(19)
(11)EP 3 647 992 A1

(12)EUROPEAN PATENT APPLICATION
published in accordance with Art. 153(4) EPC

(43)Date of publication:
06.05.2020 Bulletin 2020/19

(21)Application number: 17916284.7

(22)Date of filing:  30.06.2017
(51)Int. Cl.: 
G06K 9/00  (2006.01)
H04N 5/232  (2006.01)
(86)International application number:
PCT/CN2017/091352
(87)International publication number:
WO 2019/000462 (03.01.2019 Gazette  2019/01)
(84)Designated Contracting States:
AL AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HR HU IE IS IT LI LT LU LV MC MK MT NL NO PL PT RO RS SE SI SK SM TR
Designated Extension States:
BA ME
Designated Validation States:
MA MD

(71)Applicant: GUANGDONG OPPO MOBILE TELECOMMUNICATIONS CORP., LTD.
Wusha, Chang'an Dongguan, Guangdong 523860 (CN)

(72)Inventor:
  • LIANG, Kun
    Dongguan, Guangdong 523860 (CN)

(74)Representative: Mewburn Ellis LLP 
Aurora Building Counterslip
Bristol BS1 6BX
Bristol BS1 6BX (GB)

  


(54)FACE IMAGE PROCESSING METHOD AND APPARATUS, STORAGE MEDIUM, AND ELECTRONIC DEVICE


(57) A method for face image processing is provided. The method includes the follows. A face image is recognized to obtain a recognition result. An image adjustment parameter set is obtained according to the recognition result. A deflection angle of a face in the face image relative to a reference face in a reference face image is determined. A target image adjustment parameter is selected from the image adjustment parameter set according to the deflection angle. The face image is processed according to the target image adjustment parameter. A device for face image processing, a storage medium, and an electronic device are further provided.




Description

TECHNICAL FIELD



[0001] This disclosure relates to the technical field of face image processing, and more particularly to a method and device for face image processing, a storage medium, and an electronic device.

BACKGROUND



[0002] With the development of Internet and mobile communication network, and with the rapid enhancement in processing capacity and storage capacity of electronic devices, massive applications have been rapidly popularized and used. Especially for applications related to face image processing, a face image processing function becomes more and more powerful. At present, many camera-related applications can provide a convenient and fast photo beautification function, which can realize instant optimization of photo effects with a simple operation.

SUMMARY



[0003] According to embodiments, a method and device for face image processing, a storage medium, and an electronic device are provided, which can improve speed of face image processing.

[0004] In a first aspect of the present disclosure, a method for face image processing is provided. A face image is recognized to obtain a recognition result. An image adjustment parameter set is obtained according to the recognition result. A deflection angle of a face in the face image relative to a reference face in a reference face image is determined. A target image adjustment parameter is selected from the image adjustment parameter set according to the deflection angle. The face image is processed according to the target image adjustment parameter.

[0005] In a second aspect of the present disclosure, a device for face image processing is provided. The device includes a recognizing module, an obtaining module, a determining module, a selecting module, and a processing module. The recognizing module is configured to recognize a face image to obtain a recognition result. The obtaining module is configured to obtain an image adjustment parameter set according to the recognition result. The determining module is configured to determine a deflection angle of a face in the face image relative to a reference face in a reference face image. The selecting module is configured to select a target image adjustment parameter from the image adjustment parameter set according to the deflection angle. The processing module is configured to process the face image according to the target image adjustment parameter.

[0006] In a third aspect of the present disclosure, a storage medium is provided. The storage medium is configured to store multiple instructions. The instructions, when executed by a processor, are operable to execute the above method for face image processing.

[0007] In a fourth aspect of the present disclosure, an electronic device is provided. The electronic device includes a memory and a processor. The processor is electrically coupled with the memory. The memory is configured to store instructions and data. The processor is configured to: recognize a face image to obtain a recognition result; obtain an image adjustment parameter set according to the recognition result; determine a deflection angle of a face in the face image relative to a reference face in a reference face image; select a target image adjustment parameter from the image adjustment parameter set according to the deflection angle; and process the face image according to the target image adjustment parameter.

BRIEF DESCRIPTION OF THE DRAWINGS



[0008] In order to describe technical solutions of embodiments more clearly, the following will give a brief description of accompanying drawings used for describing the embodiments. Apparently, accompanying drawings described below are merely some embodiments. Those of ordinary skill in the art can also obtain other accompanying drawings based on the accompanying drawings described below without creative efforts.

FIG. 1 is a schematic flowchart illustrating a method for face image processing according to embodiments.

FIG. 2 is a schematic diagram illustrating face images according to embodiments.

FIG. 3 is a schematic flowchart illustrating a method for face image processing according to other embodiments.

FIG. 4 is a schematic diagram illustrating face images according to other embodiments.

FIG. 5 is a schematic flowchart illustrating a method for face image processing according to other embodiments.

FIG. 6 is a schematic diagram illustrating face images according to other embodiments.

FIG. 7 is a schematic comparison diagram illustrating face images before and after face image processing according to embodiments.

FIG. 8 is a schematic structural diagram illustrating a device for face image processing according to a first embodiment.

FIG. 9 is a schematic structural diagram illustrating a device for face image processing according to a second embodiment.

FIG. 10 is a schematic structural diagram illustrating a device for face image processing according to a third embodiment.

FIG. 11 is a schematic structural diagram illustrating a device for face image processing according to a fourth embodiment.

FIG. 12 is a schematic structural diagram illustrating a device for face image processing according to a fifth embodiment.

FIG. 13 is a schematic structural diagram illustrating an electronic device according to embodiments.

FIG. 14 is a schematic structural diagram illustrating an electronic device according to other embodiments.


DETAILED DESCRIPTION



[0009] Hereinafter, technical solutions of embodiments will be described in a clear and comprehensive manner with reference to the accompanying drawings intended for the embodiments. It is evident that the embodiments described herein constitute merely some rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments without creative efforts shall fall within protection scope of the present disclosure.

[0010] The terms "first", "second", "third", and the like (if present) used in the specification, claims, and drawings of the present disclosure are intended for distinguishing different objects rather than describing a particular order or a sequential order. It should be understood that the objects so described are interchangeable where appropriate. In addition, the terms "include", "comprise", "have" and any variations thereof are intended to cover non-exclusive inclusion. Such as a process and method that includes a series of steps, or an apparatus, electronic device, and system that includes a series of modules or units, is not limited to the listed steps or units, but optionally may also include steps or units not listed, or alternatively other steps or units inherent to these processes, methods, apparatuses, electronic device, or system.

[0011] According to embodiments, a method for face image processing is provided. As illustrated in FIG. 1, the method for face image processing includes the following.

[0012] At S110, a face image is recognized to obtain a recognition result.

[0013] In embodiments, the face image can be obtained in a variety of manners. As an example, the face image is collected by an electronic device through a camera of the electronic device in real time or is pre-stored in the electronic device. As another example, the face image is obtained from a network. Specifically, the face image is obtained from a message sent by a friend through an Instant Messenger App (e.g., WeChat®). Face image recognition can be comprehended as recognizing a face in the face image.

[0014] Face recognition is a biometric recognition technology based on human face feature information to identity a person. Generally, a face recognition system mainly relates to four aspects, namely: face-image acquisition and detection, face-image preprocessing, face-image feature extraction, and matching and recognition.

[0015] In some embodiments, an original image cannot be used directly because the original image is often limited by various conditions and subjected to random interference. For this reason, at an early stage of image processing, it is necessary to pre-process (e.g., gradation correction, noise filtering, etc.) the original image after the face image is obtained. Then features of the face image are extracted by representation methods based on knowledge, and algebra features, or statistical learning. For example, information such as histogram features, color features, template features, structural features, and so on can be extracted from the face image, useful information is then selected from such information extracted to implement face detection.

[0016] For example, after features are extracted, the face image is recognized according to the extracted features as follows. Feature templates stored in a database are searched according to extracted feature data of the face image. When a similarity between the extracted feature data and a feature template exceeds a setting threshold, a search result (that is, the feature template) is outputted and face features to-be-recognized are compared with the face feature template obtained. The face in the face image is identified according to similarity between the face featured extracted and the face feature template to obtain a recognition result.

[0017] In some embodiments, face image information may be analyzed by means of a specified classification model. The classification model may be a trained deep neural network based on machine deep learning, such as a convolutional neural network (CNN). The CNN is a multi-layer neural network comprising an input layer, a convolution layer, a pooling layer, a fully connected layer, and an output layer. CNN allows direct input of an image of multi-dimensional input vectors, which avoids data reconstruction in feature extraction and classification, and thus reduces complexity of image processing. Upon input of target image information into the CNN, the target image information is transmitted from the input layer to the output layer through a multi-stage transformation. Calculation performed in the CNN generally relates to a dot product between an input (i.e., the face image) and a weight matrix of each layer to obtain a final output (e.g., gender, age, personal-identity, etc.).

[0018] A CNN model is obtained by training according to samples and classification information in advance. For example, a large number of sample face images can be collected in advance, and each sample face image can be manually marked with information (e.g., gender, age, identity, etc.) of a person to which the face image belongs. These sample face images are then input into the CNN for training. The training mainly includes two phases, that is, a forward propagation phase and a backward propagation phase. In the forward propagation phase, each sample Xi (i.e., a sample image) can be input into an n-layer convolutional neural network to obtain an actual output Oi, where Oi =Fn(...(F2(F1(XiW(1))W(2))...)W(n)), i is a positive integer, W(n) is a weight of an nth layer, and F is an activation function (e.g., a sigmoid function or a hyperbolic tangent function). A weight matrix can be obtained by inputting sample images into the convolutional neural network. Thereafter, in the backward propagation phase, a difference between each actual output Oi and an ideal output Yi can be calculated, and the weight matrix is adjusted by backward propagation according to an error minimization method, where Yi is obtained according to the marked information of sample Xi. For instance, if sample image Xi is marked that the gender is female, then Yi is set to 1; if sample image Xi is marked that the gender is male, then Yi is set to 0. Finally, trained convolutional neural network is determined according to the weight matrix adjusted, and images can be intelligently analyzed through the trained convolutional neural network. In this way, information such as the gender, age, personal-identity, and the like of a person in the image can be accurately determined as a recognition result.

[0019] At SI20, an image adjustment parameter set is obtained according to the recognition result.

[0020] In embodiments, the image adjustment parameter set includes multiple groups of image adjustment parameters (i.e., multiple image adjustment parameter groups). The image adjustment parameters are historical image adjustment parameters of the face in the face image, that is, previously-used image adjustment parameters. In some embodiments, the electronic device can record processing habits for different face images in a time period in the past, and can analyze and learn the processing habits of the user in the time period based on machine learning. The electronic device can generate image adjustment parameters for processing different face images through self-analysis and self-learning processes, to obtain different image adjustment parameter sets.

[0021] In some embodiments, an image adjustment parameter includes a luminance value, a exposure value, contrast, sharpness, a liquefaction value, a pixel blur radius, and other parameter values, which are not enumerated herein.

[0022] In practice, influence of external factors or issues with a face image per se may cause recognition failure. For example, when face comparison is performed, comparison failure may occur if there is a discrepancy between a face in a face image stored in advance and the face in a face image to be recognized, the discrepancy includes: changes in beard, changes in hair style, changes in accessories such as glasses, changes in expression, and so on. When the recognition of the face image fails, the user can manually determine the recognition result. For instance, when receiving a result selection instruction triggered by a user through a display screen, the electronic device obtains and displays a result list according to the result selection instruction, and uses a target result selected from the result list as the recognition result. Then, the electronic device can obtain a image adjustment parameter set corresponding to the recognition result.

[0023] In some embodiments, the recognition result includes a personal-identity. The personal-identity is used to distinguish the person to whom the face in the face image stored in the electronic device belongs. The personal-identity includes, but is not limited to, an identity number, name, nickname, a serial number, or other information that is able to group face images of a same person into one category.

[0024] In some embodiments, the image adjustment parameter set is obtained according to the recognition result as follows. A preset personal-identity that matches the personal-identity is obtained from a preset identity set. A preset image adjustment parameter set corresponding to the preset personal-identity is obtained, and the preset image adjustment parameter set is used as the recognition result to obtain the image adjustment parameter set.

[0025] In above embodiments, a correspondence relationship between personal-identities and image adjustment parameter sets is set in advance. In other words, image adjustment parameters corresponding to face images belonging to a same person will be grouped into a same image adjustment parameter set. As an example, when personal identity of a face image is recognized, once the personal-identity is recognized, the preset image adjustment parameter set is obtained from multiple image adjustment parameter sets according to the personal-identity.

[0026] At S130, a deflection angle of a face in the face image relative to a reference face in a reference face image is determined.

[0027] The reference face image is used as a reference and may be any image that includes a face belonging to a same person as the face in the face image to be recognized.

[0028] For example, as illustrated in FIG. 2, image A is a front-view image of a face in a face image. Image A is used as a reference face image, then image B is a left-deflection image of the face in image A, and a deflection angle of the face in image B relative to the face in image A is horizontally to the left by 45°. Likewise, image C is a right-deflection image of the face in image A, and a deflection angle of the face in image C relative to the face in image A is horizontally to the right by 45°. For another example, Image B is used as a reference face image, a deflection angle of the face in image A relative to the face in image B is horizontally to the right by 45°, and a deflection angle of the face in image C relative to the face in image B is horizontally to the right by 90°. Any of image A, image B, and image C can be used as a reference face image.

[0029] At S140, a target image adjustment parameter is selected from the image adjustment parameter set according to the deflection angle.

[0030] In practice, images taken at different angles are different, and so image adjustment parameters required for the images taken at different angles are different.

[0031] In embodiments, multiple groups of image adjustment parameters included in the image adjustment parameter set may include image adjustment parameters required for face images of different angles to achieve a preset effect such as a picture effect. For instance, a front-view image of a face and a side-view image of the same face are different, and accordingly, image adjustment parameters required for the front-view face image are different from image adjustment parameters required for the side-view face image. For example, there is a scar in the left face while no scar in the right face, when due to the shooting angle, most of the left face is presented in the face image and the scar in the left face is exposed, the scar appeared in the face image needs to be processed; on the other hand, when most of the right face is presented in the face image and the scar in the left face cannot be seen, no scar-related process is needed for the face image. Therefore, according to the deflection angle of the face in the face image relative to the reference face in the reference face image, an orientation of the face in the face image can be determined, and the target image adjustment parameter to-be-selected can be determined.

[0032] In some embodiments, the target image adjustment parameter is selected from the image adjustment parameter set according to the deflection angle as follows. A target angle interval to which the deflection angle belongs is determined from multiple preset angle intervals. A preset image adjustment parameter corresponding to the target angle interval is obtained from the image adjustment parameter set. The preset image adjustment parameter is used as the target image adjustment parameter.

[0033] The multiple preset angle intervals may be obtained by the electronic device by deep-learning user behavior habits for a period of time and analyzing the same.

[0034] At SI50, the face image is processed according to the target image adjustment parameter.

[0035] In embodiments, the manner in which the face image is processed according to the target image adjustment parameter may be various. As an example, an area to-be-adjusted in the face image is determined according to the target image adjustment parameter, and the area determined is adjusted according to the target image adjustment parameter obtained.

[0036] In some embodiments, image features of the face image is adjusted according to the target image adjustment parameter to process the face image. In this case, the face image is processed according to the target image adjustment parameter as follows. Multiple image features of the face image are obtained. For each of multiple image features, an initial layer corresponding to thereto is generated, to obtain a layer set. According to the target image adjustment parameter, at least one target layer is selected from the layer set and the at least one target layer is adjusted. The at least one target layer adjusted is combined with unselected initial layers in the layer set.

[0037] In some embodiments, the multiple image features may include color features, shape features, texture features, and spatial relationship features. In some embodiments, the face image is analyzed, and an initial layer corresponding to each image feature is generated according to a analysis result. For example, a color feature corresponds to a color layer that characterizes colors of the face image, and a shape feature corresponds to a shape layer that characterizes shape of the face image.

[0038] Since not all image features need to be modified, there may be parameters having a null value in the target image adjustment parameter (group), therefore, only parameters whose values are not null in the target image adjustment parameter are used. Accordingly, at least one target layer corresponding to at least one feature to-be-modified is selected for adjustment from a layer set according to at least one parameter whose value is not null, to obtain at least one target layer adjusted. The at least one target layer adjusted is then combined with remaining unadjusted initial layers to synthesize a processed face image.

[0039] As illustrated in FIG. 7, compared to an unprocessed face image, in a processed face image in FIG. 7, shape features, texture features, and color features are processed. Specifically, the luminance of the face image is enhanced, the sharpness of the face image is improved, the contrast of the face image is enhanced, localized skinning is applied, hairstyle is modified, and the like.

[0040] In some embodiments, as illustrated in FIG. 3, the deflection angle of the face in the face image relative to the reference face in the reference face image is determined as follows. At S131, the reference face image corresponding to the face image is obtained according to the recognition result. At S132, a first offset reference point in the reference face of the reference face image is determined. At S133, a second offset reference point corresponding to the first offset reference point is determined in the face of the face image. At S134, position offset information of the second offset reference point relative to the first offset reference point is obtained. At S135, the deflection angle of the face in the face image relative to the reference face in the reference face image is determined according to the position offset information.

[0041] In some embodiments, the position offset information includes an offset distance and an offset direction. The following describes a front-view face image as an example of the reference face image. As illustrated in FIG. 4, using image A as the reference face image, a first offset reference point X1 (corresponding to a pupil) is determined in image A. Using image B as the currently obtained face image, a second offset reference point X2 (corresponding to the pupil) is determined in image B according to the first offset reference point X1. As an example, a position of the second offset reference point X2 is decomposed through a graphic method, and an equivalent point X2' corresponding to the second offset reference point X2 is determined in image A. As illustrated in FIG. 4, there is no distance difference in a vertical direction between the equivalent point X2' and the first offset reference point X1, and the equivalent point X2' offsets horizontally to the left (referred to as "offset direction") by distance d (referred to as "offset distance") with respect to the first offset reference point X1. Then, the offset distance and the offset direction obtained are taken into a corresponding deflection angle-related algorithm obtained, to obtain the deflection angle.

[0042] In some embodiments, as illustrated in FIG. 5, the angle of deflection of the face in the face image relative to the reference face in the reference face image is determined as follows. At S131, the reference face image corresponding to the face image is obtained according to the recognition result. At S136, a first reference vector in the face of the face image is determined. At S137, a second reference vector corresponding to the first reference vector is determined in the reference face of the reference face image. At S138, an angle between the first reference vector and the second reference vector is determined. At S139, the deflection angle of the face in the face image relative to the reference face in the reference face image is determined according to the angle.

[0043] In embodiments, the manner in which the first reference vector is determined may be various. As an example, feature points of the reference face image are obtained to obtain a feature-point set, and then any two representative target feature points are selected from the feature-point set to obtain the first reference vector. As an example, the first reference vector in the face of the face image is determined as follows. A first feature point and a second feature point in the face image are determined. The first reference vector is determined according to the first feature point and the second feature point.

[0044] The following describes a front-view face image as an example of the reference face image. As illustrated in FIG. 6, using image A as the reference face image. A first feature point Y1 (corresponding to a pupil) and a second feature point Z1 (corresponding to a nose) are determined in image A, to obtain a first reference vector a in a direction of Y1 to Z1. Using image B as the currently obtained face image. A third feature point Y2 (corresponding to the pupil) in image B is determined according to the first feature point Y1 and a second feature point Z2 (corresponding to the nose) in image B is determined according to the second feature point Z1, to obtain a second reference vector b in a direction of Y2 to Z2.

[0045] In embodiments, the manner in which the angle between the first reference vector and the second reference vector is determined may be various. In some embodiments, the angle between the first reference vector a and the second reference vector b can be directly calculated using a formula. In some embodiments, as illustrated in FIG. 6, an angleα between the first reference vector a and a horizontal plane and an angleβ between the second reference vector b and the horizontal plane can be calculated respectively, and a difference between the angleα and the angleβ is determined as the angle between the first reference vector and the second reference vector.

[0046] In some embodiments, a deflection angle-related algorithm is invoked, the angle obtained is taken into the deflection angle-related algorithm, and the deflection angle is obtained through operation.

[0047] Likewise, methods of embodiments are further applicable to processing of a face image that produces certain angle deflection in a vertical direction, as well as processing of a face image that simultaneously produces certain angle deflection in a horizontal direction and a vertical direction. For details, reference may be made to the related descriptions in the foregoing processing method, and it will not be described in further detail herein.

[0048] According to the method for face image processing of embodiments, the recognition result can be obtained by recognizing the face image. The image adjustment parameter set is obtained according to the recognition result. The deflection angle of the face in the face image relative to the reference face in the reference face image is determined. The target image adjustment parameter is selected from the image adjustment parameter set according to the deflection angle determined. The face image is processed according to the target image adjustment parameter. By adopting the present disclosure, according to the result of the recognition of the face image and the deflection angle of the face in the face image, the target image adjustment parameter can be determined from the image adjustment parameter set to process the face image, without manually determining how to adjust the face image, which can improve speed and efficiency of image processing and reduce operating time while reducing power consumption of an electronic device.

[0049] According to embodiments, a device for face image processing is further provided. The device can be integrated with a server. As illustrated in FIG. 8, a device 300 for face image processing includes a recognizing module 301, an obtaining module 302, a determining module 303, a selecting module 304, and a processing module 305.

[0050] The recognizing module 301 is configured to recognize a face image to obtain a recognition result.

[0051] In embodiments, the face image can be obtained in a variety of manners. As an example, the face image is collected by an electronic device through a camera of the electronic device in real time or is pre-stored in the electronic device. As another example, the face image is obtained from a network. Specifically, the face image is obtained from a message sent by a friend through an Instant Messenger App (e.g., WeChat®). Face image recognition can be comprehended as recognizing a face in the face image. The recognition result can be used to determine a person to whom respective face image belongs.

[0052] The obtaining module 302 is configured to obtain an image adjustment parameter set according to the recognition result.

[0053] In embodiments, the image adjustment parameter set includes multiple groups of image adjustment parameters. The image adjustment parameters are historical image adjustment parameters of the face in the face image, that is, previously-used image adjustment parameters. In some embodiments, the electronic device can record processing habits for different face images in a time period in the past, and can analyze and learn the processing habits of the user in the time period based on machine learning. The electronic device can generate image adjustment parameters for processing different face images through self-analysis and self-learning processes to obtain different image adjustment parameter sets.

[0054] In some embodiments, the image adjustment parameter is a luminance value, a exposure value, contrast, sharpness, a liquefaction value, a pixel blur radius, or other parameter values, which are not enumerated herein.

[0055] In some embodiments, the recognition result includes a personal-identity. The personal-identity is used to distinguish the person to whom the face in the face image stored in the electronic device belongs. The personal-identity includes, but is not limited to, an identity number, name, nickname, a serial number, or other information that is able to group face images of a same person into one category. In some embodiments, the obtaining module 302 is configured to: obtain, from a preset identity set, a preset personal-identity that matches the personal-identity; and obtain a preset image adjustment parameter set corresponding to the preset personal-identity, and use the preset image adjustment parameter set as the recognition result to obtain the image adjustment parameter set.

[0056] In above embodiments, a correspondence relationship between personal-identities and image adjustment parameter sets is set in advance. In other words, image adjustment parameters corresponding to face images belonging to a same person will be grouped to a same image adjustment parameter set. As an example, when personal identity of a face image is recognized, the preset image adjustment parameter set is obtained from multiple image adjustment parameter sets according to the personal-identity.

[0057] The determining module 303 is configured to determine a deflection angle of a face in the face image relative to a reference face in a reference face image.

[0058] The reference face image is used as a reference and may be any image that includes a face belonging to a same person as the face in the face image to be recognized.

[0059] The selecting module 304 is configured to select a target image adjustment parameter from the image adjustment parameter set according to the deflection angle.

[0060] In embodiments, multiple groups of image adjustment parameters included in the image adjustment parameter set may include image adjustment parameters required for face images of different angles to achieve a preset effect such as a picture effect. For instance, a front-view image of a face and a side-view image of the same face are different. Therefore, according to the deflection angle of the face in the face image relative to the reference face in the reference face image, an orientation of the face in the face image can be determined, and the target image adjustment parameter to-be-selected can be determined.

[0061] In some embodiments, the selecting module 304 is configured to: determine, from multiple preset angle intervals, a target angle interval to which the deflection angle belongs; obtain, from the image adjustment parameter set, a preset image adjustment parameter corresponding to the target angle interval; and use the preset image adjustment parameter as the target image adjustment parameter.

[0062] The processing module 305 is configured to process the face image according to the target image adjustment parameter.

[0063] In embodiments, the manner in which the face image is processed according to the target image adjustment parameter may be various. As an example, an area to-be-adjusted in the face image is determined according to the target image adjustment parameter, and the area determined is adjusted according to the target image adjustment parameter obtained.

[0064] In some embodiments, the recognition result includes a personal-identity. As illustrated in FIG. 9, the obtaining module 302 includes an identity obtaining sub-module 3021 and a set obtaining sub-module 3022. The identity obtaining sub-module 3021 is configured to obtain, from a preset identity set, a preset personal-identity that matches the personal-identity. The set obtaining sub-module 3022 is configured to obtain a preset image adjustment parameter set corresponding to the preset personal-identity, and use the preset image adjustment parameter set as the recognition result to obtain the image adjustment parameter set.

[0065] In some embodiments, as illustrated in FIG. 10, the determining module 303 includes an image obtaining sub-module 3031, a first determining sub-module 3032, an information obtaining sub-module 3033, and a second determining sub-module 3034. The image obtaining sub-module 3031 is configured to obtain the reference face image corresponding to the face image according to the recognition result. The first determining sub-module 3032 is configured to determine a first offset reference point in the reference face of the reference face image, and determine, in the face of the face image, a second offset reference point corresponding to the first offset reference point. The information obtaining sub-module 3033 is configured to obtain position offset information of the second offset reference point relative to the first offset reference point. The second determining sub-module 3034 is configured to determine, according to the position offset information, the deflection angle of the face in the face image relative to the reference face in the reference face image.

[0066] In some embodiments, as illustrated in FIG. 11, the determining module 303 includes the image obtaining sub-module 3031, a third determining sub-module 3035, an angle obtaining sub-module 3036, and a fourth determining sub-module 3037. The image obtaining sub-module 3031 is configured to obtain the reference face image corresponding to the face image according to the recognition result. The third determining sub-module 3035 is configured to determine a first reference vector in the face of the face image, and determine, in the reference face of the reference face image, a second reference vector corresponding to the first reference vector. The angle obtaining sub-module 3036 is configured to determine an angle between the first reference vector and the second reference vector. The fourth determining sub-module 3037 is configured to determine, according to the angle, the deflection angle of the face in the face image relative to the reference face in the reference face image.

[0067] In some embodiments, as illustrated in FIG. 12, the processing module 305 includes a feature obtaining sub-module 3051, a generating sub-module 3052, a selecting sub-module 3053, and a combining sub-module 3054. The feature obtaining sub-module 3051 is configured to obtain multiple image features of the face image. The generating sub-module 3052 is configured to, for each of the multiple image features, generate an initial layer corresponding thereto, to obtain a layer set. The selecting sub-module 3053 is configured to select and adjust, according to the target image adjustment parameter, at least one target layer from the layer set. The combining sub-module 3054 is configured to combine the at least one target layer adjusted with unselected initial layers in the layer set.

[0068] In some embodiments, the multiple image features may include color features, shape features, texture features, and spatial relationship features. In some embodiments, the face image is analyzed, and an initial layer corresponding to each image feature is generated according to a analysis result. For example, a color feature corresponds to a color layer that characterizes colors of the face image, and a shape feature corresponds to a shape layer that characterizes shape of the face image.

[0069] Since not all image features need to be modified, there may be parameters having a null value in the target image adjustment parameter, therefore, only parameters whose values are not null in the target image adjustment parameter are used. Accordingly, at least one target layer corresponding to at least one feature to-be-modified is selected for adjustment from a layer set according to at least one parameter whose value is not null, to obtain at least one target layer adjusted. The at least one target layer adjusted is then combined with remaining unadjusted initial layers to synthesize a processed face image.

[0070] In some embodiments, the recognition result includes a personal-identity. The obtaining module 302 is configured to: obtain, from a preset identity set, a preset personal-identity that matches the personal-identity; and obtain a preset image adjustment parameter set corresponding to the preset personal-identity, and use the preset image adjustment parameter set as the recognition result to obtain the image adjustment parameter set.

[0071] In some embodiments, the selecting module is configured to: determine, from multiple preset angle intervals, a target angle interval to which the deflection angle belongs; obtain, from the image adjustment parameter set, a preset image adjustment parameter corresponding to the target angle interval; and use the preset image adjustment parameter as the target image adjustment parameter.

[0072] According to the device for face image processing of embodiments, the recognition result can be obtained by recognizing the face image. The image adjustment parameter set is obtained according to the recognition result. The deflection angle of the face in the face image relative to the reference face in the reference face image is determined. The target image adjustment parameter is selected from the image adjustment parameter set according to the deflection angle determined. The face image is processed according to the target image adjustment parameter. By adopting the present disclosure, according to the result of the recognition of the face image and the deflection angle of the face in the face image, the target image adjustment parameter can be determined from the image adjustment parameter set to process the face image, without manually determining how to adjust the face image, thereby improving speed and efficiency of image processing and reducing operating time while reducing power consumption of an electronic device.

[0073] According to embodiments, a storage medium is further provided. The storage medium is configured to store multiple instructions. The instructions, when executed by a processor, are operable with the processor to execute any of the foregoing methods for face image processing.

[0074] According to embodiments, an electronic device is further provided. The electronic device may be a device such as a smart phone, a tablet computer, or the like. As illustrated in FIG. 13, an electronic device 400 includes a processor 401 and a memory 402. The processor 401 is electrically coupled with the memory 402.

[0075] The processor 401 is a control center of the electronic device 400. The processor 401 is configured to connect various parts of the entire electronic device 400 through various interfaces and lines, configured to execute various functions of the electronic device and process data by running or loading programs stored in the memory 402 and invoking data stored in the memory 402, thereby monitoring the electronic device 400.

[0076] In this embodiment, the processor 401 of the electronic device 400 is configured to load instructions corresponding to processes of one or more programs into the memory 402 according to following operations, and to run programs stored in the memory 402, thereby implementing various functions. A face image is recognized to obtain a recognition result. An image adjustment parameter set is obtained according to the recognition result. A deflection angle of a face in the face image relative to a reference face in a reference face image is determined. A target image adjustment parameter is selected from the image adjustment parameter set according to the deflection angle. The face image is processed according to the target image adjustment parameter.

[0077] In some embodiments, the processor 401 is configured to: obtain the reference face image corresponding to the face image according to the recognition result; determine a first offset reference point in the reference face of the reference face image; determine, in the face of the face image, a second offset reference point corresponding to the first offset reference point; obtain position offset information of the second offset reference point relative to the first offset reference point; and determine, according to the position offset information, the deflection angle of the face in the face image relative to the reference face in the reference face image.

[0078] In some embodiments, the processor 401 is configured to: obtain the reference face image corresponding to the face image according to the recognition result; determine a first reference vector in the face of the face image; determine, in the reference face of the reference face image, a second reference vector corresponding to the first reference vector; determine an angle between the first reference vector and the second reference vector; and determine, according to the angle, the deflection angle of the face in the face image relative to the reference face in the reference face image.

[0079] In some embodiments, the processor 401 is configured to: determine a first feature point and a second feature point in the face image; determine the first reference vector according to the first feature point and the second feature point.

[0080] In some embodiments, the processor 401 is configured to: obtain multiple image features of the face image; for each of the multiple image features, generate an initial layer corresponding thereto, to obtain a layer set; select, according to the target image adjustment parameter, at least one target layer from the layer set for adjustment; combine the at least one target layer adjusted with unselected initial layers in the layer set.

[0081] In some embodiments, the recognition result includes a personal-identity, and the processor 401 configured to obtain the image adjustment parameter set according to the recognition result is configured to: obtain, from a preset identity set, a preset personal-identity that matches the personal-identity; obtain a preset image adjustment parameter set corresponding to the preset personal-identity, and use the preset image adjustment parameter set as the recognition result to obtain the image adjustment parameter set.

[0082] In some embodiments, the processor 401 is configured to: determine, from multiple preset angle intervals, a target angle interval to which the deflection angle belongs; obtain, from the image adjustment parameter set, a preset image adjustment parameter corresponding to the target angle interval; use the preset image adjustment parameter as the target image adjustment parameter.

[0083] The memory 402 is configured to store programs and data. The programs stored in the memory 402 include instructions that are executable in the processor. The programs can form various functional modules. The processor 401 executes various functional applications and data processing by running the programs stored in the memory 402.

[0084] In some embodiments, as illustrated in FIG. 14, the electronic device 400 further includes a radio frequency circuit 403, a display screen 404, a control circuit 405, an input unit 406, an audio circuit 407, a sensor 408, and a power source 409. The processor 401 is electrically coupled with the radio frequency circuit 403, the display screen 404, the control circuit 405, the input unit 406, the audio circuit 407, the sensor 408, and the power source 409, respectively.

[0085] The radio frequency circuit 403 is configured to transmit and receive (i.e., transceive) radio frequency signals to communicate with a server or other electronic devices through a wireless communication network.

[0086] The display screen 404 is configured to display information entered by a user or information provided for the user as well as various graphical user interfaces of the electronic device. The graphical user interface may be composed of images, text, icons, video, and any combination thereof.

[0087] The control circuit 405 is electrically coupled with the display screen 404 and is configured to control the display screen 404 to display information.

[0088] The input unit 406 is configured to receive inputted digits information, character information, or user feature information (e.g., fingerprints), and to generate, according to user settings and function controls, keyboard-based, mouse-based, joystick-based, optical or trackball signal inputs, and other signal inputs.

[0089] The audio circuit 407 is configured to provide an audio interface between a user and the electronic device through a speaker and a microphone.

[0090] The sensor 408 is configured to collect external environmental information. The sensor 408 may include one or more of an ambient light sensor, an acceleration sensor, an angular velocity sensor (e.g., a gyroscope), and the like.

[0091] The power source 409 is configured for power supply of various components of electronic device 400. In some embodiments, the power source 409 may be logically coupled with the processor 401 via a power management system to enable management of charging, discharging, and power consumption through the power management system.

[0092] Although not illustrated in FIG. 14, the electronic device 400 may further include a camera, a Bluetooth module, and the like, and the disclosure will not elaborate herein.

[0093] According to the electronic device of embodiments, the recognition result can be obtained by recognizing the face image. The image adjustment parameter set is obtained according to the recognition result. The deflection angle of the face in the face image relative to the reference face in the reference face image is determined. The target image adjustment parameter is selected from the image adjustment parameter set according to the deflection angle determined. By adopting the present disclosure, according to the result of the recognition of the face image and the deflection angle of the face in the face image, the target image adjustment parameter can be determined from the image adjustment parameter set to process the face image, without manually determining how to adjust the face image, which can improve speed and efficiency of image processing and reduce operating time while reducing power consumption of an electronic device.

[0094] According to embodiments, a storage medium is further provided. The storage medium is configured to store multiple instructions. The instructions, when executed by a processor, are operable with the processor to execute any of the foregoing methods for face image processing described in embodiments.

[0095] It will be understood by those of ordinary skill in the art that all or a part of the various methods of the embodiments described above may be accomplished by means of a program to instruct associated hardware, the program may be stored in a computer-readable memory, which may include a read-only memory (ROM), a random access memory (RAM), Disk or compact disc (CD), and so on.

[0096] While the disclosure has been described in connection with certain embodiments, it is to be understood that the disclosure is not to be limited to the disclosed embodiments but, on the contrary, is intended to cover various modifications and equivalent arrangements included within the scope of the appended claims, which scope is to be accorded the broadest interpretation so as to encompass all such modifications and equivalent structures as is permitted under the law.


Claims

1. A method for face image processing, comprising:

recognizing a face image to obtain a recognition result;

obtaining an image adjustment parameter set according to the recognition result;

determining a deflection angle of a face in the face image relative to a reference face in a reference face image;

selecting a target image adjustment parameter from the image adjustment parameter set according to the deflection angle; and

processing the face image according to the target image adjustment parameter.


 
2. The method for face image processing of claim 1, wherein determining the deflection angle of the face in the face image relative to the reference face in the reference face image comprises:

obtaining the reference face image corresponding to the face image according to the recognition result;

determining a first offset reference point in the reference face of the reference face image;

determining, in the face of the face image, a second offset reference point corresponding to the first offset reference point;

obtaining position offset information of the second offset reference point relative to the first offset reference point; and

determining, according to the position offset information, the deflection angle of the face in the face image relative to the reference face in the reference face image.


 
3. The method for face image processing of claim 1, wherein determining the deflection angle of the face in the face image relative to the reference face in the reference face image comprises:

obtaining the reference face image corresponding to the face image according to the recognition result;

determining a first reference vector in the face of the face image;

determining, in the reference face of the reference face image, a second reference vector corresponding to the first reference vector;

determining an angle between the first reference vector and the second reference vector; and

determining, according to the angle, the deflection angle of the face in the face image relative to the reference face in the reference face image.


 
4. The method for face image processing of claim 3, wherein determining the first reference vector in the face of the face image comprises:

determining a first feature point and a second feature point in the face image; and

determining the first reference vector according to the first feature point and the second feature point.


 
5. The method for face image processing of claim 1, wherein processing the face image according to the target image adjustment parameter comprises:

obtaining a plurality of image features of the face image;

for each of the plurality of image features, generating an initial layer corresponding thereto, to obtain a layer set;

selecting, according to the target image adjustment parameter, at least one target layer from the layer set and adjusting the at least one target layer; and

combining the at least one target layer adjusted with unselected initial layers in the layer set.


 
6. The method for face image processing of claim 1, wherein the recognition result comprises a personal-identity, and obtaining the image adjustment parameter set according to the recognition result comprises:

obtaining, from a preset identity set, a preset personal-identity that matches the personal-identity; and

obtaining a preset image adjustment parameter set corresponding to the preset personal-identity, and using the preset image adjustment parameter set as the recognition result to obtain the image adjustment parameter set.


 
7. The method for face image processing of claim 1, wherein selecting the target image adjustment parameter from the image adjustment parameter set according to the deflection angle comprises:

determining, from a plurality of preset angle intervals, a target angle interval to which the deflection angle belongs;

obtaining, from the image adjustment parameter set, a preset image adjustment parameter corresponding to the target angle interval; and

using the preset image adjustment parameter as the target image adjustment parameter.


 
8. A device for face image processing, comprising:

a recognizing module, configured to recognize a face image to obtain a recognition result;

an obtaining module, configured to obtain an image adjustment parameter set according to the recognition result;

a determining module, configured to determine a deflection angle of a face in the face image relative to a reference face in a reference face image;

a selecting module, configured to select a target image adjustment parameter from the image adjustment parameter set according to the deflection angle; and

a processing module, configured to process the face image according to the target image adjustment parameter.


 
9. The device for face image processing of claim 8, wherein the determining module comprises:

an image obtaining sub-module, configured to obtain the reference face image corresponding to the face image according to the recognition result;

a first determining sub-module, configured to determine a first offset reference point in the reference face of the reference face image, and determine, in the face of the face image, a second offset reference point corresponding to the first offset reference point;

an information obtaining sub-module, configured to obtain position offset information of the second offset reference point relative to the first offset reference point; and

a second determining sub-module, configured to determine, according to the position offset information, the deflection angle of the face in the face image relative to the reference face in the reference face image.


 
10. The device for face image processing of claim 8, wherein the determining module comprises:

an image obtaining sub-module, configured to obtain the reference face image corresponding to the face image according to the recognition result;

a third determining sub-module, configured to determine a first reference vector in the face of the face image, and determine, in the reference face of the reference face image, a second reference vector corresponding to the first reference vector;

an angle obtaining sub-module, configured to determine an angle between the first reference vector and the second reference vector; and

a fourth determining sub-module, configured to determine, according to the angle, the deflection angle of the face in the face image relative to the reference face in the reference face image.


 
11. The device for face image processing of claim 8, wherein the processing module comprises:

a feature obtaining sub-module, configured to obtain a plurality of image features of the face image;

a generating sub-module, configured to, for each of the plurality of image features, generate an initial layer corresponding thereto, to obtain a layer set;

a selecting sub-module, configured to select, according to the target image adjustment parameter, at least one target layer from the layer set and adjust the at least one target layer;

a combining sub-module, configured to combine the at least one target layer adjusted with unselected initial layers in the layer set.


 
12. The device for face image processing of claim 8, wherein the recognition result comprises a personal-identity, and the obtaining module comprises:

an identity obtaining sub-module, configured to obtain, from a preset identity set, a preset personal-identity that matches the personal-identity;

a set obtaining sub-module, configured to obtain a preset image adjustment parameter set corresponding to the preset personal-identity, and use the preset image adjustment parameter set as the recognition result to obtain the image adjustment parameter set.


 
13. A storage medium, configured to store a plurality of instructions, the instructions which, when executed by a processor, are operable to execute the method for face image processing of any of claims 1 to 7.
 
14. An electronic device, comprising:

a memory, configured to store instructions and data; and

a processor, electrically coupled with the memory;

the processor being configured to:

recognize a face image to obtain a recognition result;

obtain an image adjustment parameter set according to the recognition result;

determine a deflection angle of a face in the face image relative to a reference face in a reference face image;

select a target image adjustment parameter from the image adjustment parameter set according to the deflection angle; and

process the face image according to the target image adjustment parameter.


 
15. The electronic device of claim 14, wherein the processor configured to determine the deflection angle of the face in the face image relative to the reference face in the reference face image is configured to:

obtain the reference face image corresponding to the face image according to the recognition result;

determine a first offset reference point in the reference face of the reference face image;

determine, in the face of the face image, a second offset reference point corresponding to the first offset reference point;

obtain position offset information of the second offset reference point relative to the first offset reference point; and

determine, according to the position offset information, the deflection angle of the face in the face image relative to the reference face in the reference face image.


 
16. The electronic device of claim 14, wherein the processor configured to determine the deflection angle of the face in the face image relative to the reference face in the reference face image is configured to:

obtain the reference face image corresponding to the face image according to the recognition result;

determine a first reference vector in the face of the face image;

determine, in the reference face of the reference face image, a second reference vector corresponding to the first reference vector;

determine an angle between the first reference vector and the second reference vector; and

determine, according to the angle, the deflection angle of the face in the face image relative to the reference face in the reference face image.


 
17. The electronic device of claim 16, wherein the processor configured to determine the first reference vector in the face of the face image is configured to:

determine a first feature point and a second feature point in the face image; and

determine the first reference vector according to the first feature point and the second feature point.


 
18. The electronic device of claim 14, wherein the processor configured to process the face image according to the target image adjustment parameter is configured to:

obtain a plurality of image features of the face image;

for each of the plurality of image features, generate an initial layer corresponding thereto, to obtain a layer set;

select, according to the target image adjustment parameter, at least one target layer from the layer set and adjust the at least one target layer; and

combine the at least one target layer adjusted with unselected initial layers in the layer set.


 
19. The electronic device of claim 14, wherein the recognition result comprises a personal-identity, and the processor configured to obtain the image adjustment parameter set according to the recognition result is configured to:

obtain, from a preset identity set, a preset personal-identity that matches the personal-identity; and

obtain a preset image adjustment parameter set corresponding to the preset personal-identity, and use the preset image adjustment parameter set as the recognition result to obtain the image adjustment parameter set.


 
20. The electronic device of claim 14, wherein the processor configured to select the target image adjustment parameter from the image adjustment parameter set according to the deflection angle is configured to:

determine, from a plurality of preset angle intervals, a target angle interval to which the deflection angle belongs;

obtain, from the image adjustment parameter set, a preset image adjustment parameter corresponding to the target angle interval; and

use the preset image adjustment parameter as the target image adjustment parameter.


 




Drawing