WO2016119339A1 - 图像修正方法、图像修正装置和视频系统 - Google Patents
图像修正方法、图像修正装置和视频系统 Download PDFInfo
- Publication number
- WO2016119339A1 WO2016119339A1 PCT/CN2015/079165 CN2015079165W WO2016119339A1 WO 2016119339 A1 WO2016119339 A1 WO 2016119339A1 CN 2015079165 W CN2015079165 W CN 2015079165W WO 2016119339 A1 WO2016119339 A1 WO 2016119339A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- image
- actual face
- eye region
- face image
- sample image
- Prior art date
Links
- 238000003702 image correction Methods 0.000 title claims abstract description 66
- 238000000034 method Methods 0.000 title claims abstract description 41
- 210000001747 pupil Anatomy 0.000 claims abstract description 62
- 238000013507 mapping Methods 0.000 claims description 19
- 238000012937 correction Methods 0.000 claims description 11
- 238000012545 processing Methods 0.000 claims description 10
- 238000004364 calculation method Methods 0.000 claims description 9
- 238000007476 Maximum Likelihood Methods 0.000 claims description 7
- 230000006854 communication Effects 0.000 abstract description 4
- 238000004891 communication Methods 0.000 abstract description 3
- 210000001508 eye Anatomy 0.000 description 247
- 210000001331 nose Anatomy 0.000 description 20
- 230000000007 visual effect Effects 0.000 description 6
- 238000010586 diagram Methods 0.000 description 3
- 210000000744 eyelid Anatomy 0.000 description 3
- 230000005540 biological transmission Effects 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 210000000887 face Anatomy 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000001953 sensory effect Effects 0.000 description 2
- 230000001427 coherent effect Effects 0.000 description 1
- 238000003708 edge detection Methods 0.000 description 1
- 210000003128 head Anatomy 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 210000000554 iris Anatomy 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 210000003786 sclera Anatomy 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
- G06T5/80—Geometric correction
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2415—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on parametric or probabilistic models, e.g. based on likelihood ratio or false acceptance rate versus a false rejection rate
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/98—Detection or correction of errors, e.g. by rescanning the pattern or by human intervention; Evaluation of the quality of the acquired patterns
- G06V10/993—Evaluation of the quality of the acquired pattern
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/161—Detection; Localisation; Normalisation
- G06V40/165—Detection; Localisation; Normalisation using facial parts and geometric relationships
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/168—Feature extraction; Face representation
- G06V40/171—Local features and components; Facial parts ; Occluding parts, e.g. glasses; Geometrical relationships
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/172—Classification, e.g. identification
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/18—Eye characteristics, e.g. of the iris
- G06V40/193—Preprocessing; Feature extraction
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N7/00—Television systems
- H04N7/14—Systems for two-way working
- H04N7/141—Systems for two-way working between two video terminals, e.g. videophone
- H04N7/142—Constructional details of the terminal equipment, e.g. arrangements of the camera and the display
- H04N7/144—Constructional details of the terminal equipment, e.g. arrangements of the camera and the display camera and display on the same optical axis, e.g. optically multiplexing the camera and display for eye to eye contact
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N7/00—Television systems
- H04N7/14—Systems for two-way working
- H04N7/141—Systems for two-way working between two video terminals, e.g. videophone
- H04N7/147—Communication arrangements, e.g. identifying the communication as a video-communication, intermediate storage of the signals
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N7/00—Television systems
- H04N7/14—Systems for two-way working
- H04N7/15—Conference systems
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30196—Human being; Person
- G06T2207/30201—Face
Definitions
- the present invention relates to the field of display technologies, and in particular, to an image correction method, an image correction device, and a video system.
- video transmission technology has been widely used, for example, video chat.
- the two sides are prone to visual inequality in video communication.
- the eyes are looking at the display screen, and the viewing position of the camera is usually on one side of the display screen. Therefore, the image of the other party that is seen is not always looking at itself, thereby reducing the sensory experience.
- an embodiment of the present invention provides an image correction method, including:
- Each eye region of the actual face image is corrected according to the target sample image such that the orientation of each eye region of the corrected actual face image coincides with the orientation of the corresponding eye region of the target sample image.
- the step of acquiring a target sample image that matches the actual face image may include:
- the feature information may include a central coordinate of each eye region, an aspect ratio of each eye region, a diameter of an iris in each eye region, a distance between a center of the left eye region and a nose, and a center of the right eye region. Any one or combination of the ratio of the distance between the noses, the distance between the center of the left eye region and the mouth, and the ratio of the distance between the center of the right eye region and the mouth.
- Obtaining a target sample image that matches the actual face image and correcting each eye region of the actual face image according to the target sample image such that each eye region of the corrected actual face image The step of matching the orientation of the corresponding eye region of the target sample image further includes:
- the image correction method may further include performing the step of acquiring an actual face image:
- Feature information of each preset sample image is acquired and stored to form the sample library.
- the image correction method may further include performing the step of acquiring and storing the feature information of each preset sample image:
- the step of separately calculating the matching degree between the feature information of the actual face image and the feature information of each preset sample image in the sample library may include:
- a matching degree of the feature information of the actual face image and the feature information of each preset sample image in the category of the preset sample image corresponding thereto is calculated.
- the predetermined rule may include a maximum likelihood classification.
- the image correction method may further include performing the step of acquiring an actual face image:
- the predetermined coordinate range being greater than a distance between central coordinates of two eye regions of the target sample image
- an embodiment of the present invention further provides an image correction apparatus, including:
- An actual image acquisition module configured to acquire an actual face image
- a target sample image obtaining module configured to acquire a target sample image that matches the actual face image
- a processing module configured to correct, according to the target sample image, each eye region of the actual face image such that an orientation of each eye region of the corrected actual face image and a corresponding eye of the target sample image The orientation of the area matches.
- the target sample image obtaining module may include:
- a feature information acquiring unit configured to acquire feature information of the actual face image
- a matching degree calculation unit configured to separately calculate feature information of the actual face image a degree of matching between the feature information of each preset sample image in the sample library, and when the degree of matching between the preset sample image and the actual face image is greater than a preset value, A preset sample image in which a degree of matching between the face images is greater than a preset value is set as the target sample image.
- the feature information includes a central coordinate of each eye region, an aspect ratio of each eye region, a diameter of an iris in each eye region, a distance between a center of the left eye region and the nose, and a center and a nose of the right eye region. Any one or any combination of the ratio of the distances between the centers of the left eye regions and the distance between the centers of the right eye regions and the distance between the centers of the right eye regions and the mouth.
- the image correction device may further include:
- a coordinate acquiring module configured to acquire a center coordinate of a pupil portion of each eye region of the actual face image and a center coordinate of a pupil portion of each eye region in the target sample image;
- a judging module configured to calculate a distance between a center coordinate of a pupil portion of each eye region in the actual face image and a center coordinate of a pupil portion of a corresponding eye region in the target sample image; when the distance is greater than a predetermined
- the processing module is capable of correcting an eye region of the actual face image according to the target sample image such that an orientation of each eye region of the corrected actual face image and a corresponding image of the target sample image The orientation of the eye area coincides.
- the feature information acquiring unit may further acquire feature information of the preset sample image.
- the image correction device may further include:
- a sample image acquisition module configured to collect a plurality of face images as a preset sample image, wherein in each preset sample image, the orientation of each eye region of the collected person faces the sample image acquisition module;
- a storage module configured to store feature information of each preset sample image to form the sample library.
- the image correction device may further include:
- a classification module configured to divide the preset sample image into a plurality of categories according to a predetermined rule
- the target sample image obtaining module may further include:
- a categorizing unit configured to determine a category of the preset sample image corresponding to the actual face image
- the matching degree calculation unit is capable of respectively calculating a matching degree between the feature information of the actual face image and the feature information of each preset sample image in the category of the preset sample image corresponding thereto.
- the predetermined rule may include a maximum likelihood classification.
- the image correction device may further include:
- a pre-positioning module for setting an coordinate axis within a field of view of the camera and setting a predetermined coordinate range within the field of view, the predetermined coordinate range being greater than a distance between central coordinates of two eye regions of the target sample image ;
- the processing module can include:
- a feature point acquiring unit configured to acquire information of a feature point of each eye region in the actual face image and a target feature point of the corresponding target sample image that matches a feature point of each eye region in the actual face image ;
- mapping calculation unit configured to calculate a mapping relationship between a feature point of each eye region in the actual face image and a corresponding target feature point
- a correction unit configured to calculate, according to the mapping relationship, a corrected position of each feature point of each eye region in the actual face image.
- an embodiment of the present invention further provides a video system, including the above image correction apparatus provided by the present invention.
- the image correction method of the present invention may correct each eye region of the actual face image such that the orientation of each eye region of the corrected actual face image is the same as the orientation of the corresponding eye region of the target sample image, Both are facing the camera.
- the actual face image collected by the first party camera is transmitted to the second party, and the second party can use the image correction method to acquire the actual face image of the other party and the eye region thereof.
- display again to make the corrected eye area The pupil portion is facing the camera of the first party, that is, for the second party, the first party user is viewed toward the second party, thereby improving the user's visual experience.
- FIG. 1 is a schematic flow chart of an image correction method in an embodiment of the present invention
- FIG. 3 is a schematic diagram of an eye region of an actual face image
- FIG. 4 is a schematic diagram of an eye region of an actual face image and an eye region of a target sample image
- Fig. 5 is a view showing the configuration of an image correcting device in an embodiment of the present invention.
- an image correction method is provided, as shown in FIG. 1, comprising:
- Each eye region of the actual face image is corrected according to the target sample image such that the orientation of each eye region of the corrected actual face image coincides with the orientation of the corresponding eye region of the target sample image.
- the acquired image is the coordinates and the gradation value of each pixel of the face image.
- the image correction method of the present invention is applied to the correction of a face image, in particular, the correction of an eye region in a face image.
- the "eye region” in the present invention refers to the visible portion in the eyelid. Divided (including pupil, iris, sclera, etc.), each person has two eyes, therefore, the actual face image has two eye regions, and the target sample image also has two eye regions.
- the left eye region on the actual face image corresponds to the left eye region on the target sample image
- the area of the eye on the right side corresponds to the orientation of the pupil area, and when the image is captured by the camera, the orientation of the eye area in the target sample image is directly opposite the camera.
- the purpose of the "acquiring a target sample image that matches the actual face image” is to find a target sample image that is identical to the actual face image at both the shooting angle and the shooting distance.
- the target sample image and the actual face image are matched, that is, the shooting angles of the two face images and the shooting distance are respectively corresponding, and the two eye regions in the target sample image are front view cameras. of. Therefore, the image correction method of the present invention corrects the actual face image such that the orientation of the two eye regions of the corrected face image is the same as the orientation of the corresponding two eye regions of the target sample image. That is, both face the camera.
- the actual face image collected by the first party camera is transmitted to the second party, and the second party can use the image correction method to acquire the actual face image of the other party and the eye region thereof.
- the corrected eye area (specifically, the pupil part) faces the camera of the first party, that is, for the second party, the first party user is viewed toward the second party, for the first Fang Yiran can thus improve the visual experience of users who make video calls.
- the present invention corrects the respective eye regions of the actual face image, and obtains the corrected actual face image by combining the information of each eye region of the actual face image with the information of each eye region of the target sample image.
- the center coordinates of the respective pupil portions of the corrected actual face image are corresponding to and the same as the central coordinates of the corresponding pupil portions of the target sample image; the coordinates of the respective pupils of the corrected actual face image and the coordinates of the points on the iris region
- Corresponding pupils of the target sample image correspond to coordinates of respective points on the iris region, and the gray values of the pupils and iris regions of the corrected actual face image are unchanged, thereby visually correcting the corrected
- the pupil part of the actual face image is facing the camera.
- the actual Face images can be acquired in real time.
- a plurality of preset sample images may be stored, and when the actual face image is corrected, a sample image in which the actual face image is most matched is used as a target sample. image.
- the step of acquiring a target sample image that matches the actual face image may include:
- the feature information is information capable of indicating the position of the face and the acquisition angle. Specifically, as shown in FIG. 2 and FIG. 3, the feature information includes central coordinates of each eye region (including the center coordinates of the left eye region and the center coordinates of the right eye region, such as O1 and O2 in FIG. 2).
- the aspect ratio W/H of each eye area (shown in Figure 3), the diameter D of the iris in each eye area, the distance between the center of the left eye area and the nose L_en, and the center of the right eye area and the nose
- the ratio of the distance R_en k_en, the distance between the center of the left eye region and the mouth L_em, and the ratio of the distance between the center of the right eye region and the mouth R_em, k_em, or any combination of any of them preferably, when the feature information of the actual face image is acquired, the foregoing specific feature information is simultaneously acquired, thereby determining an accurate position and an acquisition angle of the face.
- the center coordinates of the eye region are used to determine the position of the eye region of the face in the image captured by the camera.
- the coordinate system for determining the center coordinates of the respective eye regions of the actual face image should correspond to the coordinate system for determining the center coordinates of the respective eye regions of the target sample image.
- the center coordinates of the respective eye regions of the actual face image coincide with the center coordinates of the corresponding eye regions of the target sample image, the relative positions of the respective eye regions of the actual face image at the face and the target sample image
- the corresponding eye areas are in the same position on the face. It should be understood that the center of the eye area here is the elliptical center formed by the eyelids, not the center of the iris or pupil.
- the posture and the photograph of the photographed face can be determined.
- the position of the head like the eye area.
- the distance between the center of the left eye region and the nose L_en and the distance between the center of the right eye region and the nose R_en the distance between the center of the left eye region and the mouth L_em and the right side
- the camera is located at the center of the line between the two eye regions.
- the matching degree of each specific feature information of the two images may be separately calculated, and the matching degree of each specific feature information is calculated.
- both are greater than the threshold, it indicates that the preset sample image is a target sample image that matches the actual face image.
- the center coordinates of the respective eye regions between the actual face image and the sample image, the aspect ratio W/H of each eye region, the diameter D of the iris in each eye region, the center of the left eye region, and the nose The distance between the distance L_en and the distance between the center of the right eye region and the nose R_en k_en, the distance between the center of the left eye region and the mouth L_em and the distance between the center of the right eye region and the mouth R_em a pair of k_em ratio.
- edge detection can be used to detect part of the edge points of the eyelid, and according to the characteristics of the ellipse of the human eye, the hough transform is used to calculate the ellipse corresponding to each eye region to calculate the respective eye regions. Center coordinates.
- the image correction method in the present invention is for correcting an eye region of each of the actual face images such that the orientation of each eye region in the corrected actual face image and the orientation of the corresponding eye region of the target sample image the same.
- the orientation of each eye region is consistent with the orientation of each eye region of the target sample image, so these actual face images may not be corrected.
- a step of acquiring a target sample image that matches the actual face image and correcting each eye region of the actual face image according to the target sample image such that the corrected actual face image is The step of matching the orientation of each eye region with the orientation of the corresponding eye region of the target sample image further includes:
- the distance between the eyes of the person is certain if the center coordinate of the pupil portion of one eye region in the actual face image and the center coordinate of the pupil portion of the corresponding eye region in the target image. The distance is greater than the predetermined value, and then the distance between the center coordinate of the pupil portion of the other eye region in the actual face image and the center coordinate of the pupil portion of the corresponding eye region is also greater than the predetermined value.
- the pupil portion in the actual face image is in the eye region
- the position is consistent with the position of the pupil part of the target sample image in the eye area.
- the actual position can be judged.
- the orientation of the eye region in the face image coincides with the orientation of the corresponding eye region in the target sample image, so there is no need to correct the actual face image.
- the predetermined value may be selected between 0 and 0.5 mm.
- the predetermined value may be selected according to the actual face image size, and no limitation is made here.
- the image correction method further includes performing the step of acquiring an actual face image:
- Feature information of each preset sample image is acquired and stored to form the sample library.
- Multiple face images can be images obtained by collecting different faces, and the camera can collect images on faces at different distances and at different angles.
- the correction method after comparing the feature information of the actual face image with the feature information of the preset sample image, selecting a preset sample image that matches the actual face image as the target sample An image, the position and an angle of the corresponding face in the target sample image are consistent with the position and angle of the corresponding face in the actual face image, that is, the feature information of the target sample image and the feature information of the actual face image. The match is greater than the preset value.
- the image is stored in the form of gray value when the image is stored, and the feature information is stored in a numerical manner.
- the gray value and the feature information of the image can be Perform normalization.
- the image correction method further includes performing the step of acquiring and storing each preset sample image:
- the step of respectively calculating the matching degree of the feature information of the actual face image and the feature information of each preset sample image in the sample library may include:
- the feature information may include a plurality of different information: each eye The central coordinate of the region, the aspect ratio of each eye region, the diameter of the iris in each eye region, the distance between the center of the left eye region and the nose, and the distance between the center of the right eye region and the nose, The ratio of the distance between the center of the left eye area to the mouth and the distance between the center of the right eye area and the mouth.
- all the sample images can be first divided into a large class according to the central coordinates of each eye region, and the range of the central coordinates of each eye region of each of the a large classes is different;
- the aspect ratio of the region is further classified into a large class to obtain b subclasses, and the range of the central coordinates of the eye regions of each of the b categories is different from each other, and the range of the aspect ratio of the eye region is different. They are also different from each other; again, b categories are sorted again according to the height of the eyes, c categories are obtained, and so on.
- the preset sample images of the m categories are obtained, and the preset sample images of any two categories have different ranges of at least one specific feature information, for example, at least the range of the aspect ratio of the eye region is different, or at least the eye region The center coordinate range is different.
- a plurality of sample images may be included in each category, when each specific feature information of the actual face image is in the sample image of the nth category.
- the preset sample image with the highest matching degree is used as the Target sample image.
- the predetermined rule described above may include a maximum likelihood classification. For example, when a preset sample image in which the aspect ratio of the eye region is between 1.5 and 2 and the iris diameter is between 1.0 and 1.2 is divided into the first category, and there is no conditional preset sample image in the sample library.
- the maximum likelihood classification method can assign a predetermined sample image of an eye region having an aspect ratio between 1.45 and 2 and an iris diameter between 1.0 and 1.2 into the class.
- the iris and the pupil portion of the actual face image deviates from the camera is large (for example, when the face is slightly faced), the iris and the pupil portion need to be adjusted when the orientation of the pupil portion of the eye region is adjusted.
- a large angle ⁇ is deflected to achieve the purpose of the pupil direct view camera.
- the human eye has a certain range, that is, the iris and the pupil part can be rotated by a certain angle, but when the angle of the iris and the pupil part is too large, only the iris and the pupil part can be exposed in the eye area. small portion.
- the image The correction method also includes before the step of obtaining the actual face image:
- the predetermined coordinate range being greater than a distance between central coordinates of two eye regions of the target sample image
- the predetermined coordinate range may be determined according to a field of view of the camera and a distance of the face from the camera. It can be understood that the predetermined coordinate range is smaller than a range that the camera can capture. For example, if the image captured by the camera is an image of 30 cm*20 cm, then the predetermined coordinate range can be set to 20 cm*10 cm, that is, the entire image center is taken as the coordinate origin, when the center of the human eye (such as the center of the left eye region) The horizontal coordinate of the coordinates is less than -10cm.
- the actual abscissa of the center of the human eye is set to -10cm, so that at the time of correction, the center of the left eye area and the center of the right eye area move less by 1cm. The correction difficulty is reduced.
- the abscissa of the center of the human eye is greater than 10 cm, the actual abscissa of the center of the human eye is set to 10 cm.
- each eye region of the actual face image is corrected according to the target sample image such that the orientation of each eye region of the corrected actual face image and the corresponding eye of the target sample image
- the steps of the orientation of the regions may include:
- the feature point of the eye region in the actual face image may be a partial edge point of the iris in the eye region, and the target feature point corresponding to the feature point of the eye region in the actual face image in the target sample image is Partial edge of the iris of the target sample image point.
- the position of the feature point in the actual face image on the iris is the same as the position of the target feature point in the target sample image on the iris, and the feature point in the actual face image
- the information of the target feature points in the target sample image may include coordinate information and gray value.
- the coordinates of the feature point in the actual face image and the target feature point in the target sample image are Information and gray values can all be different.
- a is an eye region of an actual face image to be corrected, wherein feature point A is located to the left of the iris
- b in FIG. 4 is an eye region in the target sample image, wherein the target feature point A' Also located on the left side of the iris, the feature point A corresponds to the target feature point A', but the coordinates and gray values of A and A' are not the same.
- a plurality of feature points may be selected in the eye region of the actual face image, and a mapping relationship between the feature point and the corresponding target feature point is calculated according to the actual coordinate value and the gray value of the feature point and the corresponding target feature point. And using the mapping relationship to calculate a coordinate position of a target point corresponding to each pixel point of the eye region in the actual face image to be corrected in the target sample image.
- the mapping relationship is that the position of the target feature point in the target sample image is 2 mm to the right of the position of the feature point in the actual face image to be corrected, and then the image may be corrected according to the mapping relationship.
- the gray value of the pixel at the point is corrected to the gray value of the pixel at 2 mm to the left of the pixel, and the gray value of the pixel is shifted to the right by 2 mm, so that the visual effect is the iris and the pupil portion.
- Moved 2 mm to the right so that the pupil portion of the corrected face image and the pupil portion of the target sample image are in the same position, so that the pupil portion of the corrected face image and the pupil portion of the target sample image are aligned. Both are facing the camera.
- an image correction apparatus including:
- the actual image acquisition module 10 is configured to acquire an actual face image
- a target sample image obtaining module 20 configured to acquire a target sample image that matches the actual face image
- the processing module 40 is configured to correct each eye region of the actual face image according to the target sample image, so that each eye of the corrected actual face image is The orientation of the region coincides with the orientation of the corresponding eye region of the target sample image.
- the image correcting device may pre-store a plurality of preset sample images, and the target sample image acquiring module 20 may compare the plurality of preset sample images with the actual face image to select the actual face image. Match the target sample image. Whether the preset sample image and the actual face image match can be judged by comparing the feature information of the two images.
- the target sample image obtaining module 20 includes:
- a feature information acquiring unit 21 configured to acquire feature information of the actual face image
- the matching degree calculation unit 22 is configured to separately calculate a matching degree between the feature information of the actual face image and the feature information of each preset sample image in the sample library, when a preset sample image and the actual person When the matching degree between the face images is greater than the preset value, the one preset sample image that matches the actual face image with the matching degree greater than the preset value is set as the target sample image.
- the feature information includes the center coordinates of the respective eye regions (O1 and O2 in the figure), the aspect ratio W/H of the eye region, the diameter D of the iris in each eye region, and the left eye region.
- the distance R_em is a combination of any one or any of k_em.
- the image correction device further includes:
- a coordinate acquiring module 90 configured to acquire a center coordinate of a pupil portion of each eye region of the actual face image and a center coordinate of a pupil portion of each eye region in the target sample image;
- the determining module 30 is configured to calculate a distance between a center coordinate of the pupil portion of the eye region in the actual face image and a center coordinate of a pupil portion of the corresponding eye region in the target sample image, when the distance is greater than
- the processing module 40 may correct each eye region of the actual face image according to the target sample image, so that the orientation of each eye region of the corrected actual face image and the target sample are corrected.
- the orientation of the corresponding eye area of the image matches.
- the judging module 30 judges When the distance between the center coordinates of the pupil portions of the respective eye regions in the actual face image and the center coordinates of the pupil portions of the corresponding eye regions in the target sample is less than a predetermined value, it is not necessary to correct the actual face image.
- the feature information acquiring unit 21 may also acquire the feature information of the preset sample image, and the image correcting device further includes:
- the sample image acquisition module 50 is configured to collect a plurality of face images as preset sample images, and in each preset sample image, the orientation of each eye region of the collected person faces the sample image acquisition module 50;
- the storage module 60 is configured to store feature information of each preset sample image to form the sample library.
- the process of acquiring the preset sample image by using the sample image acquisition module 50 may be performed before the actual face image is acquired. Before the actual image acquisition module 10 acquires the actual face image, a plurality of preset sample images have been stored in the storage module 60.
- the image correction device may further include:
- a classification module 70 configured to divide the plurality of preset sample images into a plurality of categories according to a predetermined rule
- the target sample image obtaining module 20 further includes: a categorizing unit 23, configured to determine a category of the preset sample image corresponding to the actual face image; and the categorizing unit 23 determines that the actual face image corresponds to the nth
- the matching degree calculation unit 22 can calculate the actual face separately The degree of matching between the feature information of the image and the feature information of each preset sample image in the category of the preset sample image corresponding thereto.
- the predetermined rule includes a maximum likelihood classification.
- the image correction apparatus may further include:
- a pre-positioning module 80 configured to set an coordinate axis within a field of view of the camera and set a predetermined coordinate range within the field of view, the predetermined coordinate range being greater than a center coordinate between two eye regions of the target sample image Distance;
- a boundary value of the predetermined coordinate range is set as an actual coordinate value of a center of an eye region exceeding the predetermined coordinate range in the actual face image.
- the first-party camera may acquire an entire image including an actual face image and a background image
- the actual image acquiring module in the second-party image correcting device may receive the first party The overall image captured by the camera.
- the predetermined coordinate range is set on the entire image.
- the pre-positioning module 80 can determine the actual coordinates of the respective eye regions according to the predetermined coordinate range, and when the processing module 40 adjusts the orientation of the pupils of the respective eye regions of the face image, the respective positions can be directly retrieved from the pre-position module 80.
- the actual coordinate value of the center of the eye area can be determined.
- processing module 40 can include:
- a feature point acquiring unit 41 configured to acquire a feature point of each eye region in the actual face image and a target feature point in the corresponding target sample image that matches a feature point of each eye region in the actual face image information
- mapping calculation unit 42 configured to calculate a mapping relationship between feature points of respective eye regions in the actual face image and corresponding target feature points
- the correcting unit 43 is configured to calculate a corrected position of each feature point of each eye region in the actual face image according to the mapping relationship.
- a video system including the above image correcting apparatus.
- the first party's camera collects the actual face image of each frame of the first party user, and transmits the image to the second party's image correction device by using the network, and the second party's image correction is performed.
- the actual image acquisition module in the device can acquire the actual face image collected by the camera of the first party, and determine and correct the actual face image, so that the orientation of each eye region of the corrected face image is correct to the camera. Therefore, for the second party user, the other party and the user are always looking at themselves during the communication process, thereby achieving the effect of peer-to-peer communication and improving the sensory experience.
- the above is a description of the image correction method, the image correction device, and the video system provided by the present invention.
- the image correction device may store a plurality of preset sample images in advance, and the orientations of the respective eye regions of the preset sample image are all facing the camera. Therefore, when the face image is corrected by the image correction method described above, the image can be acquired and matched with the actual face image.
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Oral & Maxillofacial Surgery (AREA)
- General Health & Medical Sciences (AREA)
- Signal Processing (AREA)
- Human Computer Interaction (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Quality & Reliability (AREA)
- Geometry (AREA)
- Data Mining & Analysis (AREA)
- Ophthalmology & Optometry (AREA)
- Evolutionary Computation (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Probability & Statistics with Applications (AREA)
- General Engineering & Computer Science (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Image Processing (AREA)
- Image Analysis (AREA)
Abstract
Description
Claims (19)
- 一种图像修正方法,包括:获取实际人脸图像;获取与所述实际人脸图像相匹配的目标样本图像;以及根据所述目标样本图像对所述实际人脸图像的各个眼睛区域进行修正,以使得修正后的实际人脸图像的各个眼睛区域的朝向和所述目标样本图像的相应的眼睛区域的朝向相符。
- 根据权利要求1所述的图像修正方法,其中,获取与所述实际人脸图像相匹配的目标样本图像的步骤包括:获取所述实际人脸图像的特征信息;以及分别计算所述实际人脸图像的特征信息与样本库中每张预设样本图像的特征信息的匹配度,当一张预设样本图像与所述实际人脸图像之间的匹配度大于预设值时,将该与所述实际人脸图像之间的匹配度大于预设值的一张预设样本图像设定为所述目标样本图像。
- 根据权利要求2所述的图像修正方法,其中,所述特征信息包括各个眼睛区域的中心坐标、各个眼睛区域的宽高比、各个眼睛区域中的虹膜的直径、左侧眼睛区域的中心与鼻子之间的距离和右侧眼睛区域的中心与鼻子之间的距离之比、左侧眼睛区域的中心与嘴巴之间的距离和右侧眼睛区域的中心与嘴巴之间的距离之比中的任意一者或任意几者的组合。
- 根据权利要求2所述的图像修正方法,其中,获取与所述实际人脸图像相匹配的目标样本图像的步骤和根据所述目标样本图像对所述实际人脸图像的各个眼睛区域进行修正,以使得修正后的实际人脸图像的各个眼睛区域的朝向和所述目标样本图像的相应的眼睛区域的朝向相符的步骤之间还包括:获取所述实际人脸图像中各个眼睛区域的瞳孔部分的中心坐标 和所述目标样本图像中各个眼睛区域的瞳孔部分的中心坐标;以及计算所述实际人脸图像中各个眼睛区域的瞳孔部分的中心坐标与所述目标样本中相应的眼睛区域的瞳孔部分的中心坐标之间的距离,当所述距离大于预定值时,执行根据所述目标样本图像对所述实际人脸图像的各个眼睛区域进行修正,以使得修正后的实际人脸图像的各个眼睛区域的朝向和所述目标样本图像的相应的眼睛区域的朝向相符的步骤。
- 根据权利要求1至4中任意一项所述的图像修正方法,其中,所述图像修正方法还包括在获取实际人脸图像的步骤之前进行的:采集多张人脸图像作为预设样本图像,在每张预设样本图像中,被采集者的各个眼睛区域的朝向均正对摄像头;以及获取并存储每张预设样本图像的特征信息,以形成所述样本库。
- 根据权利要求5所述的图像修正方法,其中,所述图像修正方法还包括在获取并存储每张预设样本图像的特征信息的步骤之后进行的:按照预定规则将所述预设样本图像划分为多个类别;并且分别计算所述实际人脸图像的特征信息与样本库中每张预设样本图像的特征信息的匹配度的步骤包括:确定所述实际人脸图像所对应的预设样本图像的类别;以及计算所述实际人脸图像的特征信息和与其对应的预设样本图像的类别中每张预设样本图像的特征信息的匹配度。
- 根据权利要求6所述的图像修正方法,其中,所述预定规则包括最大似然分类法。
- 根据权利要求1至4中任意一项所述的图像修正方法,其中,所述图像修正方法还包括在获取实际人脸图像的步骤之前进行的:在所述摄像头的视野范围内设立坐标轴;在所述视野范围内设置预定坐标范围,所述预定坐标范围大于所述目标样本图像的两个眼睛区域的中心坐标之间的距离;以及当所述实际人脸图像中的任意一个眼睛区域的中心坐标值超出所述预定坐标范围时,将所述预定坐标范围的边界值设置为所述实际人脸图像中超出所述预定坐标范围的眼睛区域的中心的实际坐标值。
- 根据权利要求1至4中任意一项所述的图像修正方法,其中,根据所述目标样本图像对所述实际人脸图像的各个眼睛区域进行修正,以使得修正后的实际人脸图像的各个眼睛区域的朝向和所述目标样本图像的相应的眼睛区域的朝向相符的步骤包括:获取所述实际人脸图像中各个眼睛区域的特征点的信息和相对应的目标样本图像中与所述实际人脸图像中各个眼睛区域的特征点对应的目标特征点的信息;计算所述实际人脸图像中各个眼睛区域的特征点与相对应的目标特征点之间的映射关系;以及根据所述映射关系计算所述实际人脸图像中各个眼睛区域的各特征点修正后的位置。
- 一种图像修正装置,其特征在于,包括;实际图像获取模块,用于获取实际人脸图像;目标样本图像获取模块,用于获取与所述实际人脸图像相匹配的目标样本图像;以及处理模块,用于根据所述目标样本图像对所述实际人脸图像的各个眼睛区域进行修正,以使得修正后的实际人脸图像的各个眼睛区域的朝向和所述目标样本图像的相应的眼睛区域的朝向相符。
- 根据权利要求10所述的图像修正装置,其中,所述目标样本图像获取模块包括:特征信息获取单元,用于获取所述实际人脸图像的特征信息;以及匹配度计算单元,用于分别计算所述实际人脸图像的特征信息与样本库中每张预设样本图像的特征信息之间的匹配度,当一张预设样本图像与所述实际人脸图像之间的匹配度大于预设值时,将该与所述实际人脸图像之间的匹配度大于预设值的一张预设样本图像设定为所述目标样本图像。
- 根据权利要求11所述的图像修正装置,其中,所述特征信息包括各个眼睛区域的中心坐标、各个眼睛区域的宽高比、各个眼睛区域中的虹膜的直径、左侧眼睛区域的中心与鼻子之间的距离和右侧眼睛区域的中心与鼻子之间的距离之比、左侧眼睛区域的中心与嘴巴之间的距离和右侧眼睛区域的中心与嘴巴之间的距离之比中的任意一者或任意几者的组合。
- 根据权利要求10至12中任意一项所述的图像修正装置,其中,所述图像修正装置还包括:坐标获取模块,用于获取所述实际人脸图像的各个眼睛区域的瞳孔部分的中心坐标和所述目标样本图像中的各个眼睛区域的瞳孔部分的中心坐标;判断模块,用于计算所述实际人脸图像中各个眼睛区域的瞳孔部分的中心坐标与所述目标样本图像中相应的眼睛区域的瞳孔部分的中心坐标之间的距离;当所述距离大于预定值时,所述处理模块能够根据所述目标样本图像对所述实际人脸图像的眼睛区域进行修正,以使得修正后的实际人脸图像的各个眼睛区域的朝向和所述目标样本图像的相应的眼睛区域的朝向相符。
- 根据权利要求10至12中任意一项所述的图像修正装置,其中,所述特征信息获取单元还能够获取预设样本图像的特征信息;并且所述图像修正装置还包括:样本图像采集模块,用于采集多张人脸图像作为预设样本图像, 在每张预设样本图像中,被采集者的各个眼睛区域的朝向正对所述样本图像采集模块;以及存储模块,用于存储每张预设样本图像的特征信息,以形成所述样本库。
- 根据权利要求14所述的图像修正装置,其中,所述图像修正装置还包括:分类模块,用于按照预定规则将所述预设样本图像划分为多个类别;并且所述目标样本图像获取模块还包括:归类单元,用于确定所述实际人脸图像所对应的预设样本图像的类别;并且所述匹配度计算单元能够分别计算所述实际人脸图像的特征信息与其对应的预设样本图像的类别中每张预设样本图像的特征信息之间的匹配度。
- 根据权利要求15所述的图像修正装置,其中,所述预定规则包括最大似然分类法。
- 根据权利要求10至12中任意一项所述的图像修正装置,其中,所述图像修正装置还包括:预定位模块,用于在摄像头的视野范围内设立坐标轴并在所述视野范围内设置预定坐标范围,所述预定坐标范围大于所述目标样本图像的两个眼睛区域的中心坐标之间的距离;并且当所述实际人脸图像中的任意一个眼睛区域的中心坐标值超出所述预定坐标范围时,将所述预定坐标范围的边界值设置为所述实际人脸图像中超出所述预定坐标范围的眼睛区域的中心的实际坐标值。
- 根据权利要求10至12中任意一项所述的图像修正装置,其中,所述处理模块包括:特征点获取单元,用于获取所述实际人脸图像中各个眼睛区域的特征点和相对应的目标样本图像中与所述实际人脸图像中各个眼睛区域的特征点匹配的目标特征点的信息;映射计算单元,用于计算所述实际人脸图像中各个眼睛区域的特征点与相对应的目标特征点之间的映射关系;以及修正单元,用于根据所述映射关系计算所述实际人脸图像中各个眼睛区域的各特征点修正后的位置。
- 一种视频系统,包括权利要求10至18中任意一项所述的图像修正装置。
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US14/891,943 US9824428B2 (en) | 2015-01-29 | 2015-05-18 | Image correction method, image correction apparatus and video system |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510045283.6 | 2015-01-29 | ||
CN201510045283.6A CN104574321B (zh) | 2015-01-29 | 2015-01-29 | 图像修正方法、图像修正装置和视频系统 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2016119339A1 true WO2016119339A1 (zh) | 2016-08-04 |
Family
ID=53090302
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/CN2015/079165 WO2016119339A1 (zh) | 2015-01-29 | 2015-05-18 | 图像修正方法、图像修正装置和视频系统 |
Country Status (3)
Country | Link |
---|---|
US (1) | US9824428B2 (zh) |
CN (1) | CN104574321B (zh) |
WO (1) | WO2016119339A1 (zh) |
Families Citing this family (23)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
KR20150015680A (ko) * | 2013-08-01 | 2015-02-11 | 씨제이씨지브이 주식회사 | 특징점의 생성을 이용한 이미지 보정 방법 및 장치 |
CN104574321B (zh) * | 2015-01-29 | 2018-10-23 | 京东方科技集团股份有限公司 | 图像修正方法、图像修正装置和视频系统 |
RU2596062C1 (ru) | 2015-03-20 | 2016-08-27 | Автономная Некоммерческая Образовательная Организация Высшего Профессионального Образования "Сколковский Институт Науки И Технологий" | Способ коррекции изображения глаз с использованием машинного обучения и способ машинного обучения |
US10198819B2 (en) | 2015-11-30 | 2019-02-05 | Snap Inc. | Image segmentation and modification of a video stream |
ES2912310T3 (es) | 2016-01-05 | 2022-05-25 | Reald Spark Llc | Corrección de la mirada en imágenes multivista |
CN106161933B (zh) * | 2016-06-30 | 2019-05-17 | 维沃移动通信有限公司 | 一种图像处理方法及移动终端 |
CN107368775A (zh) * | 2017-04-21 | 2017-11-21 | 阿里巴巴集团控股有限公司 | 一种虹膜识别过程中的预览方法及装置 |
EP3665553B1 (en) | 2017-08-08 | 2023-12-13 | RealD Spark, LLC | Adjusting a digital representation of a head region |
CN110020578A (zh) | 2018-01-10 | 2019-07-16 | 广东欧珀移动通信有限公司 | 图像处理方法、装置、存储介质及电子设备 |
US11017575B2 (en) | 2018-02-26 | 2021-05-25 | Reald Spark, Llc | Method and system for generating data to provide an animated visual representation |
CN108416747B (zh) * | 2018-02-27 | 2020-07-10 | 平安科技(深圳)有限公司 | 元素位置修正方法、装置、计算机设备及存储介质 |
CN108491775B (zh) * | 2018-03-12 | 2021-04-23 | 维沃移动通信有限公司 | 一种图像修正方法及移动终端 |
CN108985245A (zh) * | 2018-07-25 | 2018-12-11 | 深圳市飞瑞斯科技有限公司 | 眼部位置的确定方法、装置、计算机设备和存储介质 |
CN108965767A (zh) * | 2018-07-26 | 2018-12-07 | 吴铁 | 一种改善人与人视频交互体验的视频处理方法及系统 |
CN110163806B (zh) * | 2018-08-06 | 2023-09-15 | 腾讯科技(深圳)有限公司 | 一种图像处理方法、装置以及存储介质 |
CN110942422A (zh) * | 2018-09-21 | 2020-03-31 | 北京市商汤科技开发有限公司 | 一种图像处理方法、装置和计算机存储介质 |
CN109587035B (zh) * | 2018-10-24 | 2020-08-07 | 北京三快在线科技有限公司 | 会话界面的头像展示方法、装置、电子设备及存储介质 |
CN110458121B (zh) * | 2019-08-15 | 2023-03-14 | 京东方科技集团股份有限公司 | 一种人脸图像生成的方法及装置 |
CN110544317A (zh) * | 2019-08-29 | 2019-12-06 | 联想(北京)有限公司 | 图像处理方法、装置、电子设备和可读存储介质 |
CN111093027B (zh) * | 2019-12-31 | 2021-04-13 | 联想(北京)有限公司 | 一种显示方法及电子设备 |
JP7451242B2 (ja) * | 2020-03-13 | 2024-03-18 | キヤノン株式会社 | 画像処理装置、画像処理方法、及びプログラム |
CN111401459A (zh) * | 2020-03-24 | 2020-07-10 | 谷元(上海)文化科技有限责任公司 | 一种动画人物形态变化视觉捕捉系统 |
CN111428707B (zh) * | 2020-06-08 | 2020-11-10 | 北京三快在线科技有限公司 | 图形识别码的识别方法、装置、存储介质及电子设备 |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102483854A (zh) * | 2009-09-11 | 2012-05-30 | 皇家飞利浦电子股份有限公司 | 图像处理系统 |
CN103310186A (zh) * | 2012-02-29 | 2013-09-18 | 三星电子株式会社 | 校正图像中用户的注视方向的方法和便携式终端 |
CN103345619A (zh) * | 2013-06-26 | 2013-10-09 | 上海永畅信息科技有限公司 | 视频聊天中的人眼自然对视自适应矫正方法 |
US20140016871A1 (en) * | 2012-07-13 | 2014-01-16 | Postech Academy-Industry | Method for correcting user's gaze direction in image, machine-readable storage medium and communication terminal |
CN104574321A (zh) * | 2015-01-29 | 2015-04-29 | 京东方科技集团股份有限公司 | 图像修正方法、图像修正装置和视频系统 |
Family Cites Families (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
KR100307854B1 (ko) * | 1999-11-10 | 2001-11-02 | 전병우 | 단일카메라를 이용한 시선 맞춤 보정 방법 및 그 장치 |
US6806898B1 (en) * | 2000-03-20 | 2004-10-19 | Microsoft Corp. | System and method for automatically adjusting gaze and head orientation for video conferencing |
CN100452081C (zh) * | 2007-06-01 | 2009-01-14 | 华南理工大学 | 一种人眼定位及人眼状态识别方法 |
JP4966816B2 (ja) * | 2007-10-25 | 2012-07-04 | 株式会社日立製作所 | 視線方向計測方法および視線方向計測装置 |
US8897553B2 (en) * | 2011-12-13 | 2014-11-25 | The Nielsen Company (Us), Llc | Image comparison using color histograms |
JP5949319B2 (ja) * | 2012-08-21 | 2016-07-06 | 富士通株式会社 | 視線検出装置及び視線検出方法 |
KR20140090538A (ko) * | 2013-01-09 | 2014-07-17 | 삼성전자주식회사 | 디스플레이 장치 및 제어 방법 |
CN105474633B (zh) * | 2013-06-17 | 2019-07-09 | 瑞尔D斯帕克有限责任公司 | 控制定向背光的光源 |
CN103426195B (zh) * | 2013-09-09 | 2016-01-27 | 天津常青藤文化传播有限公司 | 生成裸眼观看三维虚拟动画场景的方法 |
JPWO2015136908A1 (ja) * | 2014-03-13 | 2017-04-06 | パナソニックIpマネジメント株式会社 | 視線検出装置 |
KR20150119621A (ko) * | 2014-04-16 | 2015-10-26 | 삼성전자주식회사 | 디스플레이 장치 및 그의 이미지 합성 방법 |
EP3134847A1 (en) * | 2014-04-23 | 2017-03-01 | Google, Inc. | User interface control using gaze tracking |
-
2015
- 2015-01-29 CN CN201510045283.6A patent/CN104574321B/zh active Active
- 2015-05-18 WO PCT/CN2015/079165 patent/WO2016119339A1/zh active Application Filing
- 2015-05-18 US US14/891,943 patent/US9824428B2/en active Active
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102483854A (zh) * | 2009-09-11 | 2012-05-30 | 皇家飞利浦电子股份有限公司 | 图像处理系统 |
CN103310186A (zh) * | 2012-02-29 | 2013-09-18 | 三星电子株式会社 | 校正图像中用户的注视方向的方法和便携式终端 |
US20140016871A1 (en) * | 2012-07-13 | 2014-01-16 | Postech Academy-Industry | Method for correcting user's gaze direction in image, machine-readable storage medium and communication terminal |
CN103345619A (zh) * | 2013-06-26 | 2013-10-09 | 上海永畅信息科技有限公司 | 视频聊天中的人眼自然对视自适应矫正方法 |
CN104574321A (zh) * | 2015-01-29 | 2015-04-29 | 京东方科技集团股份有限公司 | 图像修正方法、图像修正装置和视频系统 |
Also Published As
Publication number | Publication date |
---|---|
US20160358318A1 (en) | 2016-12-08 |
CN104574321A (zh) | 2015-04-29 |
US9824428B2 (en) | 2017-11-21 |
CN104574321B (zh) | 2018-10-23 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
WO2016119339A1 (zh) | 图像修正方法、图像修正装置和视频系统 | |
US9355314B2 (en) | Head-mounted display apparatus and login method thereof | |
WO2021036436A1 (zh) | 一种人脸识别方法及装置 | |
CN106843821A (zh) | 自动调整屏幕的方法和装置 | |
WO2017161867A1 (zh) | 一种调节屏幕亮度的方法、装置及智能终端 | |
US8538044B2 (en) | Line-of-sight direction determination device and line-of-sight direction determination method | |
US8957943B2 (en) | Gaze direction adjustment for video calls and meetings | |
CN109670390B (zh) | 活体面部识别方法与系统 | |
US8428313B2 (en) | Object image correction apparatus and method for object identification | |
CN106981078B (zh) | 视线校正方法、装置、智能会议终端及存储介质 | |
JP2006510109A (ja) | 表情不変顔認識方法及び装置 | |
US20110142297A1 (en) | Camera Angle Compensation in Iris Identification | |
KR101202448B1 (ko) | 홍채 인식 장치 및 홍채 인식 방법 | |
CN110612530A (zh) | 用于选择脸部处理中使用的帧的方法 | |
US20090074234A1 (en) | System and method for capturing images | |
CN104317385A (zh) | 一种手势识别方法和系统 | |
CN109034108A (zh) | 一种视线估计的方法、装置和系统 | |
JP6396357B2 (ja) | 顔画像認証装置 | |
WO2019154012A1 (zh) | 匹配光源与光斑的方法和装置 | |
CN109840453B (zh) | 一种人脸匹配方法及装置 | |
US8165353B2 (en) | Computer system and method for calculating straightness of facial | |
Waizenegger et al. | Model based 3D gaze estimation for provision of virtual eye contact | |
CN111582003A (zh) | 一种视线跟踪学生教室预防近视系统 | |
CN113191197B (zh) | 一种图像还原方法及装置 | |
KR102194511B1 (ko) | 대표 영상프레임 결정시스템 및 이를 이용한 방법 |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
WWE | Wipo information: entry into national phase |
Ref document number: 14891943 Country of ref document: US |
|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 15879545 Country of ref document: EP Kind code of ref document: A1 |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
122 | Ep: pct application non-entry in european phase |
Ref document number: 15879545 Country of ref document: EP Kind code of ref document: A1 |
|
32PN | Ep: public notification in the ep bulletin as address of the adressee cannot be established |
Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205A DATED 01/02/2018) |
|
122 | Ep: pct application non-entry in european phase |
Ref document number: 15879545 Country of ref document: EP Kind code of ref document: A1 |