the shape information of input face image is transformed into the shape information of reference picture;
(a5) according to the texture proper vector that is stored in advance in the mask database, the texture information of the image of conversion in step (a4) is carried out linearity decompose and linear superposition, thus the texture information of definite input face image;
(a6) refer step (a4) and result (a5), repeating step (a2) is satisfied up to predetermined condition to (a5), thereby generates the shape information and the texture information of input face image according to reference picture.
11, method as claimed in claim 10 is characterized in that, step (a) further comprises the steps:
(a0) according to shape information and the texture information from a large amount of model face images, extracted, covariance to the form mean value, texture mean value, poor morphology covariance and the texture difference that obtain from reference picture is carried out principal component analysis, and generates shape facility vector sum texture proper vector.
12, method as claimed in claim 9 is characterized in that, step (b) comprises following substep:
(b1) select a cartoon image according to user control command from the cartoon image of different styles, this image is stored in the image data base and with reference picture has identical form; And
(b2) synthesize selected cartoon image or an image in step (b1), the image that this image is a form of utilizing input face image by the texture information that selected caricature and a reaction are extracted is weighted summation and generates, thereby generates a cartoon image that reacts the shape information of input face image.
13, method as claimed in claim 12 is characterized in that, step (b) further comprises substep:
(b3) when user command confirm to change shape information, control the control command that shape information changes according to the user, change the shape information of input face image, and repeat (b1) and (b2).
14, method as claimed in claim 12 is characterized in that, further comprises the steps:
utilizes with being stored in the image data base with synthetic cartoon image has part or the integral body that the sample image of maximum comparability is replaced the cartoon image that is synthesized in step (b).
15, method as claimed in claim 14, it is characterized in that, its similarity is passed through shape information difference sum between synthetic cartoon image and sample image, reach the weighted sum of texture information gap between synthetic cartoon image and the sample image, perhaps the difference of the proper vector coefficient of the proper vector coefficient of the synthetic cartoon image shape information of expression and expression sample image shape information and, and the difference of the proper vector coefficient of the texture information of the proper vector coefficient of the synthetic cartoon image texture information of expression and expression sample image with weighted sum and definite.
16, as claim 12 or 14 described methods, it is characterized in that, further comprise step:
(d) synthetic cartoon image is added an annex map picture, this annex map similarly is to select from the annex map picture that is stored in image data base according to user control command.
17, method as claimed in claim 16 is characterized in that, is according to the shape information of the input face image that is extracted in step (a) and definite in the size of the annex map picture that step (d) is added and position.
18, method as claimed in claim 9 is characterized in that, step (b) comprises following substep:
(b1) according to user control command, from the Cosmetic Design image of various styles, select a Cosmetic Design image, this Cosmetic Design image is stored in the image data base, and has identical form with reference picture;
(b2) synthesize at the selected Cosmetic Design image of step (b1), perhaps image, this image is the shape information that utilizes input face image, image by texture information that selected Cosmetic Design image and one reaction is extracted is weighted that summation generates, thereby generates a Cosmetic Design image that reacts the shape information of input face image; And
(b3) indicate the texture information that the control command of making up changes reference picture of revising according to the user.
19, method as claimed in claim 18 is characterized in that, further comprises step:
adds an annex map picture to synthetic Cosmetic Design image, and this annex map similarly is according to selecting in the annex map picture of user control command from be stored in image data base.
20, method as claimed in claim 19 is characterized in that, the position and the size of the annex map picture that is added in step are determined by the shape information of the input face image that is extracted in the step (a).
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| KR1020000048616A KR20000064110A (en) | 2000-08-22 | 2000-08-22 | Device and method for automatic character generation based on a facial image |
| KR20000048616 | 2000-08-22 |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN1447955A true CN1447955A (en) | 2003-10-08 |
Family
ID=19684433
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN01814412A Pending CN1447955A (en) | 2000-08-22 | 2001-07-07 | Appts. and method for generating synthetic face image based on shape information about face image |
Country Status (5)
| Country | Link |
|---|---|
| JP (1) | JP2004506996A (en) |
| KR (2) | KR20000064110A (en) |
| CN (1) | CN1447955A (en) |
| AU (1) | AU2001269581A1 (en) |
| WO (1) | WO2002017234A1 (en) |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103530631A (en) * | 2012-07-06 | 2014-01-22 | 索尼电脑娱乐公司 | Image processing device and image processing method |
| CN102054287B (en) * | 2009-11-09 | 2015-05-06 | 腾讯科技（深圳）有限公司 | Facial animation video generating method and device |
| CN104616330A (en) * | 2015-02-10 | 2015-05-13 | 广州视源电子科技股份有限公司 | Picture generation method and device |
| CN104751408A (en) * | 2015-03-26 | 2015-07-01 | 广东欧珀移动通信有限公司 | Face image adjusting method and device |
| CN105427238A (en) * | 2015-11-30 | 2016-03-23 | 维沃移动通信有限公司 | Image processing method and mobile terminal |
| CN105513125A (en) * | 2014-10-08 | 2016-04-20 | 韩国科学技术研究院 | Apparatus and method for generating composite image, recording medium for performing the method |
| WO2017035966A1 (en) * | 2015-08-28 | 2017-03-09 | 百度在线网络技术（北京）有限公司 | Method and device for processing facial image |
| CN107705240A (en) * | 2016-08-08 | 2018-02-16 | 阿里巴巴集团控股有限公司 | Virtual examination cosmetic method, device and electronic equipment |
| CN108492344A (en) * | 2018-03-30 | 2018-09-04 | 中国科学院半导体研究所 | A kind of portrait-cartoon generation method |
| CN112862920A (en) * | 2021-02-18 | 2021-05-28 | 清华大学 | Human body image generation method and system based on hand-drawn sketch |
| CN113780047A (en) * | 2021-01-11 | 2021-12-10 | 北京沃东天骏信息技术有限公司 | Virtual makeup trying method and device, electronic equipment and storage medium |
Families Citing this family (50)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR100385896B1 (en) * | 1999-12-28 | 2003-06-02 | 김남규 | Method and Apparatus for Providing and Using of 3 Dimensional Image Which Represents the User in Cyber-Space |
| KR20010091743A (en) * | 2000-03-17 | 2001-10-23 | 박호성 | A formation method of an automatic caricature |
| KR20010092618A (en) * | 2000-03-22 | 2001-10-26 | 이민호 | Automatic generation and output of caricature of a face using image information |
| KR20020014176A (en) * | 2000-08-16 | 2002-02-25 | 김세진 | Apparatus and method for instant photographing and characterizing user's feature |
| KR20000064110A (en) * | 2000-08-22 | 2000-11-06 | 이성환 | Device and method for automatic character generation based on a facial image |
| KR20010000426A (en) * | 2000-09-28 | 2001-01-05 | 김용환 | Method of Intelligent Image Interface |
| KR20020057447A (en) * | 2001-01-04 | 2002-07-11 | 심한억 | The Method of Making a 3D Animation Movie By Controling 3D Character Directly |
| KR100407685B1 (en) * | 2001-01-12 | 2003-12-01 | 윤경현 | Method for representing Color paper mosaic using computer |
| KR100422470B1 (en) * | 2001-02-15 | 2004-03-11 | 비쥬텍쓰리디(주) | Method and apparatus for replacing a model face of moving image |
| KR20020069595A (en) * | 2001-02-27 | 2002-09-05 | 강석령 | System and method for producing caricatures |
| KR20020082328A (en) * | 2001-04-20 | 2002-10-31 | 김장휘 | The techknowledge maken my animation on network |
| KR20010079219A (en) * | 2001-06-23 | 2001-08-22 | 조한수 | Story board game system using video and its method |
| KR20030042403A (en) * | 2001-11-22 | 2003-05-28 | 조윤석 | Facial character manufacturing method by fitting facial edgeline |
| KR100912872B1 (en) * | 2002-10-09 | 2009-08-19 | 삼성전자주식회사 | Apparatus and method for producing three-dimensional caricature |
| CN1313979C (en) | 2002-05-03 | 2007-05-02 | 三星电子株式会社 | Apparatus and method for generating 3-D cartoon |
| KR100473593B1 (en) * | 2002-05-03 | 2005-03-08 | 삼성전자주식회사 | Apparatus and method for producing three-dimensional caricature |
| KR20030091306A (en) * | 2002-05-27 | 2003-12-03 | 이채헌 | The dynamic character and image character making system and the method using face components relationship on face image. |
| KR20040009460A (en) * | 2002-07-23 | 2004-01-31 | 주식회사 페이스쓰리디 | System and method for constructing three dimensional montaged geometric face |
| KR20040049759A (en) * | 2002-12-07 | 2004-06-12 | 김창모 | Caricature to the contents of mobile phone,PDA or Internet |
| KR101028257B1 (en) * | 2003-06-16 | 2011-04-11 | 엘지전자 주식회사 | Avatar editing method for mobile communication device |
| RU2358319C2 (en) * | 2003-08-29 | 2009-06-10 | Самсунг Электроникс Ко., Лтд. | Method and device for photorealistic three dimensional simulation of face based on image |
| KR100682889B1 (en) | 2003-08-29 | 2007-02-15 | 삼성전자주식회사 | Method and Apparatus for image-based photorealistic 3D face modeling |
| KR100791034B1 (en) * | 2004-09-02 | 2008-01-03 | (주)제니텀 엔터테인먼트 컴퓨팅 | Method of Hair-Style Shaping based-on Face Recognition and apparatus thereof |
| KR100764130B1 (en) * | 2005-03-29 | 2007-10-05 | (주)제니텀 엔터테인먼트 컴퓨팅 | Method of virtual face shaping based on automatic face extraction and apparatus thereof |
| US9583141B2 (en) | 2005-07-01 | 2017-02-28 | Invention Science Fund I, Llc | Implementing audio substitution options in media works |
| US9230601B2 (en) | 2005-07-01 | 2016-01-05 | Invention Science Fund I, Llc | Media markup system for content alteration in derivative works |
| US9426387B2 (en) | 2005-07-01 | 2016-08-23 | Invention Science Fund I, Llc | Image anonymization |
| US20080086380A1 (en) * | 2005-07-01 | 2008-04-10 | Searete Llc, A Limited Liability Corporation Of The State Of Delaware | Alteration of promotional content in media works |
| JP5035524B2 (en) * | 2007-04-26 | 2012-09-26 | 花王株式会社 | Facial image composition method and composition apparatus |
| US9215512B2 (en) | 2007-04-27 | 2015-12-15 | Invention Science Fund I, Llc | Implementation of media content alteration |
| KR100929561B1 (en) * | 2007-08-31 | 2009-12-03 | (주)에프엑스기어 | Specialized video contents providing system reflecting user-specified facial image / audio data |
| KR100929564B1 (en) * | 2007-08-31 | 2009-12-03 | (주)에프엑스기어 | Specialized virtual avatar providing system reflecting user specified facial image |
| KR100967895B1 (en) * | 2007-08-31 | 2010-07-06 | (주)에프엑스기어 | The system which provide a specialized teaching contents where the data which the user designates is reflected |
| KR100902995B1 (en) * | 2007-10-23 | 2009-06-15 | 에스케이 텔레콤주식회사 | Method for making face image of golden ratio, and apparatus applied to the same |
| KR100952382B1 (en) * | 2009-07-29 | 2010-04-14 | 숭실대학교산학협력단 | Animation automatic generating apparatus of user-based and its method |
| JP2011228936A (en) * | 2010-04-20 | 2011-11-10 | Shiseido Co Ltd | Moving image transmission system, transmitter, receiver, moving image management device, transmission program, reception program, and moving image management program |
| CN101847268B (en) * | 2010-04-29 | 2015-03-04 | 北京中星微电子有限公司 | Cartoon human face image generation method and device based on human face images |
| WO2012043910A1 (en) * | 2010-10-01 | 2012-04-05 | 엘지전자 주식회사 | Image display device and image displaying method thereof |
| KR101862128B1 (en) * | 2012-02-23 | 2018-05-29 | 삼성전자 주식회사 | Method and apparatus for processing video information including face |
| KR101374313B1 (en) * | 2012-08-14 | 2014-03-13 | 주식회사 바른기술 | An apparatus for transmitting simplified motion information excluding background images and displaying the information by utilizing avatar and the methods thereof |
| KR101494880B1 (en) * | 2012-11-07 | 2015-02-25 | 한국과학기술연구원 | Apparatus and method for generating cognitive avatar |
| KR101418878B1 (en) * | 2013-04-22 | 2014-07-17 | 명지대학교 산학협력단 | System for generating montage using facial feature and method therefor |
| KR102288280B1 (en) * | 2014-11-05 | 2021-08-10 | 삼성전자주식회사 | Device and method to generate image using image learning model |
| US10860841B2 (en) | 2016-12-29 | 2020-12-08 | Samsung Electronics Co., Ltd. | Facial expression image processing method and apparatus |
| US10762665B2 (en) | 2018-05-23 | 2020-09-01 | Perfect Corp. | Systems and methods for performing virtual application of makeup effects based on a source image |
| EP3664035B1 (en) * | 2018-12-03 | 2021-03-03 | Chanel Parfums Beauté | Method for simulating the realistic rendering of a makeup product |
| KR102400609B1 (en) * | 2021-06-07 | 2022-05-20 | 주식회사 클레온 | A method and apparatus for synthesizing a background and a face by using deep learning network |
| KR102623592B1 (en) | 2022-01-06 | 2024-01-11 | (주)키미티즈 | system and method for manufacturing smart character design |
| CN115171199B (en) * | 2022-09-05 | 2022-11-18 | 腾讯科技（深圳）有限公司 | Image processing method, image processing device, computer equipment and storage medium |
| KR102627033B1 (en) * | 2023-05-08 | 2024-01-19 | 주식회사 알마로꼬 | System and method for generating participatory content using artificial intelligence technology |
Family Cites Families (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH04199474A (en) * | 1990-11-29 | 1992-07-20 | Matsushita Electric Ind Co Ltd | Face picture synthetic device |
| JPH07325934A (en) * | 1992-07-10 | 1995-12-12 | Walt Disney Co:The | Method and equipment for provision of graphics enhanced to virtual world |
| JP3533717B2 (en) * | 1994-09-27 | 2004-05-31 | 松下電器産業株式会社 | Image synthesis device |
| JP3799633B2 (en) * | 1995-06-16 | 2006-07-19 | セイコーエプソン株式会社 | Face image processing method and face image processing apparatus |
| US5774591A (en) * | 1995-12-15 | 1998-06-30 | Xerox Corporation | Apparatus and method for recognizing facial expressions and facial gestures in a sequence of images |
| JP2918499B2 (en) * | 1996-09-17 | 1999-07-12 | 株式会社エイ・ティ・アール人間情報通信研究所 | Face image information conversion method and face image information conversion device |
| US6661906B1 (en) * | 1996-12-19 | 2003-12-09 | Omron Corporation | Image creating apparatus |
| JP3551668B2 (en) * | 1996-12-20 | 2004-08-11 | オムロン株式会社 | Portrait transmission device, portrait communication device and method |
| KR20010091743A (en) * | 2000-03-17 | 2001-10-23 | 박호성 | A formation method of an automatic caricature |
| KR20000037042A (en) * | 2000-04-06 | 2000-07-05 | 김정렬 | Automatic character producing system |
| KR20000063344A (en) * | 2000-06-26 | 2000-11-06 | 김성호 | Facial Caricaturing method |
| KR100376760B1 (en) * | 2000-07-05 | 2003-03-19 | 케이포테크놀로지 주식회사 | Method for manufacturing caricature |
| KR20000059236A (en) * | 2000-07-24 | 2000-10-05 | 조경식 | On the internet, the way of putting photo image on 3D-modeling, bringing each layer object, painting user's face, wearing a wig, putting on glasses, ....etc |
| KR20000064110A (en) * | 2000-08-22 | 2000-11-06 | 이성환 | Device and method for automatic character generation based on a facial image |
-
2000
- 2000-08-22 KR KR1020000048616A patent/KR20000064110A/en active Search and Examination
-
2001
- 2001-07-06 KR KR10-2001-0040472A patent/KR100407111B1/en not_active IP Right Cessation
- 2001-07-07 WO PCT/KR2001/001167 patent/WO2002017234A1/en active Application Filing
- 2001-07-07 AU AU2001269581A patent/AU2001269581A1/en not_active Abandoned
- 2001-07-07 JP JP2002521224A patent/JP2004506996A/en active Pending
- 2001-07-07 CN CN01814412A patent/CN1447955A/en active Pending
Cited By (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102054287B (en) * | 2009-11-09 | 2015-05-06 | 腾讯科技（深圳）有限公司 | Facial animation video generating method and device |
| CN103530631A (en) * | 2012-07-06 | 2014-01-22 | 索尼电脑娱乐公司 | Image processing device and image processing method |
| CN103530631B (en) * | 2012-07-06 | 2016-12-28 | 索尼电脑娱乐公司 | Image processing apparatus and image processing method |
| CN105513125B (en) * | 2014-10-08 | 2018-09-25 | 韩国科学技术研究院 | Composograph generating means and method, the recording medium for executing this method |
| CN105513125A (en) * | 2014-10-08 | 2016-04-20 | 韩国科学技术研究院 | Apparatus and method for generating composite image, recording medium for performing the method |
| CN104616330A (en) * | 2015-02-10 | 2015-05-13 | 广州视源电子科技股份有限公司 | Picture generation method and device |
| CN104751408A (en) * | 2015-03-26 | 2015-07-01 | 广东欧珀移动通信有限公司 | Face image adjusting method and device |
| WO2017035966A1 (en) * | 2015-08-28 | 2017-03-09 | 百度在线网络技术（北京）有限公司 | Method and device for processing facial image |
| US10599914B2 (en) | 2015-08-28 | 2020-03-24 | Baidu Online Network Technology (Beijing) Co., Ltd. | Method and apparatus for human face image processing |
| CN105427238A (en) * | 2015-11-30 | 2016-03-23 | 维沃移动通信有限公司 | Image processing method and mobile terminal |
| CN105427238B (en) * | 2015-11-30 | 2018-09-04 | 维沃移动通信有限公司 | A kind of image processing method and mobile terminal |
| CN107705240A (en) * | 2016-08-08 | 2018-02-16 | 阿里巴巴集团控股有限公司 | Virtual examination cosmetic method, device and electronic equipment |
| CN107705240B (en) * | 2016-08-08 | 2021-05-04 | 阿里巴巴集团控股有限公司 | Virtual makeup trial method and device and electronic equipment |
| CN108492344A (en) * | 2018-03-30 | 2018-09-04 | 中国科学院半导体研究所 | A kind of portrait-cartoon generation method |
| CN113780047A (en) * | 2021-01-11 | 2021-12-10 | 北京沃东天骏信息技术有限公司 | Virtual makeup trying method and device, electronic equipment and storage medium |
| CN112862920A (en) * | 2021-02-18 | 2021-05-28 | 清华大学 | Human body image generation method and system based on hand-drawn sketch |
Also Published As
| Publication number | Publication date |
|---|---|
| WO2002017234A1 (en) | 2002-02-28 |
| AU2001269581A1 (en) | 2002-03-04 |
| KR100407111B1 (en) | 2003-11-28 |
| JP2004506996A (en) | 2004-03-04 |
| KR20000064110A (en) | 2000-11-06 |
| KR20020015642A (en) | 2002-02-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN1447955A (en) | Appts. and method for generating synthetic face image based on shape information about face image | |
| US11861936B2 (en) | Face reenactment | |
| JP6639700B2 (en) | Method and system for generating a multimodal digital image | |
| Hertzmann et al. | Image analogies | |
| JP6611107B2 (en) | Computer system and method for upsampling images | |
| Kondratyuk et al. | Videopoet: A large language model for zero-shot video generation | |
| US8861800B2 (en) | Rapid 3D face reconstruction from a 2D image and methods using such rapid 3D face reconstruction | |
| CN1147822C (en) | Method and system for image processing | |
| CN101044761A (en) | Method for processing an input image securely | |
| US11024060B1 (en) | Generating neutral-pose transformations of self-portrait images | |
| WO2023000895A1 (en) | Image style conversion method and apparatus, electronic device and storage medium | |
| JP2021077365A (en) | Using convolutional neural network-style transfer to automate graphic design creation | |
| CN110853119A (en) | Robust reference picture-based makeup migration method | |
| Zeng et al. | Avatarbooth: High-quality and customizable 3d human avatar generation | |
| CN109272539A (en) | The decomposition method of image texture and structure based on guidance figure Total Variation | |
| Abdollahzadeh et al. | A survey on generative modeling with limited data, few shots, and zero shot | |
| CN111833413A (en) | Image processing method, image processing device, electronic equipment and computer readable storage medium | |
| CN1290064C (en) | Equipment and method for synthetizing higher-dimensional grain | |
| Hubball et al. | Image‐based aging using evolutionary computing | |
| JP7106144B2 (en) | Image analysis device | |
| CN109308726B (en) | Method and system for rapidly generating facial image editing template | |
| Lu et al. | Pano-NeRF: Synthesizing High Dynamic Range Novel Views with Geometry from Sparse Low Dynamic Range Panoramic Images | |
| Fernandes | An improved contrast enhancing approach for color-to-grayscale mappings | |
| Duan et al. | Realistic Image-to-Image Translation with Enhanced Texture | |
| CN112634441A (en) | 3D human body model generation method, system and related equipment |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| AD01 | Patent right deemed abandoned | ||
| C20 | Patent right or utility model deemed to be abandoned or is abandoned |