WO2017185452A1 - 图片还原方法及系统 - Google Patents
图片还原方法及系统 Download PDFInfo
- Publication number
- WO2017185452A1 WO2017185452A1 PCT/CN2016/084060 CN2016084060W WO2017185452A1 WO 2017185452 A1 WO2017185452 A1 WO 2017185452A1 CN 2016084060 W CN2016084060 W CN 2016084060W WO 2017185452 A1 WO2017185452 A1 WO 2017185452A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- area
- processed
- body unit
- human body
- picture
- Prior art date
Links
- 238000000034 method Methods 0.000 title claims abstract description 51
- 230000036548 skin texture Effects 0.000 claims description 74
- 239000011148 porous material Substances 0.000 claims description 58
- 238000000605 extraction Methods 0.000 claims description 24
- 230000037303 wrinkles Effects 0.000 claims description 20
- 238000012545 processing Methods 0.000 abstract description 11
- 230000003796 beauty Effects 0.000 description 50
- 210000001508 eye Anatomy 0.000 description 17
- 239000000284 extract Substances 0.000 description 12
- 210000004709 eyebrow Anatomy 0.000 description 10
- 238000005516 engineering process Methods 0.000 description 9
- 230000002441 reversible effect Effects 0.000 description 7
- 230000006870 function Effects 0.000 description 6
- 230000001815 facial effect Effects 0.000 description 5
- 230000002829 reductive effect Effects 0.000 description 5
- 238000010586 diagram Methods 0.000 description 4
- 210000001061 forehead Anatomy 0.000 description 4
- 210000002837 heart atrium Anatomy 0.000 description 4
- 238000005452 bending Methods 0.000 description 2
- 210000005252 bulbus oculi Anatomy 0.000 description 2
- 238000012937 correction Methods 0.000 description 2
- 210000000720 eyelash Anatomy 0.000 description 2
- 210000000744 eyelid Anatomy 0.000 description 2
- 230000000670 limiting effect Effects 0.000 description 2
- 230000000717 retained effect Effects 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000010365 information processing Effects 0.000 description 1
- 230000002452 interceptive effect Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 210000000214 mouth Anatomy 0.000 description 1
- 210000001331 nose Anatomy 0.000 description 1
- 238000000638 solvent extraction Methods 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
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
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
Definitions
- the present invention relates to the field of image technologies, and in particular, to a picture restoration method and system.
- the current inverse beauty algorithm only smears the beauty image, such as darkening the skin and making the eyes smaller by default. Although the appearance of the person being photographed in the beauty picture was restored to some extent, the actual details of the subject's appearance in the beauty picture were also maliciously distorted. At present, there is no practical basis for these methods of restoring beauty pictures.
- the beauty marks and portrait features in the beauty pictures are not used to restore the beauty pictures, and the true appearance of the person being photographed in the beauty pictures cannot be truly reflected.
- a picture restoration method comprising:
- the human body unit information included in each of the to-be-processed areas and each of the to-be-processed areas The feature information of the corresponding contrast area is subjected to a restoration process on the body unit of each of the to-be-processed areas.
- the picture to be restored includes a background area and a human body area;
- the contrast area includes any one or a combination of the following: a background area, an edge area of each of the to-be-processed areas, and a human body Remove other areas of the area to be processed in the area.
- the human body unit information includes an outline of a human body unit, and a skin texture of the to-be-processed area where the human body unit is located;
- the characteristic information includes noise information of the contrast area, and skin texture of the contrast area;
- the skin texture includes the size of the pores, the density of the pores, and the thickness of the wrinkles; and the contour of the human body unit and the skin texture in each of the areas to be treated can be reduced by any one or more of the following methods:
- Extracting a noise of the to-be-processed area determining a contour of the human body unit, and a skin texture corresponding to the human body unit according to a noise density of the to-be-processed area and a noise density of a comparison area corresponding to the to-be-processed area.
- Reduction direction and reduction range or
- the human body unit information includes a color of the human body unit; the characteristic information includes a color of the contrast area; and the color of the human body unit in each of the to-be-processed areas may pass any one or more of the following Ways to restore:
- a pixel value of the pixel is calculated according to pixel values of a plurality of adjacent pixels of the pixel.
- the method further includes:
- Extracting a special mark of each of the to-be-processed areas the special mark including a deformed standard geometric shape, a shape customized according to the picture to be restored;
- the human body unit of each of the to-be-processed areas is adjusted according to the special mark of each of the to-be-processed areas.
- adjusting the human body unit of each of the to-be-processed areas includes:
- a picture restoration system comprising:
- a dividing module configured to divide the to-be-restored picture into multiple to-be-processed areas
- An extraction module configured to extract body unit information in each of the plurality of to-be-processed areas
- the extraction module is further configured to extract feature information of the comparison area corresponding to each of the to-be-processed areas.
- a restoration module configured to perform a restoration process on the human body unit of each of the to-be-processed areas according to the human body unit information included in each of the to-be-processed areas and the feature information of the comparison area corresponding to each of the to-be-processed areas.
- the picture to be restored includes a background area and a human body area;
- the contrast area includes any one or a combination of the following: a background area, an edge area of each of the to-be-processed areas, and a human body Remove other areas of the area to be processed in the area.
- the human body unit information includes an outline of a human body unit, a skin texture of a to-be-processed area where the human body unit is located, and the characteristic information includes a noise signal of the contrast area.
- the skin texture of the contrast area; the skin texture includes the size of the pores, the density of the pores, the thickness of the wrinkles; and the contour and skin texture of the human body unit in each of the areas to be treated may pass through any one or more of the following Way to restore:
- Extracting a noise of the to-be-processed area determining a contour of the human body unit, and a skin texture corresponding to the human body unit according to a noise density of the to-be-processed area and a noise density of a comparison area corresponding to the to-be-processed area.
- Reduction direction and reduction range or
- the human body unit information includes a color of the human body unit; the characteristic information includes a color of the contrast area; and the color of the human body unit in each of the to-be-processed areas may pass any one or more of the following Ways to restore:
- a pixel value of the pixel is calculated according to pixel values of a plurality of adjacent pixels of the pixel.
- the system further comprises:
- the extraction module is further configured to extract a special mark of each of the to-be-processed areas, where the special mark includes a deformed standard geometric shape, and a shape customized according to the picture to be restored;
- the restoration module is further configured to: according to the special mark of each of the to-be-processed areas, each of the to-be-processed The body unit in the treatment area is adjusted.
- the system further includes: the obtaining module is further configured to acquire a standard shape corresponding to the special mark of the to-be-processed area;
- the reduction module is further configured to: restore the special mark according to the special mark and its corresponding standard shape, and obtain a reduction parameter;
- the reduction module is further configured to adjust a human body unit in the to-be-processed area according to the obtained reduction parameter.
- the present invention restores the to-be-restored picture according to the feature information in the picture to be restored, and can restore the true appearance of the person in the picture, thereby improving the user experience.
- FIG. 1 is a diagram showing an application environment of a preferred embodiment of a picture restoration system of the present invention.
- FIG. 2 is a flow chart of a preferred embodiment of the picture restoration method of the present invention.
- FIG. 3 is a functional block diagram of a preferred embodiment of the picture restoration system of the present invention.
- Figure 4 is a schematic illustration of the reduction of a beauty photograph of the present invention.
- FIG. 1 is an application environment diagram of a preferred embodiment of a picture restoration system provided by the present invention.
- An electronic device 1000 includes a picture restoration system 10.
- the electronic device 1000 further includes a memory 20 and a processor 30.
- the electronic device 1000 includes an electronic device capable of automatically performing numerical calculation and/or information processing according to an instruction set or stored in advance, the hardware including but not limited to a microprocessor, an application specific integrated circuit (ASIC), and a programmable Gate array (FPGA), digital processor (DSP), embedded devices, etc.
- a microprocessor an application specific integrated circuit (ASIC)
- ASIC application specific integrated circuit
- FPGA programmable Gate array
- DSP digital processor
- embedded devices etc.
- the electronic device 1000 may also include a network device and/or a user device.
- the network device includes, but is not limited to, a single network server, a server group composed of multiple network servers, or a cloud computing-based cloud composed of a large number of hosts or network servers, where the cloud computing is distributed computing.
- a super virtual computer consisting of a group of loosely coupled computers.
- the user equipment includes, but is not limited to, any electronic product that can interact with a user through a keyboard, a mouse, a remote controller, a touch pad, or a voice control device, such as a personal computer, a tablet computer, a smart phone, and a personal digital device.
- Assistant PDA, game console, or interactive network TV IPTV a variety of digital cameras, smart cameras and other equipment.
- the network in which the user equipment and the network device are located includes, but is not limited to, the Internet, a wide area network, a metropolitan area network, a local area network, a virtual private network VPN, and the like.
- the picture restoration system 10 is included in a picture processing plugin.
- the picture processing plug-in can be loaded into various independently operable programs such as various operating systems, software, applications (APPs), etc., to enable the loaded program to control the audio to be played to perform the corresponding processing as described in the present invention.
- APP Mito application software
- the photo control plug-in can be constructed by combining an existing plug-in with the picture restoration system 10, for example, including a picture restoration system 10 in an existing WeChat plug-in to obtain the picture processing plug-in, etc. .
- the picture restoration system 10 of the present invention can be used to implement a picture restoration method for restoring the picture to be restored according to feature information in a picture to be restored (for example, a beauty picture).
- the invention can restore the true appearance of the person to be photographed in the picture to be restored, and improves the user experience.
- FIG. 2 is a flow chart of a preferred embodiment of the picture restoration method of the present invention.
- the order of execution of the steps in the flowchart shown in the figure may be changed according to different requirements, and some steps may be omitted.
- the image restoration method includes:
- the picture restoration system 10 obtains a picture to be restored.
- the picture to be restored includes but is not limited to a picture to be reversed, and may also be a picture including a human body unit in other forms.
- the picture restoration system 10 acquires the to-be-reverse beauty picture.
- the restore function option can be selected on the user interface provided by the picture restoration system 10.
- the user may display the image to be reversed on the current user interface, and then input the “inverse beauty” voice, and after the image restoration system 10 receives the voice,
- the "reverse beauty” is recognized by the voice recognition technology, and then the to-be-reverse beauty picture is obtained.
- the picture restoration system 10 divides the picture to be restored into a plurality of to-be-processed areas.
- the picture restoration system 10 divides the picture to be restored into a background area and a human body area by using a picture recognition technology, and then divides the human body area into a plurality of to-be-processed areas.
- the plurality of to-be-processed areas include, but are not limited to, an eyebrow area, a nose area, an eye area, a mouth area, and a face non-five facial area.
- the picture restoration system 10 may divide the non-manifold area of the face based on the position information, for example, the non-five area above the eyebrow area is used as the forehead area, and the non-five area under the lip area is used as the chin area. In addition to the forehead area, the non-five areas outside the chin area serve as the atrium area.
- Each of the areas to be treated includes a human body unit including, but not limited to, eyebrows, nose, eyes, mouth, and face non-five features (such as cheeks).
- the eyebrow area contains two eyebrows and the nose area contains a nose.
- the restoration operation is performed on each human body unit in the subsequent processing area.
- the plurality of areas to be treated may also include other units of the human body, such as hands, arms, and the like.
- the picture restoration system 10 extracts each of the plurality of to-be-processed areas to be processed. Human body unit information.
- the human body unit information includes at least one of the following: a position of the human body unit, a contour of the human body unit, a color of the human body unit, and a skin texture of the to-be-processed area where the human body unit is located.
- the position of the human body unit is a position of the human body unit in the picture to be restored, and the outline of the human body unit includes the shape, size, and the like of the human body unit.
- the color of the human body unit includes the chromaticity, brightness, and the like of the human body unit.
- the skin texture includes skin color, size of pores, thickness of wrinkles, and the like.
- the human body unit in the to-be-processed area is first identified by using a picture recognition technology, and the human body unit information in the to-be-processed area is extracted by using a feature extraction technique.
- the image recognition technique is used to determine the location of the nose from the nose region, and then feature extraction techniques are used to extract the position of the nose, the contours (ie, the bridge of the nose and the nose), and the texture of the skin on the nose.
- the image recognition technology is used to determine the position of the eye from the eye region, and then the feature extraction technique is used to extract the position of the eye, the contour (including the eyelid, the eyeball, the eyelash) and the pixel grayscale of the eye.
- the picture restoration system 10 may also extract noise information in each of the areas to be processed. Taking the anti-beauty picture as an example, since the picture is subjected to the beauty treatment, the noise of the part of the picture is also reduced. Therefore, the noise of the area to be treated and other areas can be compared to restore the body of the area to be treated. unit.
- the picture restoration system 10 may also extract special indicia for each area to be processed.
- Each special mark corresponds to a standard shape.
- the special indicia include, but are not limited to, a standard geometric shape of the deformation. For example, a deformed straight line, a deformed circle, a deformed ellipse, a deformed rectangle, or the like.
- a certain part of the line may be deformed (such as a protrusion). , extract the line of deformation. Subsequent correction of the deformed line can be used to restore the face after the beauty.
- the standard shape corresponding to the special mark may be pre-stored in the memory 20.
- the picture restoration system 10 may directly extract the special mark in the to-be-processed area by using a feature extraction technique.
- the picture restoration system 10 can also receive an area selected by the user in the to-be-processed area, and extract a special standard in the selected area by using feature extraction technology.
- the special mark is subsequently restored to obtain a reduction parameter, and the human body unit in the area to be treated is restored according to the reduction parameter.
- the special mark may also be other shapes defined by the user according to the picture to be restored.
- the picture restoration system 10 also needs to receive a standard shape corresponding to the special mark input by the user.
- the picture restoration system 10 extracts feature information of the comparison area corresponding to each of the to-be-processed areas.
- each area to be processed corresponds to at least one contrast area.
- the contrast area may be any one or more of the following: an edge area of each to-be-processed area, a background area of a picture to be restored, and other areas in the body area from which the area to be processed is removed, for example, the The area to be treated is a nose area, and other areas in the body area from which the area to be treated is removed may be an atrium area, a chin area, or the like.
- the feature information of the contrast area includes at least one of the following: noise information of the contrast area, color of the contrast area, and skin texture of the contrast area.
- the color of the contrast area includes the chromaticity, brightness, and the like of the contrast area.
- the skin texture of the contrast area includes the skin color in the contrast area, the size of the pores, the thickness of the wrinkles, and the like.
- the picture restoration system 10 performs, according to the human body unit information included in each to-be-processed area and the feature information of the comparison area corresponding to each of the to-be-processed areas, the human body unit of each of the to-be-processed areas. Restore processing.
- the picture restoration system 10 determines the restoration of the human body unit of the to-be-processed area according to the human body unit information included in each of the to-be-processed areas and the feature information of the comparison area corresponding to each of the to-be-processed areas.
- Direction and extent of reduction For example, taking the reverse beauty picture as an example, the reduction direction of the human body unit is that the eyes become smaller, the skin becomes darker, the face area becomes larger, and the like.
- the picture restoration system 10 determines the skin texture or noise of the contrast area corresponding to the area to be processed according to the skin texture or noise information of the area to be processed.
- the point information determines a contour of the human body unit in the area to be treated, a reduction direction corresponding to the skin texture of the human body unit, and a reduction width.
- the determination of the contour of the human body unit, the reduction direction corresponding to the skin texture of the human body unit, and the reduction amplitude may be implemented by any one or more of the following combinations:
- the contrast region of the nose region is the edge region of the nose.
- the average size of the pores of the edge region is calculated as A by extracting the pores of the edge region.
- the pores in the nose region are extracted, and the average size of the pores in the nose region is calculated as B. If A is greater than B, the contour of the nose region corresponds to a reduction direction from small to large. If A is smaller than B, the contour of the nose corresponds to a reduction direction from large to small.
- the contour of the nose corresponds to a reduction of A/B.
- the reduction direction and the reduction range corresponding to the contour of the nose can also be determined by calculating the density of the pores in the edge region and the density of the pores on the nose.
- the picture restoration system 10 simultaneously determines the contour of the human body unit and the skin texture of the human body unit in each of the to-be-processed areas by using the skin texture and noise information of the contrast area corresponding to each to-be-processed area. Direction and extent of reduction.
- the skin texture and noise information of the to-be-processed area and the to-be-processed area are The skin texture and the noise information of the contrast area corresponding to the domain determine the contour of the human body unit in the area to be processed, the reduction direction and the reduction range corresponding to the skin texture of the human body unit. For example, an average value of the pore size is calculated according to the pore size of the area to be processed, and an average value of the noise size is calculated according to the noise size of the area to be processed, and an average value of the pore size and an average value of the noise size of the area to be treated are obtained. Weighting is performed to obtain a weighting value corresponding to the to-be-processed area.
- the direction of reduction of the area to be treated may be determined according to the direction of bending of the special mark relative to its corresponding standard shape.
- the picture restoration system 10 adjusts the color of the human body unit in the to-be-processed area according to the color of the to-be-processed area and the color of the contrast area corresponding to the to-be-processed area for each to-be-processed area.
- the adjustment of the color of the human body unit in the to-be-processed area may be implemented by any one or more of the following combinations:
- the average pixel value of the mouth area is P
- the average pixel value of the edge area corresponding to the mouth area is Q
- the pixel value of each pixel point in the mouth area is 0.5P+0.5Q.
- the adjacent pixel points may be located in an edge region of the area to be processed.
- the adjacent four pixels are pixel point a, pixel point b, pixel point c, pixel point d, where pixel point a, pixel point b, pixel point c is located in the edge area.
- the pixel value of the first pixel is an average value of the adjacent four pixel points.
- the picture when a special mark is extracted in the area to be processed, the picture is also The original system 10 acquires a standard shape corresponding to the special mark; according to the special mark and its corresponding standard shape, the special mark is restored, and a reduction parameter is obtained. Adjusting the human body unit in the to-be-processed area according to the obtained reduction parameter.
- the picture restoration system 10 further acquires portrait distribution data, which may be pre-stored in the memory 20 or may be input by the user according to the picture to be restored.
- the portrait distribution data includes an average position of the human body unit in the human body and a color interval range of the human body unit.
- the picture restoration system 10 fine-tunes the human body unit in the picture to be restored according to the portrait distribution data. This will prevent errors during the restore process.
- the picture restoration system 10 may also acquire other pictures of the user in the picture to be restored, and the other pictures may be pictures without beauty.
- the portrait distribution data is then acquired based on the other pictures.
- the picture restoration system 10 fine-tunes and verifies the human body unit in the picture to be restored according to the portrait distribution data. This will prevent errors during the restore process.
- the picture restoration system 10 of the present invention can restore the human body unit in the picture to be restored according to the features in the picture to be restored, instead of simply performing the ugly processing.
- a beauty photo is reversed, and the photo after the beauty is more realistic. In this way, the user can see the true appearance of the person being photographed in the beauty picture.
- the picture restoration system 10 includes an acquisition module 100, a division module 101, an extraction module 102, and a restoration module 103.
- the obtaining module 100 is configured to obtain a picture to be restored.
- the picture to be restored includes, but is not limited to, a picture to be reversed, and may also be a picture of other forms including a human body unit.
- the obtaining module 100 acquires the to-be-reverse beauty picture.
- the restore function option can be selected on the user interface provided by the picture restoration system 10.
- the user may display the image to be reversed on the current user interface, and then input the “inverse beauty” voice, and the acquiring module 100 receives the voice.
- the "reverse beauty” is recognized by the voice recognition technology, and then the to-be-reverse beauty picture is obtained.
- the dividing module 101 is configured to divide the picture to be restored into a plurality of to-be-processed areas.
- the dividing module 101 divides the picture to be restored into a background area and a human body area by using a picture recognition technology, and then divides the human body area into a plurality of to-be-processed areas.
- the plurality of to-be-processed areas include, but are not limited to, an eyebrow area, a nose area, an eye area, a mouth area, and a face non-five facial area.
- the dividing module 101 may divide the non-manifold region of the face based on the position information, for example, the non-five facial region above the eyebrow region is used as the forehead region, and the non-five facial region below the lip region is used as the chin region.
- the non-Five-Autonomous Region outside the forehead area and the chin area serves as the atrium area.
- Each of the areas to be treated includes a human body unit including, but not limited to, an eyebrow, a nose, an eye, a mouth, and a non-five facial unit such as a cheek.
- the eyebrow area contains two eyebrows and the nose area contains a nose.
- the restoration operation is performed on each human body unit in the subsequent processing area.
- the plurality of areas to be treated may also include other units of the human body, such as hands, arms, and the like.
- the extraction module 102 is configured to extract body unit information in each of the plurality of to-be-processed areas.
- the human body unit information includes at least one of the following: a position of the human body unit, a contour of the human body unit, a color of the human body unit, and a skin texture of the to-be-processed area where the human body unit is located.
- the position of the human body unit is a position of the human body unit in the picture to be restored, and the outline of the human body unit includes the shape, size, and the like of the human body unit.
- the color of the human body unit includes the chromaticity, brightness, and the like of the human body unit.
- the skin texture includes skin color, size of pores, thickness of wrinkles, and the like.
- the human body unit in the to-be-processed area is first identified by using a picture recognition technology, and the human body unit information in the to-be-processed area is extracted by using a feature extraction technique.
- the image recognition technique is used to determine the location of the nose from the nose region, and then feature extraction techniques are used to extract the position of the nose, the contours (ie, the bridge of the nose and the nose), and the texture of the skin on the nose.
- the image recognition technology is used to determine the position of the eye from the eye region, and then the feature extraction technique is used to extract the position of the eye, the contour (including the eyelid, the eyeball, the eyelash) and the pixel grayscale of the eye.
- the extraction module 102 may also extract noise information in each area to be processed. Taking the anti-beauty picture as an example, since the picture is subjected to the beauty treatment, the noise of the part of the picture is also reduced. Therefore, the noise of the area to be treated and other areas can be compared to restore the body of the area to be treated. unit.
- the extraction module 102 may also extract special indicia for each area to be processed.
- Each special mark corresponds to a standard shape.
- the special indicia include, but are not limited to, a standard geometric shape of the deformation. For example, a deformed straight line, a deformed circle, a deformed ellipse, a deformed rectangle, or the like.
- a certain part of the line may be deformed (such as a protrusion). , extract the line of deformation. Subsequent correction of the deformed line can be used to restore the face after the beauty.
- the standard shape corresponding to the special mark may be pre-stored in the memory 20.
- the extraction module 102 may directly extract the special mark in the to-be-processed area by using a feature extraction technique.
- the picture restoration system 10 may also receive an area selected by the user in the area to be processed, and extract a special mark in the selected area by using feature extraction techniques. The special mark is subsequently restored to obtain a reduction parameter, and the human body unit in the area to be treated is restored according to the reduction parameter.
- the special mark may also be other shapes defined by the user according to the picture to be restored.
- the extraction module 102 also needs to receive a standard shape corresponding to the special mark input by the user.
- the extraction module 102 is further configured to extract feature information of the comparison area corresponding to each of the to-be-processed areas.
- each area to be processed corresponds to at least one contrast area.
- the contrast area may be any one or more of the following: an edge area of each to-be-processed area, a background area of a picture to be restored, and other areas in the body area from which the area to be processed is removed, for example, the The area to be treated is a nose area, and other areas in the body area from which the area to be treated is removed may be an atrium area, a chin area, or the like.
- the feature information of the contrast area includes at least one of the following: noise information of the contrast area, color of the contrast area, and skin texture of the contrast area.
- the color of the contrast area includes the chromaticity, brightness, and the like of the contrast area.
- the skin texture of the contrast area includes the skin color in the contrast area, the size of the pores, the thickness of the wrinkles, and the like.
- the restoration module 103 is configured to perform a restoration process on the human body unit of each of the to-be-processed areas according to the human body unit information included in each of the to-be-processed areas and the feature information of the comparison area corresponding to each of the to-be-processed areas.
- the restoration module 103 determines the restoration direction of the human body unit of the to-be-processed area according to the human body unit information included in each of the to-be-processed areas and the feature information of the comparison area corresponding to each of the to-be-processed areas. And the extent of reduction. For example, taking the reverse beauty picture as an example, the reduction direction of the human body unit is that the eyes become smaller, the skin becomes darker, the face area becomes larger, and the like.
- the restoration module 103 determines, in each to-be-processed area, the skin texture or noise information of the contrast area corresponding to the to-be-processed area according to the skin texture or noise information of the to-be-processed area.
- the determination of the contour of the human body unit, the reduction direction corresponding to the skin texture of the human body unit, and the reduction amplitude may be implemented by any one or more of the following combinations:
- the contrast region of the nose region is the edge region of the nose.
- the average size of the pores of the edge region is calculated as A by extracting the pores of the edge region.
- the pores in the nose region are extracted, and the average size of the pores in the nose region is calculated as B. If A is greater than B, the contour of the nose region corresponds to a reduction direction from small to large. If A is smaller than B, the contour of the nose corresponds to a reduction direction from large to small.
- the contour of the nose corresponds to a reduction of A/B.
- the reduction direction and the reduction range corresponding to the contour of the nose can also be determined by calculating the density of the pores in the edge region and the density of the pores on the nose.
- the restoration module 103 simultaneously determines the contour of the human body unit and the skin texture of the human body unit in each of the to-be-processed regions by using skin texture and noise information of the contrast region corresponding to each to-be-processed region. Restore direction and reduction range. Specifically, determining, according to skin texture and noise information of the to-be-processed area, skin texture and noise information of the contrast area corresponding to the to-be-processed area, corresponding to the contour of the human body unit and the skin texture of the human body unit in the to-be-processed area The direction of reduction and the extent of reduction.
- an average value of the pore size is calculated according to the pore size of the area to be processed, and an average value of the noise size is calculated according to the noise size of the area to be processed, and an average value of the pore size and an average value of the noise size of the area to be treated are obtained. Weighting is performed to obtain a weighting value corresponding to the to-be-processed area. Similarly, calculating an average value of the pore size and an average value of the noise size of the contrast area corresponding to the area to be processed, and weighting the average value of the pore size and the average value of the noise size of the contrast area corresponding to the area to be processed to obtain the to-be-processed The weighted value corresponding to the contrast area corresponding to the area.
- the direction of reduction of the area to be treated may be determined according to the direction of bending of the special mark relative to its corresponding standard shape.
- the restoration module 103 is configured according to the to-be-processed area for each to-be-processed area.
- the color of the contrast of the region corresponding to the area to be processed adjusts the color of the human body unit in the area to be processed.
- the adjustment of the color of the human body unit in the to-be-processed area may be implemented by any one or more of the following combinations:
- the average pixel value of the mouth area is P
- the average pixel value of the edge area corresponding to the mouth area is Q
- the pixel value of each pixel point in the mouth area is 0.5P+0.5Q.
- the adjacent pixel points may be located in an edge region of the area to be processed.
- the adjacent four pixels are pixel point a, pixel point b, pixel point c, pixel point d, where pixel point a, pixel point b, pixel point c is located in the edge area.
- the pixel value of the first pixel is an average value of the adjacent four pixel points.
- the restoration module 103 acquires a standard shape corresponding to the special mark; according to the special mark and its corresponding standard shape, the special The tag is restored and the restore parameters are obtained. Adjusting the human body unit in the to-be-processed area according to the obtained reduction parameter.
- the restoration module 103 further acquires portrait distribution data, which may be pre-stored in the memory 20 or may be input by the user according to the data to be restored.
- the portrait distribution data includes an average position of the human body unit in the human body and a color interval range of the human body unit.
- the restoration module 103 fine-tunes the human body unit in the picture to be restored according to the portrait distribution data. This will prevent errors during the restore process.
- the restoration module 103 may also acquire other pictures of the user in the picture to be restored, and the other pictures may be pictures without beauty.
- the portrait distribution data is then acquired based on the other pictures.
- the restoration module 103 performs fine adjustment and verification on the human body unit in the picture to be restored according to the portrait distribution data. This will prevent errors during the restore process.
- the invention can restore the human body unit in the picture to be restored according to the features in the picture to be restored, instead of simply doing the ugly processing.
- a beauty photo is reversed, and the photo after the beauty is more realistic. This way the user can see the beauty picture being The true look of the shooter.
- each functional module in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.
- the above integrated unit can be implemented in the form of hardware or in the form of hardware plus software function modules.
- the above-described integrated unit implemented in the form of a software function module can be stored in a computer readable storage medium.
- the software function modules described above are stored in a storage medium and include instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) or a processor to perform the methods of the various embodiments of the present invention. Part of the steps.
Abstract
一种图片还原方法和图片还原系统，该还原方法包括：获取待还原图片（S10）；将所述待还原图片划分成多个待处理区域（S11）；提取所述多个待处理区域中的每个待处理区域中的人体单元信息（S12）；提取所述每个待处理区域对应的对比区域的特征信息（S13）；及根据所述每个待处理区域所包含的人体单元信息及所述每个待处理区域对应的对比区域的特征信息，对所述每个待处理区域的人体单元进行还原处理（S14）。图片还原方法和图片还原系统能还原图片中被拍者的真实容貌，提高了用户的体验。
Description
本申请要求于2016年4月27日提交中国专利局，申请号为201610274328.1、发明名称为“图片还原方法及系统”的中国专利申请的优先权，其全部内容通过引用结合在本申请中。
本发明涉及图像技术领域，尤其涉及一种图片还原方法及系统。
越来越多的人喜欢使用美颜的方法获得美观的肖像照片。虽然美颜拍照的方法使被拍者达到了美观的效果，但在某些程度上也误导了看这些美颜图片的用户。相机美颜算法在给用户带来美观、乐趣的同时，用户也越来越想获知美颜图片中被拍者的真实的容貌情况。假如用户没有见过被拍者，用户通过观看美颜图片根本无法获知被拍者的实际容貌。
目前的逆美颜算法只是对美颜图片做丑化处理，例如将皮肤变黑，眼睛按照默认设置变小等。尽管在一定程度上还原了美颜图片中被拍者的容貌，但美颜图片中被拍者的容貌的实际细节也同时被恶意扭曲。而且目前这些还原美颜图片的方法并没有实际的依据，没有利用美颜图片中的美化痕迹及肖像特点来还原美颜图片，不能真实反应美颜图片中被拍者的真实容貌。
发明内容
鉴于以上内容，有必要提出一种图片还原方法及系统，能还原图片中被拍者的真实容貌，提高了用户的体验。
一种图片还原方法，所述方法包括：
获取待还原图片；
将所述待还原图片划分成多个待处理区域；
提取所述多个待处理区域中的每个待处理区域中的人体单元信息；
提取所述每个待处理区域对应的对比区域的特征信息；及
根据所述每个待处理区域所包含的人体单元信息及所述每个待处理区域
对应的对比区域的特征信息，对所述每个待处理区域的人体单元进行还原处理。
根据本发明一优选实施例，所述待还原图片包括背景区域及人体区域；所述对比区域包括以下任一种或者多种的组合：背景区域、所述每个待处理区域的边缘区域、人体区域中除去待处理区域的其他区域。
根据本发明一优选实施例，所述人体单元信息包括人体单元的轮廓、所述人体单元所在的待处理区域的皮肤纹理；所述特征信息包括对比区域的噪点信息、对比区域的皮肤纹理；所述皮肤纹理包括毛孔的大小、毛孔的密度、皱纹的粗细；及所述每个待处理区域中人体单元的轮廓及皮肤纹理可通过以下任一种或者多种方式进行还原：
根据所述待处理区域的毛孔大小与所述待处理区域对应的对比区域的毛孔大小来确定所述人体单元的轮廓、所述人体单元的皮肤纹理所对应的还原方向及还原幅度；或
根据所述待处理区域的毛孔密度与所述待处理区域对应的对比区域的毛孔密度来确定所述人体单元的轮廓、所述人体单元的皮肤纹理所对应的还原方向及还原幅度；或
提取所述待处理区域的噪点，根据所述待处理区域的噪点大小与所述待处理区域对应的对比区域的噪点大小来确定人体单元的轮廓、人体单元的皮肤纹理所对应的还原方向及还原幅度；或
提取所述待处理区域的噪点，根据所述待处理区域的噪点密度与所述待处理区域对应的对比区域的噪点密度来确定所述人体单元的轮廓、所述人体单元的皮肤纹理所对应的还原方向及还原幅度；或
根据所述待处理区域的皱纹粗细与所述待处理区域对应的对比区域的皱纹粗细来确定所述人体单元的轮廓、所述人体单元的皮肤纹理所对应的还原方向及还原幅度。
根据本发明一优选实施例，所述人体单元信息包括人体单元的色彩；所述特征信息包括对比区域的色彩；及所述每个待处理区域中人体单元的色彩可通过以下任一种或者多种方式进行还原：
将所述待处理区域的像素值与所述待处理区域对应的对比区域的像素值进行加权，根据加权后的像素值确定所述待处理区域中每个像素点的像素值；
或
针对所述待处理区域中每个像素点，根据像素点的多个相邻像素点的像素值计算所述像素点的像素值。
根据本发明一优选实施例，所述方法还包括：
提取所述每个待处理区域的特殊标记，所述特殊标记包括变形的标准几何形状、根据待还原图片自定义的形状；
根据所述每个待处理区域的特殊标记，对所述每个待处理区域的人体单元进行调整。
根据本发明一优选实施例，对所述每个待处理区域的人体单元进行调整包括：
获取所述待处理区域的特殊标记所对应的标准形状；
根据所述特殊标记及其对应的标准形状，对所述特殊标记进行还原，并得到还原参数；及
根据所得到的还原参数对所述待处理区域中的人体单元进行调整。
一种图片还原系统，所述系统包括：
获取模块，用于获取待还原图片；
划分模块，用于将所述待还原图片划分成多个待处理区域；
提取模块，用于提取所述多个待处理区域中的每个待处理区域中的人体单元信息；
所述提取模块还用于提取所述每个待处理区域对应的对比区域的特征信息；及
还原模块，用于根据所述每个待处理区域所包含的人体单元信息及所述每个待处理区域对应的对比区域的特征信息，对所述每个待处理区域的人体单元进行还原处理。
根据本发明一优选实施例，所述待还原图片包括背景区域及人体区域；所述对比区域包括以下任一种或者多种的组合：背景区域、所述每个待处理区域的边缘区域、人体区域中除去待处理区域的其他区域。
根据本发明一优选实施例，所述人体单元信息包括人体单元的轮廓、所述人体单元所在的待处理区域的皮肤纹理；所述特征信息包括对比区域的噪点信
息、对比区域的皮肤纹理；所述皮肤纹理包括毛孔的大小、毛孔的密度、皱纹的粗细；及所述每个待处理区域中人体单元的轮廓及皮肤纹理可通过以下任一种或者多种方式进行还原：
根据所述待处理区域的毛孔大小与所述待处理区域对应的对比区域的毛孔大小来确定所述人体单元的轮廓、所述人体单元的皮肤纹理所对应的还原方向及还原幅度；或
根据所述待处理区域的毛孔密度与所述待处理区域对应的对比区域的毛孔密度来确定所述人体单元的轮廓、所述人体单元的皮肤纹理所对应的还原方向及还原幅度；或
提取所述待处理区域的噪点，根据所述待处理区域的噪点大小与所述待处理区域对应的对比区域的噪点大小来确定人体单元的轮廓、人体单元的皮肤纹理所对应的还原方向及还原幅度；或
提取所述待处理区域的噪点，根据所述待处理区域的噪点密度与所述待处理区域对应的对比区域的噪点密度来确定所述人体单元的轮廓、所述人体单元的皮肤纹理所对应的还原方向及还原幅度；或
根据所述待处理区域的皱纹粗细与所述待处理区域对应的对比区域的皱纹粗细来确定所述人体单元的轮廓、所述人体单元的皮肤纹理所对应的还原方向及还原幅度。
根据本发明一优选实施例，所述人体单元信息包括人体单元的色彩；所述特征信息包括对比区域的色彩；及所述每个待处理区域中人体单元的色彩可通过以下任一种或者多种方式进行还原：
将所述待处理区域的像素值与所述待处理区域对应的对比区域的像素值进行加权，根据加权后的像素值确定所述待处理区域中每个像素点的像素值；或
针对所述待处理区域中每个像素点，根据像素点的多个相邻像素点的像素值计算所述像素点的像素值。
根据本发明一优选实施例，所述系统还包括：
所述提取模块还用于提取所述每个待处理区域的特殊标记，所述特殊标记包括变形的标准几何形状、根据待还原图片自定义的形状；
所述还原模块还用于根据所述每个待处理区域的特殊标记，对所述每个待
处理区域的人体单元进行调整。
根据本发明一优选实施例，该系统还包括：所述获取模块还用于获取所述待处理区域的特殊标记所对应的标准形状；
所述还原模块还用于根据所述特殊标记及其对应的标准形状，对所述特殊标记进行还原，并得到还原参数；及
所述还原模块还用于根据所得到的还原参数对所述待处理区域中的人体单元进行调整。
相较于现有技术，本发明根据待还原图片中的特征信息，将所述待还原图片进行还原，能还原图片中被拍者的真实容貌，提高了用户的体验。
图1是本发明图片还原系统较佳实施例的应用环境图。
图2是本发明图片还原方法的较佳实施例的流程图。
图3是本发明图片还原系统较佳实施例的功能模块图。
图4是本发明对美颜照片还原的示意图。
【主要元件符号说明】
电子设备 1000
图片还原系统 10
存储器 20
处理器 30
获取模块 100
划分模块 101
提取模块 102
还原模块 103
下面将结合本发明实施例中的附图，对本发明实施例中的技术方案进行清除、完整地描述，显然，所描述的实施例仅仅是本发明的一部分实施
例，而不是全部的实施例。
基于本发明中的实施例，本领域普通技术人员在没有做出创造性劳动的前提下所获得的所有其他实施例，都属于本发明保护的范围。
请参考图1，是本发明提供的图片还原系统较佳实施例的应用环境图。一个电子设备1000包括图片还原系统10。所述电子设备1000还包括存储器20及处理器30。
所述电子设备1000包括一种能够按照事先设定或存储的指令，自动进行数值计算和/或信息处理的电子设备，其硬件包括但不限于微处理器、专用集成电路(ASIC)、可编程门阵列(FPGA)、数字处理器(DSP)、嵌入式设备等。
所述电子设备1000还可包括网络设备和/或用户设备。其中，所述网络设备包括但不限于单个网络服务器、多个网络服务器组成的服务器组或基于云计算(Cloud Computing)的由大量主机或网络服务器构成的云，其中，云计算是分布式计算的一种，由一群松散耦合的计算机集组成的一个超级虚拟计算机。所述用户设备包括但不限于任何一种可与用户通过键盘、鼠标、遥控器、触摸板或声控设备等方式进行人机交互的电子产品，例如，个人计算机、平板电脑、智能手机、个人数字助理PDA、游戏机、或交互式网络电视IPTV，各种数码相机、智能相机等设备。其中，所述用户设备及网络设备所处的网络包括但不限于互联网、广域网、城域网、局域网、虚拟专用网络VPN等。所述图片还原系统10包含于图片处理插件中。所述图片处理插件可被加载至各类操作系统、软件、应用(APP)等可独立运行的程序中，以使被加载程序能够控制待播放音频执行如本发明所述的相应处理。例如，实现时，在现有的美图应用软件(APP)，如美图秀秀的基础上，增加相应的代码实现。
更优选地，可通过将现有的插件与所述图片还原系统10相结合来构成所述拍照控制插件，例如，在现有的微信插件中包含图片还原系统10，以获得该图片处理插件等。
本发明的图片还原系统10可以用于实现一种图片还原方法，用于根据待还原图片(例如美颜图片)中的特征信息，将所述待还原图片进行还原。本发明能还原待还原图片中被拍者的真实容貌，提高了用户的体验。
如图2所示，是本发明图片还原方法的较佳实施例的流程图。根据不同的需求，该图所示流程图中步骤的执行顺序可以改变，某些步骤可以省略。
所述图片还原方法包括：
S10，图片还原系统10获取待还原图片。
在本实施例中，所述待还原图片包括但不限于待逆美颜图片，也可以是其他形式包括人体单元的图片。以待逆美颜图片为例，图片还原系统10接收到用户对一张待逆美颜图片进行还原的请求时，图片还原系统10获取所述待逆美颜图片。
当用户需要对一张待还原图片进行还原时，可以在图片还原系统10提供的用户界面上选择还原功能选项。
在其他实施例中，以待逆美颜图片为例，用户可以将待逆美颜图片显示在当前用户界面上，然后语音输入“逆美颜”，图片还原系统10接收到所述语音后，通过语音识别技术识别出“逆美颜”，然后获取所述待逆美颜图片。
S11，图片还原系统10将所述待还原图片划分成多个待处理区域。
在本实施例中，图片还原系统10利用图片识别技术将所述待还原图片先划分成背景区域与人体区域，再将人体区域划分成多个待处理区域。现实生活中，所述多个待处理区域包括但不限于眉毛区域，鼻子区域，眼睛区域，嘴巴区域，脸部非五官区域。本实施例中，所述图片还原系统10可以基于位置信息对脸部非五官区域进行划分，例如将眼眉区域上方的非五官区域作为额头区域，将嘴唇区域下方的非五官区域作为下巴区域，将除额头区域、下巴区域外的非五官区域作为中庭区域等。
每个待处理区域包含一个人体单元，所述人体单元包括但不限于眉毛，鼻子，眼睛，嘴巴，脸部非五官单元(如脸颊)。例如，眉毛区域包含两条眉毛，鼻子区域包含一个鼻子等。后续对待处理区域中的每个人体单元进行还原操作。
在其他实施例中，所述多个待处理区域也可以包括人体的其他单元，例如，手，胳膊等。
S12，图片还原系统10提取所述多个待处理区域中每个待处理区域中
的人体单元信息。
在本实施例中，所述人体单元信息包括以下至少一种：人体单元的位置、人体单元的轮廓、人体单元的色彩、所述人体单元所在的待处理区域的皮肤纹理。所述人体单元的位置为人体单元在待还原图片中的位置，所述人体单元的轮廓包括人体单元的形状，大小等。所述人体单元的色彩包括人体单元的色度，亮度等。所述皮肤纹理包括皮肤颜色、毛孔的大小、皱纹的粗细等。
在本实施例中，对任意一个待处理区域而言，先利用图片识别技术识别出所述待处理区域中的人体单元，再利用特征提取技术提取所述待处理区域中的人体单元信息。
例如，利用图片识别技术从鼻子区域中确定鼻子所在的位置，然后利用特征提取技术提取鼻子的位置，轮廓(即鼻梁和鼻翼)及鼻子上的皮肤纹理等。利用图片识别技术从眼睛区域中确定眼睛所在的位置，然后利用特征提取技术提取眼睛的位置，轮廓(包括眼眶、眼珠、眼睫毛)及眼睛的像素灰度等。
在本发明其他实施例中，图片还原系统10还可提取每个待处理区域中的噪点信息。以待逆美颜图片为例，由于图片经过美颜处理后，图片上被美颜部分的噪点也会减少，因此，后续可以对比待处理区域与其他区域的噪点信息来还原待处理区域的人体单元。
在本发明其他实施例中，图片还原系统10还可提取每个待处理区域的特殊标记。每种特殊标记对应一种标准形状。所述特殊标记包括但不限于，变形的标准几何形状。例如，变形的直线、变形的圆、变形的椭圆、变形的矩形等。例如，以待逆美颜图片为例，当图片中人脸上有阳光时，所述阳光为直线形状，在美颜的瘦脸处理中，所述直线的某一部分可能会变形(如凸起)，提取变形的直线。后续可根据对变形直线的修正来还原美颜后的脸部。当特殊标记为变形的标准几何形状时，所述特殊标记对应的标准形状可以预先存储于存储器20中。
所述图片还原系统10可以利用特征提取技术在所述待处理区域中直接提取所述特殊标记。图片还原系统10也可以接收用户在所述待处理区域中选择的一个区域，利用特征提取技术在所选择的区域中提取一个特殊标
记。后续对所述特殊标记进行还原，从而得到还原参数，根据所述还原参数还原待处理区域中的人体单元。
当然，在其他实施例中，所述特殊标记也可以是用户根据待还原图片自己定义的其他形状。为了修正所述特殊标记，图片还原系统10还需接收用户输入的所述特殊标记对应的标准形状。
S13，图片还原系统10提取所述每个待处理区域对应的对比区域的特征信息。
在本实施例中，每个待处理区域至少对应一个对比区域。以待逆美颜图片为例，在美颜图片时，一般为了保持图片的真实性，边缘区域的细节特征，例如鼻子的边缘区域、眼睛的边缘区域等都是保留的。另外，背景区域一般是不作美颜处理。因此，所述对比区域可以是以下任一种或者多种形式：每个待处理区域的边缘区域、待还原图片的背景区域、人体区域中除去所述待处理区域的其他区域，例如，所述待处理区域为鼻子区域，人体区域中除去所述待处理区域的其他区域可以是中庭区域，也可以是下巴区域等。
所述对比区域的特征信息包括以下至少一种：对比区域的噪点信息、对比区域的色彩、对比区域的皮肤纹理。所述对比区域的色彩包括对比区域的色度，亮度等。所述对比区域的皮肤纹理包括对比区域中的皮肤颜色、毛孔的大小、皱纹的粗细等。后续通过对比每个待处理区域中的人体单元信息与所述待处理区域对应的对比区域的特征信息的差异来还原所述待处理区域中的人体单元。
S14，所述图片还原系统10根据所述每个待处理区域所包含的人体单元信息及所述每个待处理区域对应的对比区域的特征信息，对所述每个待处理区域的人体单元进行还原处理。
在本实施例中，所述图片还原系统10根据所述每个待处理区域所包含的人体单元信息及所述每个待处理区域对应的对比区域的特征信息确定待处理区域的人体单元的还原方向及还原幅度。例如，以待逆美颜图片为例，人体单元的还原方向为眼睛变小，皮肤变暗淡，脸部区域变大等等。
优选地，图片还原系统10针对每个待处理区域，根据所述待处理区域的皮肤纹理或噪点信息与所述待处理区域对应的对比区域的皮肤纹理或噪
点信息确定所述待处理区域中人体单元的轮廓、人体单元的皮肤纹理所对应的还原方向及还原幅度。进一步地，所述人体单元的轮廓、人体单元的皮肤纹理所对应的还原方向及还原幅度的确定可以通过以下任一种或者多种的组合方式来实现：
(1)根据所述待处理区域的毛孔大小与所述待处理区域对应的对比区域的毛孔大小来确定人体单元的轮廓、人体单元的皮肤纹理所对应的还原方向及还原幅度。
(2)根据所述待处理区域的毛孔密度与所述待处理区域对应的对比区域的毛孔密度来确定人体单元的轮廓、人体单元的皮肤纹理所对应的还原方向及还原幅度。
(3)根据所述待处理区域的噪点大小与所述待处理区域对应的对比区域的噪点大小来确定人体单元的轮廓、人体单元的皮肤纹理所对应的还原方向及还原幅度。
(4)根据所述待处理区域的噪点密度与所述待处理区域对应的对比区域的噪点密度来确定人体单元的轮廓、人体单元的皮肤纹理所对应的还原方向及还原幅度。
(5)根据所述待处理区域的皱纹粗细与所述待处理区域对应的对比区域的皱纹粗细来确定人体单元的轮廓、人体单元的皮肤纹理所对应的还原方向及还原幅度。
举例而言，针对鼻子区域，所述鼻子区域的对比区域为鼻子的边缘区域。通过提取所述边缘区域的毛孔，计算所述边缘区域的毛孔的平均大小为A。提取鼻子区域中的毛孔，计算所述鼻子区域中的毛孔的平均大小为B。若A大于B，则鼻子区域的轮廓所对应的还原方向为从小到大，若A小于B，则鼻子的轮廓所对应的还原方向为从大到小。鼻子的轮廓所对应的还原幅度为A/B。当然，也可以通过计算边缘区域的毛孔的密度与鼻子上毛孔的密度来确定鼻子的轮廓所对应的还原方向及还原幅度。
在其他实施例中，图片还原系统10同时利用每个待处理区域对应的对比区域的皮肤纹理及噪点信息确定所述每个待处理区域中人体单元的轮廓、人体单元的皮肤纹理所对应的还原方向及还原幅度。
具体地，根据所述待处理区域的皮肤纹理及噪点信息与所述待处理区
域对应的对比区域的皮肤纹理及噪点信息确定所述待处理区域中人体单元的轮廓、人体单元的皮肤纹理所对应的还原方向及还原幅度。例如，根据所述待处理区域的毛孔大小计算出毛孔大小平均值，根据所述待处理区域的噪点大小计算出噪点大小平均值，将所述待处理区域的毛孔大小平均值及噪点大小平均值进行加权得到所述待处理区域对应的加权值。同理计算所述待处理区域对应的对比区域的毛孔大小平均值及噪点大小平均值，将所述待处理区域对应的对比区域的毛孔大小平均值及噪点大小平均值进行加权得到所述待处理区域对应的对比区域对应的加权值。根据所述待处理区域对应的加权值及所述待处理区域对应的对比区域对应的加权值确定所述待处理区域中人体单元的轮廓、人体单元的皮肤纹理所对应的还原方向及还原幅度。
在其他实施例中，当所述待处理区域有特殊标记时，可以根据所述特殊标记相对于其对应的标准形状的弯曲方向，确定所述待处理区域的还原方向。
优选地，图片还原系统10针对每个待处理区域，根据所述待处理区域的色彩与所述待处理区域对应的对比区域的色彩调整所述待处理区域中人体单元的色彩。
进一步地，针对每个待处理区域，所述待处理区域中人体单元的色彩的调整可以通过以下任一种或者多种的组合方式来实现：
(1)将所述待处理区域的像素值与所述待处理区域对应的对比区域的像素值进行加权，根据加权后得到的像素值确定所述待处理区域中每个像素点的像素值。例如，嘴巴区域的平均像素值为P，嘴巴区域对应的边缘区域的平均像素值为Q，则嘴巴区域中每个像素点的像素值为0.5P+0.5Q。
(2)针对待处理区域中每个像素点，根据所述像素点的多个相邻像素点的像素值计算所述像素点的像素值。所述相邻像素点可能位于所述待处理区域的边缘区域中。例如，待处理区域中的第一个像素点，其相邻的四个像素点分别为像素点a，像素点b，像素点c，像素点d，其中像素点a，像素点b，像素点c位于边缘区域中。则所述第一个像素点的像素值为所述相邻的四个像素点的平均值。
在其他实施例中，当在待处理区域中提取了特殊标记时，所述图片还
原系统10获取所述特殊标记所对应的标准形状；根据所述特殊标记及其对应的标准形状，对所述特殊标记进行还原，并得到还原参数。根据所得到的还原参数对所述待处理区域中的人体单元进行调整。
在本实施例中，图片还原系统10进一步获取肖像分布数据，所述肖像分布数据可以是预先存储于存储器20中，也可以由用户根据待还原图片输入的数据。所述肖像分布数据包括人体单元在人体中的平均位置及人体单元的色彩区间范围。所述图片还原系统10根据所述肖像分布数据对待还原图片中人体单元进行微调。这样可以防止在还原过程中出现错误。
在其他实施例中，图片还原系统10还可以获取所述待还原图片中用户的其他图片，所述其他图片可以是没有美颜的图片。然后根据所述其他图片获取肖像分布数据。所述图片还原系统10根据所述肖像分布数据对待还原图片中人体单元进行微调及验证。这样可以防止在还原过程中出现错误。
本发明所述图片还原系统10能根据待还原图片中的特征，将待还原图片中的人体单元进行还原，而不是简单的做丑化处理。如图4所示，利用本发明将一张美颜照片进行逆美颜，逆美颜后的照片更真实。这样用户可以看到美颜图片中被拍者的真实容貌。
如图3所示，是本发明图片还原系统较佳实施例的功能模块图。所述图片还原系统10包括获取模块100、划分模块101、提取模块102及还原模块103。
获取模块100用于获取待还原图片。
在本实施例中，所述待还原图片包括但不限于，待逆美颜图片，也可以是其他形式包括人体单元的图片。以待逆美颜图片为例，所述获取模块100接收到用户对一张待逆美颜图片进行还原的请求时，所述获取模块100获取所述待逆美颜图片。
当用户需要对一张待还原图片进行还原时，可以在图片还原系统10提供的用户界面上选择还原功能选项。
在其他实施例中，以待逆美颜图片为例，用户可以将待逆美颜图片显示在当前用户界面上，然后语音输入“逆美颜”，所述获取模块100接收到所述语音后，通过语音识别技术识别出“逆美颜”，然后获取所述待逆美颜图片。
划分模块101用于将所述待还原图片划分成多个待处理区域。
在本实施例中，所述划分模块101利用图片识别技术将所述待还原图片先划分成背景区域与人体区域，再将人体区域划分成多个待处理区域。现实生活中，所述多个待处理区域包括但不限于，眉毛区域，鼻子区域，眼睛区域，嘴巴区域，脸部非五官区域。本实施例中，所述划分模块101可以基于位置信息对脸部非五官区域进行划分，例如将眼眉区域上方的非五官区域作为额头区域，将嘴唇区域下方的非五官区域作为下巴区域，将除额头区域、下巴区域外的非五官区域作为中庭区域等。
每个待处理区域包含一个人体单元，所述人体单元包括但不限于，眉毛，鼻子，眼睛，嘴巴，脸部非五官单元(如脸颊)。例如，眉毛区域包含两条眉毛，鼻子区域包含一个鼻子等。后续对待处理区域中的每个人体单元进行还原操作。
在其他实施例中，所述多个待处理区域也可以包括人体的其他单元，例如，手，胳膊等。
提取模块102用于提取所述多个待处理区域中每个待处理区域中的人体单元信息。
在本实施例中，所述人体单元信息包括以下至少一种：人体单元的位置、人体单元的轮廓、人体单元的色彩、所述人体单元所在的待处理区域的皮肤纹理。所述人体单元的位置为人体单元在待还原图片中的位置，所述人体单元的轮廓包括人体单元的形状，大小等。所述人体单元的色彩包括人体单元的色度，亮度等。所述皮肤纹理包括皮肤颜色、毛孔的大小、皱纹的粗细等。
在本实施例中，对任意一个待处理区域而言，先利用图片识别技术识别出所述待处理区域中的人体单元，再利用特征提取技术提取所述待处理区域中的人体单元信息。
例如，利用图片识别技术从鼻子区域中确定鼻子所在的位置，然后利用特征提取技术提取鼻子的位置，轮廓(即鼻梁和鼻翼)及鼻子上的皮肤纹理等。利用图片识别技术从眼睛区域中确定眼睛所在的位置，然后利用特征提取技术提取眼睛的位置，轮廓(包括眼眶、眼珠、眼睫毛)及眼睛的像素灰度等。
在本发明其他实施例中，所述提取模块102还可提取每个待处理区域中的噪点信息。以待逆美颜图片为例，由于图片经过美颜处理后，图片上被美颜部分的噪点也会减少，因此，后续可以对比待处理区域与其他区域的噪点信息来还原待处理区域的人体单元。
在本发明其他实施例中，所述提取模块102还可提取每个待处理区域的特殊标记。每种特殊标记对应一种标准形状。所述特殊标记包括但不限于，变形的标准几何形状。例如，变形的直线、变形的圆、变形的椭圆、变形的矩形等。例如，以待逆美颜图片为例，当图片中人脸上有阳光时，所述阳光为直线形状，在美颜的瘦脸处理中，所述直线的某一部分可能会变形(如凸起)，提取变形的直线。后续可根据对变形直线的修正来还原美颜后的脸部。当特殊标记为变形的标准几何形状时，所述特殊标记对应的标准形状可以预先存储于存储器20中。
所述提取模块102可以利用特征提取技术在所述待处理区域中直接提取所述特殊标记。图片还原系统10也可以接收用户在所述待处理区域中选择的一个区域，利用特征提取技术在所选择的区域中提取一个特殊标记。后续对所述特殊标记进行还原，从而得到还原参数，根据所述还原参数还原待处理区域中的人体单元。
当然，在其他实施例中，所述特殊标记也可以是用户根据待还原图片自己定义的其他形状。为了修正所述特殊标记，所述提取模块102还需接收用户输入的所述特殊标记对应的标准形状。
所述提取模块102还用于提取所述每个待处理区域对应的对比区域的特征信息。
在本实施例中，每个待处理区域至少对应一个对比区域。以待逆美颜图片为例，在美颜图片时，一般为了保持图片的真实性，边缘区域的细节特征，例如鼻子的边缘区域、眼睛的边缘区域等都是保留的。另外，背景区域一般是不作美颜处理。因此，所述对比区域可以是以下任一种或者多种形式：每个待处理区域的边缘区域、待还原图片的背景区域、人体区域中除去所述待处理区域的其他区域，例如，所述待处理区域为鼻子区域，人体区域中除去所述待处理区域的其他区域可以是中庭区域，也可以是下巴区域等。
所述对比区域的特征信息包括以下至少一种：对比区域的噪点信息、对比区域的色彩、对比区域的皮肤纹理。所述对比区域的色彩包括对比区域的色度，亮度等。所述对比区域的皮肤纹理包括对比区域中的皮肤颜色、毛孔的大小、皱纹的粗细等。后续通过对比每个待处理区域中的人体单元信息与所述待处理区域对应的对比区域的特征信息的差异来还原所述待处理区域中的人体单元。
还原模块103用于根据所述每个待处理区域所包含的人体单元信息及所述每个待处理区域对应的对比区域的特征信息，对所述每个待处理区域的人体单元进行还原处理。
在本实施例中，所述还原模块103根据所述每个待处理区域所包含的人体单元信息及所述每个待处理区域对应的对比区域的特征信息确定待处理区域的人体单元的还原方向及还原幅度。例如，以待逆美颜图片为例，人体单元的还原方向为眼睛变小，皮肤变暗淡，脸部区域变大等等。
优选地，所述还原模块103针对每个待处理区域，根据所述待处理区域的皮肤纹理或噪点信息与所述待处理区域对应的对比区域的皮肤纹理或噪点信息确定所述待处理区域中人体单元的轮廓、人体单元的皮肤纹理所对应的还原方向及还原幅度。
进一步地，所述人体单元的轮廓、人体单元的皮肤纹理所对应的还原方向及还原幅度的确定可以通过以下任一种或者多种的组合方式来实现：
(1)根据所述待处理区域的毛孔大小与所述待处理区域对应的对比区域的毛孔大小来确定人体单元的轮廓、人体单元的皮肤纹理所对应的还原方向及还原幅度。
(2)根据所述待处理区域的毛孔密度与所述待处理区域对应的对比区域的毛孔密度来确定人体单元的轮廓、人体单元的皮肤纹理所对应的还原方向及还原幅度。
(3)根据所述待处理区域的噪点大小与所述待处理区域对应的对比区域的噪点大小来确定人体单元的轮廓、人体单元的皮肤纹理所对应的还原方向及还原幅度。
(4)根据所述待处理区域的噪点密度与所述待处理区域对应的对比区域的噪点密度来确定人体单元的轮廓、人体单元的皮肤纹理所对应的还原
方向及还原幅度。
(5)根据所述待处理区域的皱纹粗细与所述待处理区域对应的对比区域的皱纹粗细来确定人体单元的轮廓、人体单元的皮肤纹理所对应的还原方向及还原幅度。
举例而言，针对鼻子区域，所述鼻子区域的对比区域为鼻子的边缘区域。通过提取所述边缘区域的毛孔，计算所述边缘区域的毛孔的平均大小为A。提取鼻子区域中的毛孔，计算所述鼻子区域中的毛孔的平均大小为B。若A大于B，则鼻子区域的轮廓所对应的还原方向为从小到大，若A小于B，则鼻子的轮廓所对应的还原方向为从大到小。鼻子的轮廓所对应的还原幅度为A/B。当然，也可以通过计算边缘区域的毛孔的密度与鼻子上毛孔的密度来确定鼻子的轮廓所对应的还原方向及还原幅度。
在其他实施例中，所述还原模块103同时利用每个待处理区域对应的对比区域的皮肤纹理及噪点信息确定所述每个待处理区域中人体单元的轮廓、人体单元的皮肤纹理所对应的还原方向及还原幅度。具体地，根据所述待处理区域的皮肤纹理及噪点信息与所述待处理区域对应的对比区域的皮肤纹理及噪点信息确定所述待处理区域中人体单元的轮廓、人体单元的皮肤纹理所对应的还原方向及还原幅度。
例如，根据所述待处理区域的毛孔大小计算出毛孔大小平均值，根据所述待处理区域的噪点大小计算出噪点大小平均值，将所述待处理区域的毛孔大小平均值及噪点大小平均值进行加权得到所述待处理区域对应的加权值。同理计算所述待处理区域对应的对比区域的毛孔大小平均值及噪点大小平均值，将所述待处理区域对应的对比区域的毛孔大小平均值及噪点大小平均值进行加权得到所述待处理区域对应的对比区域对应的加权值。根据所述待处理区域对应的加权值及所述待处理区域对应的对比区域对应的加权值确定所述待处理区域中人体单元的轮廓、人体单元的皮肤纹理所对应的还原方向及还原幅度。
在其他实施例中，当所述待处理区域有特殊标记时，可以根据所述特殊标记相对于其对应的标准形状的弯曲方向，确定所述待处理区域的还原方向。
优选地，所述还原模块103针对每个待处理区域，根据所述待处理区
域的色彩与所述待处理区域对应的对比区域的色彩调整所述待处理区域中人体单元的色彩。
进一步地，针对每个待处理区域，所述待处理区域中人体单元的色彩的调整可以通过以下任一种或者多种的组合方式来实现：
(1)将所述待处理区域的像素值与所述待处理区域对应的对比区域的像素值进行加权，根据加权后得到的像素值确定所述待处理区域中每个像素点的像素值。例如，嘴巴区域的平均像素值为P，嘴巴区域对应的边缘区域的平均像素值为Q，则嘴巴区域中每个像素点的像素值为0.5P+0.5Q。
(2)针对待处理区域中每个像素点，根据所述像素点的多个相邻像素点的像素值计算所述像素点的像素值。所述相邻像素点可能位于所述待处理区域的边缘区域中。例如，待处理区域中的第一个像素点，其相邻的四个像素点分别为像素点a，像素点b，像素点c，像素点d，其中像素点a，像素点b，像素点c位于边缘区域中。则所述第一个像素点的像素值为所述相邻的四个像素点的平均值。
在其他实施例中，当在待处理区域中提取了特殊标记时，所述还原模块103获取所述特殊标记所对应的标准形状；根据所述特殊标记及其对应的标准形状，对所述特殊标记进行还原，并得到还原参数。根据所得到的还原参数对所述待处理区域中的人体单元进行调整。
在本实施例中，所述还原模块103进一步获取肖像分布数据，所述肖像分布数据可以是预先存储于存储器20中，也可以由用户根据待还原图片输入的数据。所述肖像分布数据包括人体单元在人体中的平均位置及人体单元的色彩区间范围。所述还原模块103根据所述肖像分布数据对待还原图片中人体单元进行微调。这样可以防止在还原过程中出现错误。
在其他实施例中，所述还原模块103还可以获取所述待还原图片中用户的其他图片，所述其他图片可以是没有美颜的图片。然后根据所述其他图片获取肖像分布数据。所述还原模块103根据所述肖像分布数据对待还原图片中人体单元进行微调及验证。这样可以防止在还原过程中出现错误。
本发明能根据待还原图片中的特征，将待还原图片中的人体单元进行还原，而不是简单的做丑化处理。如图4所示，利用本发明将一张美颜照片进行逆美颜，逆美颜后的照片更真实。这样用户可以看到美颜图片中被
拍者的真实容貌。
在本发明所提供的几个实施例中，应所述理解到，所揭露的系统，装置和方法，可以通过其它的方式实现。例如，以上所描述的装置实施例仅仅是示意性的，例如，所述模块的划分，仅仅为一种逻辑功能划分，实际实现时可以有另外的划分方式。
另外，在本发明各个实施例中的各功能模块可以集成在一个处理单元中，也可以是各个单元单独物理存在，也可以两个或两个以上单元集成在一个单元中。上述集成的单元既可以采用硬件的形式实现，也可以采用硬件加软件功能模块的形式实现。
上述以软件功能模块的形式实现的集成的单元，可以存储在一个计算机可读取存储介质中。上述软件功能模块存储在一个存储介质中，包括若干指令用以使得一台计算机设备(可以是个人计算机，服务器，或者网络设备等)或处理器(processor)执行本发明各个实施例所述方法的部分步骤。
对于本领域技术人员而言，显然本发明不限于上述示范性实施例的细节，而且在不背离本发明的精神或基本特征的情况下，能够以其他的具体形式实现本发明。因此，无论从哪一点来看，均应将实施例看作是示范性的，而且是非限制性的，本发明的范围由所附权利要求而不是上述说明限定，因此旨在将落在权利要求的等同要件的含义和范围内的所有变化涵括在本发明内。不应将权利要求中的任何附图标记视为限制所涉及的权利要求。此外，显然“包括”一词不排除其他单元或步骤，单数不排除复数。系统权利要求中陈述的多个单元或装置也可以由一个单元或装置通过软件或者硬件来实现。第一，第二等词语用来表示名称，而并不表示任何特定的顺序。
最后应说明的是，以上实施例仅用以说明本发明的技术方案而非限制，尽管参照较佳实施例对本发明进行了详细说明，本领域的普通技术人员应当理解，可以对本发明的技术方案进行修改或等同替换，而不脱离本发明技术方案的精神和范围。
Claims (12)
- 一种图片还原方法，其特征在于，所述方法包括：获取待还原图片；将所述待还原图片划分成多个待处理区域；提取所述多个待处理区域中的每个待处理区域中的人体单元信息；提取所述每个待处理区域对应的对比区域的特征信息；及根据所述每个待处理区域所包含的人体单元信息及所述每个待处理区域对应的对比区域的特征信息，对所述每个待处理区域的人体单元进行还原处理。
- 如权利要求1所述的图片还原方法，其特征在于，所述待还原图片包括背景区域及人体区域；所述对比区域包括以下任一种或者多种的组合：背景区域、所述每个待处理区域的边缘区域、人体区域中除去待处理区域的其他区域。
- 如权利要求1所述的图片还原方法，其特征在于，所述人体单元信息包括人体单元的轮廓、所述人体单元所在的待处理区域的皮肤纹理；所述特征信息包括对比区域的噪点信息、对比区域的皮肤纹理；所述皮肤纹理包括毛孔的大小、毛孔的密度、皱纹的粗细；及所述每个待处理区域中人体单元的轮廓及皮肤纹理可通过以下任一种或者多种方式进行还原：根据所述待处理区域的毛孔大小与所述待处理区域对应的对比区域的毛孔大小来确定所述人体单元的轮廓、所述人体单元的皮肤纹理所对应的还原方向及还原幅度；或根据所述待处理区域的毛孔密度与所述待处理区域对应的对比区域的毛孔密度来确定所述人体单元的轮廓、所述人体单元的皮肤纹理所对应的还原方向及还原幅度；或提取所述待处理区域的噪点，根据所述待处理区域的噪点大小与所述待处理区域对应的对比区域的噪点大小来确定人体单元的轮廓、人体单元的皮肤纹理所对应的还原方向及还原幅度；或提取所述待处理区域的噪点，根据所述待处理区域的噪点密度与所述待处理区域对应的对比区域的噪点密度来确定所述人体单元的轮廓、所述人体单元 的皮肤纹理所对应的还原方向及还原幅度；或根据所述待处理区域的皱纹粗细与所述待处理区域对应的对比区域的皱纹粗细来确定所述人体单元的轮廓、所述人体单元的皮肤纹理所对应的还原方向及还原幅度。
- 如权利要求1所述的图片还原方法，其特征在于，所述人体单元信息包括人体单元的色彩；所述特征信息包括对比区域的色彩；及所述每个待处理区域中人体单元的色彩可通过以下任一种或者多种方式进行还原：将所述待处理区域的像素值与所述待处理区域对应的对比区域的像素值进行加权，根据加权后的像素值确定所述待处理区域中每个像素点的像素值；或针对所述待处理区域中每个像素点，根据像素点的多个相邻像素点的像素值计算所述像素点的像素值。
- 如权利要求1所述的图片还原方法，其特征在于，所述方法还包括：提取所述每个待处理区域的特殊标记，所述特殊标记包括变形的标准几何形状、根据待还原图片自定义的形状；根据所述每个待处理区域的特殊标记，对所述每个待处理区域的人体单元进行调整。
- 如权利要求5所述的图片还原方法，其特征在于，对所述每个待处理区域的人体单元进行调整包括：获取所述待处理区域的特殊标记所对应的标准形状；根据所述特殊标记及其对应的标准形状，对所述特殊标记进行还原，并得到还原参数；及根据所得到的还原参数对所述待处理区域中的人体单元进行调整。
- 一种图片还原系统，其特征在于，所述系统包括：获取模块，用于获取待还原图片；划分模块，用于将所述待还原图片划分成多个待处理区域；提取模块，用于提取所述多个待处理区域中的每个待处理区域中的人体单元信息；所述提取模块还用于提取所述每个待处理区域对应的对比区域的特征信息；及还原模块，用于根据所述每个待处理区域所包含的人体单元信息及所述每个待处理区域对应的对比区域的特征信息，对所述每个待处理区域的人体单元进行还原处理。
- 如权利要求7所述的图片还原系统，其特征在于，所述待还原图片包括背景区域及人体区域；所述对比区域包括以下任一种或者多种的组合：背景区域、所述每个待处理区域的边缘区域、人体区域中除去待处理区域的其他区域。
- 如权利要求7所述的图片还原系统，其特征在于，所述人体单元信息包括人体单元的轮廓、所述人体单元所在的待处理区域的皮肤纹理；所述特征信息包括对比区域的噪点信息、对比区域的皮肤纹理；所述皮肤纹理包括毛孔的大小、毛孔的密度、皱纹的粗细；及所述每个待处理区域中人体单元的轮廓及皮肤纹理可通过以下任一种或者多种方式进行还原：根据所述待处理区域的毛孔大小与所述待处理区域对应的对比区域的毛孔大小来确定所述人体单元的轮廓、所述人体单元的皮肤纹理所对应的还原方向及还原幅度；或根据所述待处理区域的毛孔密度与所述待处理区域对应的对比区域的毛孔密度来确定所述人体单元的轮廓、所述人体单元的皮肤纹理所对应的还原方向及还原幅度；或提取所述待处理区域的噪点，根据所述待处理区域的噪点大小与所述待处理区域对应的对比区域的噪点大小来确定人体单元的轮廓、人体单元的皮肤纹理所对应的还原方向及还原幅度；或提取所述待处理区域的噪点，根据所述待处理区域的噪点密度与所述待处理区域对应的对比区域的噪点密度来确定所述人体单元的轮廓、所述人体单元的皮肤纹理所对应的还原方向及还原幅度；或根据所述待处理区域的皱纹粗细与所述待处理区域对应的对比区域的皱纹粗细来确定所述人体单元的轮廓、所述人体单元的皮肤纹理所对应的还原方向及还原幅度。
- 如权利要求7所述的图片还原系统，其特征在于，所述人体单元信息包括人体单元的色彩；所述特征信息包括对比区域的色彩；及所述每个待处理区域中人体单元的色彩可通过以下任一种或者多种方式进行还原：将所述待处理区域的像素值与所述待处理区域对应的对比区域的像素值进行加权，根据加权后的像素值确定所述待处理区域中每个像素点的像素值；或针对所述待处理区域中每个像素点，根据像素点的多个相邻像素点的像素值计算所述像素点的像素值。
- 如权利要求7所述的图片还原系统，其特征在于，所述系统还包括：所述提取模块还用于提取所述每个待处理区域的特殊标记，所述特殊标记包括变形的标准几何形状、根据待还原图片自定义的形状；所述还原模块还用于根据所述每个待处理区域的特殊标记，对所述每个待处理区域的人体单元进行调整。
- 如权利要求11所述的图片还原系统，其特征在于，该系统还包括：所述获取模块还用于获取所述待处理区域的特殊标记所对应的标准形状；所述还原模块还用于根据所述特殊标记及其对应的标准形状，对所述特殊标记进行还原，并得到还原参数；及所述还原模块还用于根据所得到的还原参数对所述待处理区域中的人体单元进行调整。
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610274328.1 | 2016-04-27 | ||
| CN201610274328.1A CN105931204B (zh) | 2016-04-27 | 2016-04-27 | 图片还原方法及系统 |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| WO2017185452A1 true WO2017185452A1 (zh) | 2017-11-02 |
Family
ID=56837647
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| PCT/CN2016/084060 WO2017185452A1 (zh) | 2016-04-27 | 2016-05-31 | 图片还原方法及系统 |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN105931204B (zh) |
| WO (1) | WO2017185452A1 (zh) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110895789A (zh) * | 2018-09-13 | 2020-03-20 | 杭州海康威视数字技术股份有限公司 | 一种人脸美颜方法及装置 |
| US11507056B1 (en) * | 2020-04-06 | 2022-11-22 | Lockheed Martin Corporation | System and method for three-dimensional (3D) computer-aided manufacturing (CAM) of an ensemble of pilot equipment and garments |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106713700A (zh) * | 2016-12-08 | 2017-05-24 | 宇龙计算机通信科技(深圳)有限公司 | 图像处理方法、装置和终端 |
| CN106780315A (zh) * | 2016-12-30 | 2017-05-31 | 李聪 | 虚拟整容软件方法 |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20090060383A1 (en) * | 2007-08-27 | 2009-03-05 | Arcsoft, Inc. | Method of restoring closed-eye portrait photo |
| CN104899905A (zh) * | 2015-06-08 | 2015-09-09 | 深圳市诺比邻科技有限公司 | 人脸图像处理方法及装置 |
| CN105354810A (zh) * | 2015-12-07 | 2016-02-24 | 厦门美图之家科技有限公司 | 一种图像祛斑处理的方法、系统及拍摄终端 |
| CN105469356A (zh) * | 2015-11-23 | 2016-04-06 | 小米科技有限责任公司 | 人脸图像处理方法及装置 |
| CN105512605A (zh) * | 2015-11-23 | 2016-04-20 | 小米科技有限责任公司 | 人脸图像处理方法及装置 |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2015127394A1 (en) * | 2014-02-23 | 2015-08-27 | Northeastern University | System for beauty, cosmetic, and fashion analysis |
| CN105069007B (zh) * | 2015-07-02 | 2018-01-19 | 广东欧珀移动通信有限公司 | 一种建立美颜数据库的方法及装置 |
| CN104992402B (zh) * | 2015-07-02 | 2019-04-09 | Oppo广东移动通信有限公司 | 一种美颜处理方法及装置 |
| CN104966267B (zh) * | 2015-07-02 | 2018-01-19 | 广东欧珀移动通信有限公司 | 一种美颜用户图像的方法及装置 |
| CN105120157B (zh) * | 2015-08-25 | 2018-04-17 | 广东欧珀移动通信有限公司 | 图像的处理方法及装置 |
-
2016
- 2016-04-27 CN CN201610274328.1A patent/CN105931204B/zh not_active Expired - Fee Related
- 2016-05-31 WO PCT/CN2016/084060 patent/WO2017185452A1/zh active Application Filing
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20090060383A1 (en) * | 2007-08-27 | 2009-03-05 | Arcsoft, Inc. | Method of restoring closed-eye portrait photo |
| CN104899905A (zh) * | 2015-06-08 | 2015-09-09 | 深圳市诺比邻科技有限公司 | 人脸图像处理方法及装置 |
| CN105469356A (zh) * | 2015-11-23 | 2016-04-06 | 小米科技有限责任公司 | 人脸图像处理方法及装置 |
| CN105512605A (zh) * | 2015-11-23 | 2016-04-20 | 小米科技有限责任公司 | 人脸图像处理方法及装置 |
| CN105354810A (zh) * | 2015-12-07 | 2016-02-24 | 厦门美图之家科技有限公司 | 一种图像祛斑处理的方法、系统及拍摄终端 |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110895789A (zh) * | 2018-09-13 | 2020-03-20 | 杭州海康威视数字技术股份有限公司 | 一种人脸美颜方法及装置 |
| CN110895789B (zh) * | 2018-09-13 | 2023-05-02 | 杭州海康威视数字技术股份有限公司 | 一种人脸美颜方法及装置 |
| US11507056B1 (en) * | 2020-04-06 | 2022-11-22 | Lockheed Martin Corporation | System and method for three-dimensional (3D) computer-aided manufacturing (CAM) of an ensemble of pilot equipment and garments |
| US20220390921A1 (en) * | 2020-04-06 | 2022-12-08 | Lockheed Martin Corporation | System and method for three-dimensional (3d) computer-aided manufacturing (cam) of an ensemble of pilot equipment and garments |
| US11822312B2 (en) | 2020-04-06 | 2023-11-21 | Lockheed Martin Corporation | System and method for three-dimensional (3D) computer-aided manufacturing (CAM) of an ensemble of pilot equipment and garments |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105931204A (zh) | 2016-09-07 |
| CN105931204B (zh) | 2019-10-11 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11893828B2 (en) | System and method for image de-identification | |
| US10997445B2 (en) | Facial recognition-based authentication | |
| CN106682632B (zh) | 用于处理人脸图像的方法和装置 | |
| TWI714225B (zh) | 注視點判斷方法和裝置、電子設備和電腦儲存介質 | |
| KR102359558B1 (ko) | 얼굴 인증 방법 및 장치 | |
| WO2017185452A1 (zh) | 图片还原方法及系统 | |
| WO2016141866A1 (zh) | 图像处理设备和方法 | |
| KR20190053602A (ko) | 얼굴 인증 방법 및 장치 | |
| WO2018177364A1 (zh) | 一种滤镜的实现方法及装置 | |
| WO2017058733A1 (en) | Head-mounted display with facial expression detecting capability | |
| CN110612530B (zh) | 用于选择脸部处理中使用的帧的方法 | |
| CN109598210B (zh) | 一种图片处理方法和装置 | |
| CN111008935B (zh) | 一种人脸图像增强方法、装置、系统及存储介质 | |
| CN104157001A (zh) | 一种头部漫画的绘制方法及装置 | |
| CN108227925A (zh) | 一种坐姿调整的方法、装置、设备及存储介质 | |
| CN114187166A (zh) | 图像处理方法、智能终端及存储介质 | |
| CN107346544B (zh) | 一种图像处理方法和电子设备 | |
| CN114862729A (zh) | 图像处理方法、装置、计算机设备和存储介质 | |
| CN113556471B (zh) | 证件照生成方法、系统及计算机可读存储介质 | |
| WO2022034779A1 (ja) | 画像処理装置および画像処理方法 | |
| CN114972014A (zh) | 图像处理方法、装置和电子设备 | |
| Lim et al. | C-Auth: Exploring the Feasibility of Using Egocentric View of Face Contour for User Authentication on Glasses | |
| CN108537552B (zh) | 基于透镜的支付方法、装置及系统 | |
| CN113421197B (zh) | 一种美颜图像的处理方法及其处理系统 | |
| US9563940B2 (en) | Smart image enhancements |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| NENP | Non-entry into the national phase |
Ref country code: DE |
|
| 121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 16899962 Country of ref document: EP Kind code of ref document: A1 |
|
| 122 | Ep: pct application non-entry in european phase |
Ref document number: 16899962 Country of ref document: EP Kind code of ref document: A1 |