CN114187166A

CN114187166A - Image processing method, intelligent terminal and storage medium

Info

Publication number: CN114187166A
Application number: CN202111349959.2A
Authority: CN
Inventors: 赵玮; 周凡贻
Original assignee: Shanghai Chuanying Information Technology Co Ltd
Current assignee: Shanghai Chuanying Information Technology Co Ltd
Priority date: 2021-11-15
Filing date: 2021-11-15
Publication date: 2022-03-15

Abstract

The application discloses an image processing method, an intelligent terminal and a storage medium. The method comprises the following steps: detecting at least one target portrait in an initial image, acquiring appearance feature information of the target portrait, performing facial analysis on the target portrait to obtain expression information and/or makeup information of the target portrait, determining beauty parameters of the target portrait according to at least one of the appearance feature information, the expression information and the makeup information, and processing the initial image according to the beauty parameters to obtain the target image. The method and the device can automatically perform personalized beauty according to at least one of appearance characteristics, expressions and makeup of the user, so that the diversity and the use efficiency of image processing are improved.

Description

Image processing method, intelligent terminal and storage medium

Technical Field

The present application relates to the field of image processing technologies, and in particular, to an image processing method, an intelligent terminal, and a storage medium.

Background

With the rapid development of internet technology and mobile clients, more and more users use intelligent terminals for learning, entertainment and the like, so that the life of the users is enriched, and convenience is brought to the users. The functions of smart terminals (such as mobile phones, tablet computers, etc.) are also becoming more powerful, for example, taking pictures, etc. using smart terminals. The requirements of the current terminal on the self-photographing technology are higher and higher, and for example, self-photographing definition, focusing, whitening, skin grinding, image enhancement and the like become key factors of the current mobile phone self-photographing technology.

In the course of conceiving and implementing the present application, the inventors found that at least the following problems existed: the intelligent degree of the beautifying mode in some implementations is low, the beautifying mode is relatively single, and the picture provided by the user can only be beautified according to the preset beautifying mode, but the personalized beautifying requirement cannot be provided. For example, when a picture is beautified, overall beautification of a static image is generally performed, but this method cannot meet the personalized requirements of different users for individually beautifying different characters, manual beautification requires very complicated steps, and user experience is poor.

The foregoing description is provided for general background information and is not admitted to be prior art.

Disclosure of Invention

In view of the above technical problems, the present application provides an image processing method, an intelligent terminal and a storage medium, which can automatically perform personalized beauty according to at least one of appearance features, expressions and makeup of a user, thereby improving the diversity and the use efficiency of image processing.

In order to solve the above technical problem, the present application provides an image processing method, optionally applied to an intelligent terminal, including the following steps:

s10: acquiring at least one target portrait in the initial image;

s20: analyzing the target portrait to obtain key information of the target portrait;

s30: and according to the key information, carrying out preset processing on the initial image to obtain a target image.

Optionally, the step of S10 includes:

detecting at least one portrait in the initial image;

respectively acquiring the portrait position and/or the face area of each portrait;

and determining a road portrait in the at least one portrait according to the portrait position and/or the face area, and removing the road portrait to obtain the at least one target portrait.

Optionally, the key information of the target portrait includes gender information and/or age information, and the step S20 includes:

acquiring feature point information of the target portrait, and determining gender information of the target portrait according to the feature point information; and/or the presence of a gas in the gas,

and acquiring skin surface layer characteristic information of the target portrait, and determining age information of the target portrait according to the skin surface layer characteristic information.

Optionally, the key information of the target portrait includes race information, and the step S20 includes:

acquiring the skin color and/or eyeball color and/or hair color and/or head contour and/or face contour of the target portrait;

and matching in a preset human species information base according to the skin color and/or the eyeball color and/or the hair color and/or the head outline and/or the face outline so as to determine the human species information of the target human figure.

Optionally, the step of S20 includes:

acquiring a face image of the target portrait, and extracting facial morphological features and/or muscle morphological features from the face image;

and determining key information of the target portrait according to the morphological characteristics of the five sense organs and/or the morphological characteristics of muscles.

Optionally, the step of S20 includes:

acquiring a facial image of the target portrait, and extracting eyebrow states and/or eyelid colors and/or cheek colors and/or lip colors and/or skin brightness difference values from the facial image;

and determining key information of the target portrait according to the eyebrow state and/or eyelid color and/or cheek color and/or lip color and/or skin brightness difference value.

Optionally, the step S30 further includes:

inputting the key information into a preset network model;

and acquiring the predicted beauty parameters output by the preset network model.

Optionally, the step of S30 includes:

segmenting the initial image to respectively obtain a portrait image and a scene image;

identifying the scene image to obtain a scene type, and identifying the portrait image to obtain a light ray type;

matching a target filter in a preset filter library according to the scene type and the light ray type;

performing first processing on the scene image according to the target filter, and performing second processing on the portrait image according to the beauty parameters;

and synthesizing the images subjected to the first processing and the second processing to obtain the target image.

Optionally, the method further comprises:

monitoring the memory occupation value and/or the CPU temperature and/or the power consumption of the terminal in unit time;

if the memory occupation value is greater than a first preset value and/or the CPU temperature is greater than a second preset value and/or the power consumption per unit time is greater than a third preset value, adjusting the beauty parameter and processing the initial image according to the adjusted beauty parameter to obtain a target image, and/or,

and reducing the number of the target figures, and processing the initial image according to the reduced target figures and the beauty parameters to obtain a target image.

In order to solve the above technical problem, the present application further provides an image processing method, optionally applied to an intelligent terminal, including the following steps:

s10: detecting at least one target portrait in an initial image, and acquiring the appearance characteristic information of the target portrait;

s20: performing facial analysis on the target portrait to obtain expression information and/or makeup information of the target portrait;

s30: determining beauty parameters of the target portrait according to at least one of the appearance feature information, the expression information and the makeup information;

s40: and processing the initial image according to the beauty parameters to obtain a target image.

Optionally, the step of S10 includes:

detecting at least one portrait in the initial image;

Optionally, the appearance feature information of the target portrait includes gender information and/or age information, and the step of obtaining the gender information and/or the age information of the target portrait includes:

Optionally, the appearance feature information of the target portrait includes race information, and the step of obtaining the race information of the target portrait includes:

Optionally, the step of S20 includes:

and determining the expression information of the target portrait according to the morphological characteristics of the five sense organs and/or the morphological characteristics of muscles.

Optionally, the step of S20 includes:

and determining the makeup information of the target portrait according to the eyebrow state and/or the eyelid color and/or the cheek color and/or the lip color and/or the skin brightness difference value.

Optionally, the step of S30 includes:

inputting at least one of the appearance feature information, the expression information and the makeup information into a preset network model, wherein optionally, the preset network model is a neural network model obtained by training by taking a user historical target image as a sample set;

Optionally, the step of S40 includes:

Optionally, the method further comprises:

if the memory occupation value is larger than a first preset value and/or the CPU temperature is larger than a second preset value and/or the power consumption per unit time is larger than a third preset value, adjusting the beauty parameter and processing the initial image according to the adjusted beauty parameter to obtain a target image and/or

The application also provides an intelligent terminal, including: a memory, a processor, wherein the memory has stored thereon an image processing program, which when executed by the processor implements the steps of any of the methods described above.

The present application also provides a computer-readable storage medium, which stores a computer program that, when executed by a processor, performs the steps of the method as set forth in any one of the above.

As described above, the image processing method of the present application is applied to an intelligent terminal, and may detect at least one target portrait in an initial image, obtain appearance feature information of the target portrait, perform face analysis on the target portrait to obtain expression information and/or makeup information of the target portrait, determine a beauty parameter of the target portrait according to at least one of the appearance feature information, the expression information, and the makeup information, and process the initial image according to the beauty parameter to obtain the target image. The method and the device can automatically perform personalized beauty according to at least one of appearance characteristics, expressions and makeup of the user, so that the diversity and the use efficiency of image processing are improved.

Drawings

The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and together with the description, serve to explain the principles of the application. In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings needed to be used in the description of the embodiments will be briefly described below, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without inventive exercise.

Fig. 1 is a schematic diagram of a hardware structure of an intelligent terminal implementing various embodiments of the present application;

fig. 2 is a communication network system architecture diagram according to an embodiment of the present application;

fig. 3 is a first flowchart of an image processing method according to an embodiment of the present application;

fig. 4 is a second flowchart of an image processing method according to an embodiment of the present application;

fig. 5 is a scene schematic diagram of determining a target portrait in an image processing method provided in an embodiment of the present application;

fig. 6 is an exemplary diagram of a feature point set in the image processing method according to the embodiment of the present invention;

fig. 7 is a diagram illustrating another example of feature point sets in an image processing method according to an embodiment of the present invention;

fig. 8 is a third flowchart illustrating an image processing method according to an embodiment of the present application;

fig. 9 is a scene schematic diagram of image segmentation in the image processing method according to the embodiment of the present application;

fig. 10 is a schematic structural diagram of an image processing apparatus according to an embodiment of the present application;

fig. 11 is a schematic structural diagram of another image processing apparatus according to an embodiment of the present application.

The implementation, functional features and advantages of the objectives of the present application will be further explained with reference to the accompanying drawings. With the above figures, there are shown specific embodiments of the present application, which will be described in more detail below. These drawings and written description are not intended to limit the scope of the inventive concepts in any manner, but rather to illustrate the inventive concepts to those skilled in the art by reference to specific embodiments.

Detailed Description

Reference will now be made in detail to the exemplary embodiments, examples of which are illustrated in the accompanying drawings. When the following description refers to the accompanying drawings, like numbers in different drawings represent the same or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present application. Rather, they are merely examples of apparatus and methods consistent with certain aspects of the present application, as detailed in the appended claims.

It should be noted that, in this document, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, the recitation of an element by the phrase "comprising an … …" does not exclude the presence of additional like elements in the process, method, article, or apparatus that comprises the element, and further, where similarly-named elements, features, or elements in different embodiments of the disclosure may have the same meaning, or may have different meanings, that particular meaning should be determined by their interpretation in the embodiment or further by context with the embodiment.

It should be understood that although the terms first, second, third, etc. may be used herein to describe various information, such information should not be limited to these terms. These terms are only used to distinguish one type of information from another. For example, first information may also be referred to as second information, and similarly, second information may also be referred to as first information, without departing from the scope herein. The word "if" as used herein may be interpreted as "at … …" or "when … …" or "in response to a determination", depending on the context. Also, as used herein, the singular forms "a", "an" and "the" are intended to include the plural forms as well, unless the context indicates otherwise. It will be further understood that the terms "comprises," "comprising," "includes" and/or "including," when used in this specification, specify the presence of stated features, steps, operations, elements, components, items, species, and/or groups, but do not preclude the presence, or addition of one or more other features, steps, operations, elements, components, species, and/or groups thereof. The terms "or," "and/or," "including at least one of the following," and the like, as used herein, are to be construed as inclusive or mean any one or any combination. For example, "includes at least one of: A. b, C "means" any of the following: a; b; c; a and B; a and C; b and C; a and B and C ", again for example," A, B or C "or" A, B and/or C "means" any of the following: a; b; c; a and B; a and C; b and C; a and B and C'. An exception to this definition will occur only when a combination of elements, functions, steps or operations are inherently mutually exclusive in some way.

It should be understood that, although the steps in the flowcharts in the embodiments of the present application are shown in order as indicated by the arrows, the steps are not necessarily performed in order as indicated by the arrows. The steps are not performed in the exact order shown and may be performed in other orders unless explicitly stated herein. Moreover, at least some of the steps in the figures may include multiple sub-steps or multiple stages that are not necessarily performed at the same time, but may be performed at different times, in different orders, and may be performed alternately or at least partially with respect to other steps or sub-steps of other steps.

The words "if", as used herein, may be interpreted as "at … …" or "at … …" or "in response to a determination" or "in response to a detection", depending on the context. Similarly, the phrases "if determined" or "if detected (a stated condition or event)" may be interpreted as "when determined" or "in response to a determination" or "when detected (a stated condition or event)" or "in response to a detection (a stated condition or event)", depending on the context.

It should be noted that step numbers such as S10 and S20 are used herein for the purpose of more clearly and briefly describing the corresponding content, and do not constitute a substantial limitation on the sequence, and those skilled in the art may perform S20 first and then S10 in specific implementation, which should be within the scope of the present application.

It should be understood that the specific embodiments described herein are merely illustrative of the present application and are not intended to limit the present application.

In the following description, suffixes such as "module", "component", or "unit" used to denote elements are used only for the convenience of description of the present application, and have no specific meaning in themselves. Thus, "module", "component" or "unit" may be used mixedly.

The smart terminal may be implemented in various forms. For example, the smart terminal described in the present application may include a mobile terminal such as a mobile phone, a tablet computer, a notebook computer, a palmtop computer, a Personal Digital Assistant (PDA), a Portable Media Player (PMP), a navigation device, a wearable device, a smart band, a pedometer, etc., and may further include a fixed terminal such as a Digital TV, a desktop computer, etc., in other embodiments.

The following description will be given taking a mobile terminal as an example, and it will be understood by those skilled in the art that the configuration according to the embodiment of the present application can be applied to a fixed type terminal in addition to elements particularly used for mobile purposes.

Referring to fig. 1, which is a schematic diagram of a hardware structure of a mobile terminal for implementing various embodiments of the present application, the mobile terminal 100 may include: RF (Radio Frequency) unit 101, WiFi module 102, audio output unit 103, a/V (audio/video) input unit 104, sensor 105, display unit 106, user input unit 107, interface unit 108, memory 109, processor 110, and power supply 111. Those skilled in the art will appreciate that the mobile terminal architecture shown in fig. 1 is not intended to be limiting of mobile terminals, which may include more or fewer components than those shown, or some components may be combined, or a different arrangement of components.

The various components of the mobile terminal are optionally described below in conjunction with fig. 1:

the radio frequency unit 101 may be configured to receive and transmit signals during information transmission and reception or during a call, and optionally, receive downlink information of a base station and then process the downlink information to the processor 110; optionally, the uplink data is sent to the base station. Typically, radio frequency unit 101 includes, but is not limited to, an antenna, at least one amplifier, a transceiver, a coupler, a low noise amplifier, a duplexer, and the like. In addition, the radio frequency unit 101 can also communicate with a network and other devices through wireless communication. The wireless communication may use any communication standard or protocol, including but not limited to GSM (Global System for Mobile communications), GPRS (General Packet Radio Service), CDMA2000(Code Division Multiple Access 2000), WCDMA (Wideband Code Division Multiple Access), TD-SCDMA (Time Division-Synchronous Code Division Multiple Access), FDD-LTE (Frequency Division duplex-Long Term Evolution), TDD-LTE (Time Division duplex-Long Term Evolution, Time Division Long Term Evolution), 5G, and so on.

WiFi belongs to short-distance wireless transmission technology, and the mobile terminal can help a user to receive and send e-mails, browse webpages, access streaming media and the like through the WiFi module 102, and provides wireless broadband internet access for the user. Although fig. 1 shows the WiFi module 102, it is understood that it does not belong to the essential constitution of the mobile terminal, and may be omitted entirely as needed within the scope not changing the essence of the invention.

The audio output unit 103 may convert audio data received by the radio frequency unit 101 or the WiFi module 102 or stored in the memory 109 into an audio signal and output as sound when the mobile terminal 100 is in a call signal reception mode, a call mode, a recording mode, a voice recognition mode, a broadcast reception mode, or the like. Also, the audio output unit 103 may also provide audio output related to a specific function performed by the mobile terminal 100 (e.g., a call signal reception sound, a message reception sound, etc.). The audio output unit 103 may include a speaker, a buzzer, and the like.

The a/V input unit 104 is used to receive audio or video signals. The a/V input Unit 104 may include a Graphics Processing Unit (GPU) 1041 and a microphone 1042, the Graphics processor 1041 Processing image data of still pictures or video obtained by an image capturing device (e.g., a camera) in a video capturing mode or an image capturing mode. The processed image frames may be displayed on the display unit 106. The image frames processed by the graphic processor 1041 may be stored in the memory 109 (or other storage medium) or transmitted via the radio frequency unit 101 or the WiFi module 102. The microphone 1042 may receive sounds (audio data) via the microphone 1042 in a phone call mode, a recording mode, a voice recognition mode, or the like, and may be capable of processing such sounds into audio data. The processed audio (voice) data may be converted into a format output transmittable to a mobile communication base station via the radio frequency unit 101 in case of a phone call mode. The microphone 1042 may implement various types of noise cancellation (or suppression) algorithms to cancel (or suppress) noise or interference generated in the course of receiving and transmitting audio signals.

The mobile terminal 100 also includes at least one sensor 105, such as a light sensor, a motion sensor, and other sensors. Optionally, the light sensor includes an ambient light sensor that may adjust the brightness of the display panel 1061 according to the brightness of ambient light, and a proximity sensor that may turn off the display panel 1061 and/or the backlight when the mobile terminal 100 is moved to the ear. As one of the motion sensors, the accelerometer sensor can detect the magnitude of acceleration in each direction (generally, three axes), can detect the magnitude and direction of gravity when stationary, and can be used for applications of recognizing the posture of a mobile phone (such as horizontal and vertical screen switching, related games, magnetometer posture calibration), vibration recognition related functions (such as pedometer and tapping), and the like; as for other sensors such as a fingerprint sensor, a pressure sensor, an iris sensor, a molecular sensor, a gyroscope, a barometer, a hygrometer, a thermometer, and an infrared sensor, which can be configured on the mobile phone, further description is omitted here.

The display unit 106 is used to display information input by a user or information provided to the user. The Display unit 106 may include a Display panel 1061, and the Display panel 1061 may be configured in the form of a Liquid Crystal Display (LCD), an Organic Light-Emitting Diode (OLED), or the like.

The user input unit 107 may be used to receive input numeric or character information and generate key signal inputs related to user settings and function control of the mobile terminal. Alternatively, the user input unit 107 may include a touch panel 1071 and other input devices 1072. The touch panel 1071, also referred to as a touch screen, may collect a touch operation performed by a user on or near the touch panel 1071 (e.g., an operation performed by the user on or near the touch panel 1071 using a finger, a stylus, or any other suitable object or accessory), and drive a corresponding connection device according to a predetermined program. The touch panel 1071 may include two parts of a touch detection device and a touch controller. Optionally, the touch detection device detects a touch orientation of a user, detects a signal caused by a touch operation, and transmits the signal to the touch controller; the touch controller receives touch information from the touch sensing device, converts the touch information into touch point coordinates, sends the touch point coordinates to the processor 110, and can receive and execute commands sent by the processor 110. In addition, the touch panel 1071 may be implemented in various types, such as a resistive type, a capacitive type, an infrared ray, and a surface acoustic wave. In addition to the touch panel 1071, the user input unit 107 may include other input devices 1072. Optionally, other input devices 1072 may include, but are not limited to, one or more of a physical keyboard, function keys (e.g., volume control keys, switch keys, etc.), a trackball, a mouse, a joystick, and the like, and are not limited thereto.

Alternatively, the touch panel 1071 may cover the display panel 1061, and when the touch panel 1071 detects a touch operation thereon or nearby, the touch panel 1071 transmits the touch operation to the processor 110 to determine the type of the touch event, and then the processor 110 provides a corresponding visual output on the display panel 1061 according to the type of the touch event. Although the touch panel 1071 and the display panel 1061 are shown in fig. 1 as two separate components to implement the input and output functions of the mobile terminal, in some embodiments, the touch panel 1071 and the display panel 1061 may be integrated to implement the input and output functions of the mobile terminal, and is not limited herein.

The interface unit 108 serves as an interface through which at least one external device is connected to the mobile terminal 100. For example, the external device may include a wired or wireless headset port, an external power supply (or battery charger) port, a wired or wireless data port, a memory card port, a port for connecting a device having an identification module, an audio input/output (I/O) port, a video I/O port, an earphone port, and the like. The interface unit 108 may be used to receive input (e.g., data information, power, etc.) from external devices and transmit the received input to one or more elements within the mobile terminal 100 or may be used to transmit data between the mobile terminal 100 and external devices.

The memory 109 may be used to store software programs as well as various data. The memory 109 may mainly include a program storage area and a data storage area, and optionally, the program storage area may store an operating system, an application program (such as a sound playing function, an image playing function, and the like) required by at least one function, and the like; the storage data area may store data (such as audio data, a phonebook, etc.) created according to the use of the cellular phone, and the like. Further, the memory 109 may include high speed random access memory, and may also include non-volatile memory, such as at least one magnetic disk storage device, flash memory device, or other volatile solid state storage device.

The processor 110 is a control center of the mobile terminal, connects various parts of the entire mobile terminal using various interfaces and lines, and performs various functions of the mobile terminal and processes data by operating or executing software programs and/or modules stored in the memory 109 and calling data stored in the memory 109, thereby performing overall monitoring of the mobile terminal. Processor 110 may include one or more processing units; preferably, the processor 110 may integrate an application processor and a modem processor, optionally, the application processor mainly handles operating systems, user interfaces, application programs, etc., and the modem processor mainly handles wireless communications. It will be appreciated that the modem processor described above may not be integrated into the processor 110.

The mobile terminal 100 may further include a power supply 111 (e.g., a battery) for supplying power to various components, and preferably, the power supply 111 may be logically connected to the processor 110 via a power management system, so as to manage charging, discharging, and power consumption management functions via the power management system.

Although not shown in fig. 1, the mobile terminal 100 may further include a bluetooth module or the like, which is not described in detail herein.

In order to facilitate understanding of the embodiments of the present application, a communication network system on which the mobile terminal of the present application is based is described below.

Referring to fig. 2, fig. 2 is an architecture diagram of a communication Network system according to an embodiment of the present disclosure, where the communication Network system is an LTE system of a universal mobile telecommunications technology, and the LTE system includes a UE (User Equipment) 201, an E-UTRAN (Evolved UMTS Terrestrial Radio Access Network) 202, an EPC (Evolved Packet Core) 203, and an IP service 204 of an operator, which are in communication connection in sequence.

Optionally, the UE201 may be the mobile terminal 100 described above, and is not described herein again.

The E-UTRAN202 includes eNodeB2021 and other eNodeBs 2022, among others. Alternatively, the eNodeB2021 may be connected with other enodebs 2022 through a backhaul (e.g., X2 interface), the eNodeB2021 is connected to the EPC203, and the eNodeB2021 may provide the UE201 access to the EPC 203.

The EPC203 may include an MME (Mobility Management Entity) 2031, an HSS (Home Subscriber Server) 2032, other MMEs 2033, an SGW (Serving gateway) 2034, a PGW (PDN gateway) 2035, and a PCRF (Policy and Charging Rules Function) 2036, and the like. Optionally, the MME2031 is a control node that handles signaling between the UE201 and the EPC203, providing bearer and connection management. HSS2032 is used to provide registers to manage functions such as home location register (not shown) and holds subscriber specific information about service characteristics, data rates, etc. All user data may be sent through SGW2034, PGW2035 may provide IP address assignment for UE201 and other functions, and PCRF2036 is a policy and charging control policy decision point for traffic data flow and IP bearer resources, which selects and provides available policy and charging control decisions for a policy and charging enforcement function (not shown).

The IP services 204 may include the internet, intranets, IMS (IP Multimedia Subsystem), or other IP services, among others.

Although the LTE system is described as an example, it should be understood by those skilled in the art that the present application is not limited to the LTE system, but may also be applied to other wireless communication systems, such as GSM, CDMA2000, WCDMA, TD-SCDMA, and future new network systems (e.g. 5G), and the like.

Based on the above mobile terminal hardware structure and communication network system, various embodiments of the present application are provided.

Referring to fig. 3, fig. 3 is a first flowchart illustrating an image processing method according to an embodiment of the present disclosure. The flow of the image processing method may include:

and S11, detecting at least one target portrait in the initial image and acquiring the appearance characteristic information of the target portrait.

The execution main body of the embodiment of the application can be an intelligent terminal, and can also be an image processing device arranged in the intelligent terminal. Alternatively, the image processing apparatus may be implemented by software, or may be implemented by a combination of software and hardware. The intelligent terminal may be a terminal device such as a smart phone, a tablet Computer, a notebook Computer, a super mobile Personal Computer (UMPC), a netbook, a Personal Digital Assistant (PDA), and the like, but is not limited thereto.

Alternatively, an initial image is acquired, and a specific implementation manner of acquiring the initial image may be to acquire a photo of the person in a digital image format (e.g., BMP, JPG, etc.), such as taking a photo of the person by a digital camera or a mobile phone instantly. Persons skilled in the art will readily appreciate that the image of the person may also be obtained by means of video capture, photo scanning, and the like, which is not limited in this embodiment of the present invention.

Optionally, the initial image is an image including a face, the electronic device may further obtain an image from a multi-user family shared album in the local device and/or other devices, or may also obtain an image shared by other devices, where the shared image is an image in a multi-user family shared album to which the electronic device has access authority, and identify and obtain the face in the image.

The electronic device may receive an image acquisition instruction, which may turn on a camera of the electronic device and take a picture of a current scene. And further judging whether the image has a face or not aiming at the acquired image, and if so, continuously detecting at least one target portrait in the initial image.

Considering that if at least one portrait exists in the current picture, a target portrait in the at least one portrait may be further determined, for example, the user clicks to select the target portrait, and then selects the appearance feature information of the target portrait. Alternatively, the appearance feature information may include facial feature information, for example, obtained by a face recognition technique, which is a biometric technique for performing identification based on facial feature information of a person. A series of related technologies, also commonly called face recognition or face recognition, etc., are used to capture an image or video stream containing a human face with a camera or a video camera, automatically detect and track the human face in the image, and then perform face recognition on the detected human face. Optionally, the face recognition technology may use an adaptive boosting (adaptive boosting) algorithm based on Haar features to detect the face in the original image, or use another algorithm to detect the face in the original image, which is not limited in this embodiment.

As a possible implementation manner, facial feature information of the registered user, for example, special mark features such as birthmarks and the like, shape and position features of five sense organs such as a nose and eyes and the like, can be obtained in advance, then a two-dimensional face image in the original image is analyzed, for example, the facial features of the user are extracted by adopting an image recognition technology, a pre-registered facial database is queried, whether corresponding facial features exist or not is judged, and if the facial features exist, the user is determined to be registered; and/or if not, determining that the user is not registered. For a registered user, the appearance feature information related to the user can be directly obtained from the database. Optionally, the database may store a mapping relationship between the registered user and the appearance feature information.

Optionally, the portrait appearance feature information may include at least one of: person identity, race, age or gender, etc.

For example, the face database stores the corresponding relationship between the face and the person identity, the race, the age, or the gender, and the person identity, the age, or the gender corresponding to the face in the image can be obtained by matching the face in the target person image with the face in the database. Optionally, for a target portrait that is not matched in the face database, feature extraction may be performed on the target portrait, and appearance feature information of the target portrait is determined according to the feature information, and optionally, the appearance feature information is used to determine beauty parameters.

And S12, carrying out facial analysis on the target portrait to obtain the expression information and/or the makeup information of the target portrait.

Optionally, facial analysis may be performed on the target portrait first, features of the five sense organs are extracted, and then feature information of the five sense organs is input into a preset expression classifier to obtain predicted expression information of the user.

Expression types may include, but are not limited to, serious, smiling, silly, sold, funny, and the like. Optionally, the expression type is used to determine a beauty parameter, the beauty parameter corresponds to the expression type information, and different beauty parameters correspond to different beauty styles.

Alternatively, the makeup information of the target portrait may also be determined by performing facial analysis on the target portrait, extracting skin color features and skin texture features. Alternatively, the skin color feature may be used to represent the color, brightness, etc. of the presentation of the skin of the human face, and the skin color feature may include color information and brightness information, etc. of the skin area in the human face area. The skin characteristics may be used to represent a state of the skin of the human face, and the skin characteristics may include texture information, edge strength, and the like of a skin region in the human face region, optionally, the texture information refers to a texture distribution condition of the skin region, such as texture thickness, density, and the like, the edge information may include pixel points in the skin region that have a step change or a roof change, and the edge strength may refer to a change degree of the pixel points that have a step change or a roof change.

In one embodiment, after the intelligent terminal performs face recognition on the target image, a skin area in the face area may be determined first, and then facial features such as skin color features and skin texture features may be extracted from the skin area. Optionally, the electronic device may calculate an average value of each component of each pixel point included in the skin region in the YUV color space, and use the average value as a skin color feature of the skin region. The YUV color space may include a luminance signal Y and two chrominance signals B-Y (i.e., U), R-Y (i.e., V), optionally the Y component representing brightness, which may be a gray scale value, and U and V representing chrominance, which may be used to describe the color and saturation of the image, the luminance signal Y and the chrominance signal U, V of the YUV color space being separate. The electronic equipment can calculate the mean values of all pixel points contained in the skin area in the Y component, the U component and the V component, and the mean values in the Y component, the U component and the V component are used as skin color features.

Optionally, the intelligent terminal extracts skin characteristics, edge detection may be performed on the skin region first, so as to obtain edge information and texture information of the skin region, and optionally, the edge information may include information such as a position or a direction of an edge pixel point. Edge detection can employ a variety of edge detection operators, such as the Roberts Cross operator, Prewitt operator, Sobel operator, Kirsch operator, compass operator, and the like. The electronic equipment can find pixel points with gray values changing in a step change or a roof change and the like in the skin area through edge detection, and the pixel points with the gray values changing in the step change or the roof change and the like can be determined as edge pixel points in the skin area. After the intelligent terminal collects information such as the position or the direction of the edge pixel point, the texture complexity can be calculated according to the information such as the position or the direction of the edge pixel point, and the skin characteristic of the skin area is obtained.

Makeup types may include, but are not limited to, no makeup, light makeup, heavy makeup, and the like. Optionally, the makeup type is used to determine a beauty parameter, the beauty parameter corresponds to the makeup type information, and different beauty parameters correspond to different beauty styles.

And S13, determining the beauty parameters of the target portrait according to at least one of the appearance feature information, the expression information and the makeup information.

Optionally, the beauty parameters include parameters respectively corresponding to a plurality of types, such as skin color, skin texture, face shape, ear shape, eyebrow color, hair color, lip color, pupil color, position of five sense organs, size ratio of five sense organs to human face, and the like. After at least one of the appearance feature information, the expression information, and the makeup information is acquired, a beauty parameter corresponding to the target portrait may be generated based thereon.

Taking an example of setting beauty parameters according to the appearance feature information, if the race of the target portrait is caucasian and yellow, the beauty parameters corresponding to the race can be the skin colors of the target portrait, which are white and ruddy; and/or if the race of the target portrait is a black person, the beauty parameter corresponding to the race can be the skin color of the target portrait, and the brightening intensity has a positive relation with the age; and/or if the gender of the target portrait is female, the beauty parameter corresponding to the gender can be the reduction of skin surface pores of the target portrait, and the reduction degree has an inverse relation with the age; and/or if the gender of the target portrait is male, the beauty parameter corresponding to the gender can be keeping or reducing skin surface pores of the target portrait; and/or selecting the skin surface wrinkles of the target portrait within a preset threshold value corresponding to the age of the user according to the age of the user.

For example, the facial beautification parameters are set according to the expression information, if the expression type is serious, the initial image containing the target portrait can be identified as a certificate photo or an image acquired in a formal occasion, the corresponding facial beautification parameters can be basic facial beautification parameters, the basic facial beautification parameters include but are not limited to image color adjustment parameters, noise point modification parameters and the like, and the noise point modification parameters are used for changing skin color (the effect of whitening and blushing can be realized by changing skin color), skin smoothness and the like; and/or if the expression type is smiling, selling or funny, the corresponding beauty parameters can comprise basic beauty parameters and can also comprise five sense organs deformation adjusting parameters, and the five sense organs deformation adjusting parameters comprise eye adjusting parameters, ear deformation adjusting parameters and face deformation adjusting parameters. The eye adjustment parameters can be used for realizing the beauty effect of big eyes, and the ear deformation adjustment parameters can realize the beauty effects of sharp ears, big ears and the like. The face deformation adjustment parameters are used to change the face contour.

Setting a beauty parameter by using the makeup information for example, if the makeup type is makeup-free, the value of the beauty parameter can be properly increased; and/or, if the makeup type is light makeup, the value of the beauty parameter can be properly reduced; and/or if the makeup type is thick makeup, the corresponding beauty parameter can be a basic beauty parameter, and the target portrait can be not beautified, namely the beauty parameter is not set.

In an embodiment of the application, before step S13, a preset network model may be obtained, and optionally: establishing a beauty model which can comprise a feature extraction convolution layer, a nonlinear mapping convolution layer and a reconstruction convolution layer; setting a beauty cost function, wherein the beauty cost function comprises a content cost function and a style cost function; and training the beauty model by using big data (for example, at least one of photos before and after beauty of a large number of users, appearance characteristic information, expression information and makeup information of the photos before beauty as a data set X, and beauty parameters of the photos after beauty as a label Y) according to the beauty model and the beauty cost function, and storing the beauty model to the intelligent terminal.

Optionally, the feature extraction convolutional layer may be used to extract the primary features of the input photograph; the nonlinear mapping convolutional layers can be used for extracting high-level features of the input photo, and the number of the nonlinear mapping convolutional layers can be several; the reconstructed convolutional layer may be used for upsampling to reconstruct a picture with the same resolution as the picture after beauty. The content cost function can be used for representing the difference between the content of the beautified photo and the content of the input photo; the style cost function may be used to characterize the difference between the style of the post-beauty photograph and the style of the input photograph. After training is finished, at least one of the obtained appearance characteristic information, expression information and makeup information of the target portrait is input into a preset network model, and then the output predicted beauty parameters can be obtained. That is, the S13 step may include:

And S14, processing the initial image according to the beauty parameters to obtain a target image.

Optionally, because the regions of the target portrait image to be beautified in the initial image are different, the brightness of the photographed portrait image may be low, and if the brightness of the portrait image is low, it is difficult to distinguish the detail region from the flat region when the face beautification is performed, so after the image of the target portrait image is obtained, the operation of raising the details of the dark region may be performed on the portrait image. For example, the pixel value of the portrait image may be adjusted by using a preset pixel value correction algorithm in response to the fact that the brightness of the portrait image is lower than a preset threshold, so that the brightness of the portrait image may reach the preset threshold.

Optionally, the step of processing the initial image according to the beauty parameters may include skin polishing according to parameters corresponding to skin color and skin type, color matching according to parameters corresponding to eyebrow color, hair color, lip color and pupil color, face thinning, ear trimming, eyebrow trimming and the like according to parameters corresponding to face shape, ear shape and eyebrow shape. Optionally, taking human face peeling as an example for description, the terminal system may store a preset human face peeling algorithm, where the preset human face peeling algorithm is an algorithm for image filtering used in human face peeling processing, such as a bilateral filtering algorithm, a Non-Local-Means (Non-Local mean) filtering algorithm, a BM3D (Block-Matching and 3D, 3-dimensional Block Matching) filtering algorithm, and the like. Due to the factors of higher algorithm complexity, limited computing resources and the like of the Non-Local-Means filtering algorithm and the BM3D filtering algorithm, in practical application, a bilateral filtering algorithm is usually adopted.

It should be noted that, in the process of face peeling, in order to keep details near eyes, nose, mouth, bang, and the like, pixels with large texture complexity weight may be slightly peeled or not peeled, and pixels with small texture complexity weight (e.g., pixels in the regions of face, forehead, and the like) may be mainly peeled. And performing weighted calculation on the portrait image so that the detail area is reserved, and the flat area is subjected to buffing treatment.

Optionally, the method for processing the target portrait in the initial image may further include, in addition to the beautifying, brightness optimization, sharpness improvement, denoising processing, obstacle processing, and the like, so as to ensure that the original human face three-dimensional model is relatively accurate.

Optionally, in this embodiment, the facial orientation of the target portrait and the three-dimensional structure of each organ of the face may be further determined, the facial shape image and the skin image are drawn on the plane according to the beauty parameters, the skin image on the plane is rotated to an angle consistent with the face through three-dimensional transformation, and finally the rotated skin image is attached to the portrait image to be beautified, so as to obtain the target image.

Optionally, the beauty parameter may further include a beauty level, and the application may receive a user instruction during the process of processing the initial image according to the beauty parameter, and adjust the beauty level according to the user instruction. For example, the beauty level is provided with four levels of 1 to 4, alternatively, a level of 1 indicates no beauty, a level of 2 indicates light beauty, a level of 3 indicates medium beauty, and a level of 4 indicates deep beauty. Alternatively, the user instruction may be that the user slides the screen page on the smart terminal through a finger or a touch tool (such as a touch pen), or the user instruction may be that the user clicks the screen page on the smart terminal through the finger or the touch tool, or the user instruction may be that the user quickly double-clicks the screen page on the smart terminal through the finger or the touch tool, or the user instruction may be that the user performs a gesture operation on the screen page of the smart terminal through the finger or the touch tool. Optionally, the number of fingers performing the operation may be one or multiple, and the corresponding number of fingers may be selected according to a specific scene. The embodiment of the present application does not particularly limit the specific form of the user instruction, as long as the user instruction represents the determination of the beauty level. Because the form of user's instruction can be various, can improve the flexibility of operation mode like this, can promote user experience simultaneously.

Optionally, a progress bar including 1 to 4 beauty levels may be displayed on the screen page of the intelligent terminal, and then a sliding operation of the user in a preset direction is received, so as to determine the beauty level set by the user, and optionally, the sliding operation in the preset direction refers to a sliding direction preset by software, and the sliding operation is performed on the screen page of the intelligent terminal in the preset sliding direction by a finger or a touch tool. Optionally, the preset direction may be a direction from top to bottom, the preset direction may be a direction from bottom to top, the preset direction may be a first up-down direction and then a left-right direction, the preset direction may be a first left-right direction and then a top-down direction, the preset direction may be a clockwise upper semicircle, the preset direction may be a counterclockwise lower semicircle, the preset direction may be a clockwise arc, and the like. The embodiment of the application does not particularly limit the specific form of the preset direction, and can be adjusted correspondingly according to specific requirements.

Optionally, the user instruction may be any one of a click, a double click, a sliding operation in a preset direction, and an air gesture, and may also be a combination of any two or more of a click, a double click, a sliding operation in a preset direction, and an air gesture. When the user command is any one of clicking, double clicking, sliding operation in a preset direction and air separating gestures, the operation time can be saved. When the user command is the combination of any two or more of clicking, double-clicking, sliding operation in a preset direction and an air separating gesture, the safety of information can be improved. Alternatively, the instruction may be a combination of a tap and a blank gesture, a combination of a double tap and a sliding operation in a preset direction, a combination of a double tap and a blank gesture, a combination of a sliding operation in a preset direction and a blank gesture, and the like. Because the mode of user's instruction can be diversified, not only can improve the flexibility of user's operation, can also satisfy different users ' hobby and use custom, promote user experience.

The image processing method provided by the invention processes and modifies the initial image, and generally carries out color matching, matting, synthesis, shading modification, chroma and chroma modification, special effect addition, editing, restoration and the like on the image through image processing software. For example, the human figure is beautified, the picture is beautified, the beautifying treatment method includes but is not limited to whitening, skin grinding, face thinning, enlarging, lengthening, acne removing and the like, and the beautifying treatment method includes but is not limited to cutting, tilting, various filters and the like. Of course, those skilled in the art can set other image processing methods according to actual needs.

As can be seen from the above, in the embodiment of the application, at least one target portrait in an initial image may be detected, appearance feature information of the target portrait may be obtained, facial analysis may be performed on the target portrait to obtain expression information and/or makeup information of the target portrait, a beauty parameter of the target portrait may be determined according to at least one of the appearance feature information, the expression information, and the makeup information, and the initial image may be processed according to the beauty parameter to obtain the target image. The method and the device can automatically perform personalized beauty according to at least one of appearance characteristics, expressions and makeup of the user, so that the diversity and the use efficiency of image processing are improved.

An image processing method is further provided in the embodiment of the present application, please refer to fig. 4, where fig. 4 is a second flowchart of the image processing method provided in the embodiment of the present application, and the method includes:

and S21, detecting at least one portrait in the initial image, and respectively acquiring the portrait position and/or the face area of each portrait.

Optionally, at least one portrait in the initial image may be obtained through a face recognition technology, and considering that the passerby is often shot into the picture when the user shoots the picture, and the passerby does not need to be subjected to the face beautifying process, the embodiment may further remove the passerby portrait in the initial image, and specifically may identify and remove the passerby according to a portrait position and/or a face area of each portrait in the initial image.

S22, determining a road portrait in at least one portrait according to the position of the portrait and/or the area of the face, and removing the road portrait to obtain at least one target portrait.

After the portrait position and/or the face area of each portrait in the initial image are/is acquired, the face area can be smaller than the preset area, the portrait of which the portrait position is located at the edge position of the initial image is determined as a passerby, and the portrait of the passerby is removed, so that at least one target portrait is acquired. The preset area may be a preset proportion of the maximum face area, for example, 50% of the maximum face area, and the edge region may be a position where a distance from the image boundary is smaller than a preset value.

For example, referring to fig. 5, fig. 5 is a schematic view of a scene for determining a target portrait according to an embodiment of the present disclosure, in the view, an initial image includes three portraits, and it can be known through calculation that a face area of a third portrait is smaller than a maximum face area, that is, 50% of a face area of a first portrait, and a position of the first portrait is also located at an edge position of the initial image, so that the third portrait can be determined as a road portrait, and the remaining first portrait and the remaining second portrait are the target portraits.

And S23, acquiring gender information and/or age information and/or race information of the target portrait.

Optionally, the step of obtaining the gender information of the target portrait may include: and acquiring the characteristic point information of the target portrait, and determining the gender information of the target portrait according to the characteristic point information.

Optionally, for the gender identification, a classifier fusion mode combining a classifier trained on the local organ (for example, five sense organs) features on the face and a classifier of the whole face features can be adopted to construct a fusion classifier for the gender identification, so that the identification accuracy is improved; and the 2DPCA method is adopted to reduce the dimension of the image, and the 2DLDA is adopted as the classification method, so that the operation amount is reduced and the training and detection speed is increased on the basis of ensuring the detection precision. Optionally:

(1) preprocessing the image: and carrying out graying, histogram equalization and median filtering processing on the image in sequence.

Graying: a weighted average method is used. The image can be generally divided into a color image and a gray image, a pixel point of the color image is composed of three colors of R (red), G (green) and B (black), the gray image only contains brightness information and does not contain color information, for example, graying of the color image is usually performed by an empirical formula: gray 0.39 × R +0.5 × G +0.11 × B.

Histogram equalization: the gray level components are evenly distributed in the space, and the histogram shows that the dense gray level distribution is changed into the uniform distribution, so that the contrast of the image is enhanced, the interference of light rays to the image can be reduced, and the features are easy to extract.

Median filtering: it is essentially a filter of statistical order. For a certain point in the original image, the median filtering process takes the statistical ordering median of all pixels in the neighborhood with the point as the center as the response of the point. Median filtering has better noise reduction capability for certain types of random noise, and does not cause higher blurring effect while reducing noise compared with linear smoothing filtering.

(2) Detecting a face area: a classifier is designed by adopting a classifier AdaBoost (adaptive boosting) method based on a cascade classification model, face detection is carried out on one image, and faces existing in the image are extracted.

Most intelligent algorithms with higher operating efficiency are realized by extracting features which are efficient and are beneficial to recognition and processing, the Adaboost algorithm is the same, Adaboost is a learning model which is provided by Freund and Schapire on the basis of a PAC (Probaby experience Correct) model, and the algorithm idea is as follows: by learning a large number of positive samples and negative samples and by learning feedback, the weak classifier adaptively adjusts the error rate and corresponding weight on the premise of not knowing a priori training error until the strong classifier reaches a preset performance, the Adaboost algorithm is applied to face detection, and simultaneously the Haar characteristic, the Cascade algorithm and the Adaboost algorithm are combined, so that the detection speed and the detection accuracy are greatly improved.

(3) Intercepting a local organ area of the human face: through the previous steps, a face image is obtained from an image, the size of the face image is normalized, at least one face local organ image is extracted from the face image, and five sense organs of the left eye, the right eye, the eyebrow, the mouth and the nose can be extracted respectively. For example, the classifier for detecting the eye region in the face region may be obtained by training using an AdaBoost learning algorithm, and it is necessary to pay attention to the problem of selecting a sample, where a positive sample is an image of the eye region, and a training negative sample is composed of two parts, one part is a whole face image from which the eye region is removed, and the other part is a sub-window image around the eye region in the original image, and the two-eye image of the face may be captured in a similar manner.

(4) And (3) feature dimensionality reduction: a method using 2DPCA (Two-Dimensional Principal Component Analysis).

PCA (principal Component analysis) is a classic feature extraction and data dimension reduction tool in the field of pattern recognition and computer vision, while the 2DPCA method is a novel principal Component analysis method developed on the basis of PCA, and compared with the conventional PCA method, the 2DPCA is based on a two-dimensional image matrix rather than a one-dimensional image vector. This processing method does not need to convert the image into a one-dimensional vector in advance, and is equivalent to removing the correlation of the row vector or the column vector of the image. In 2DPCA, a covariance matrix is directly constructed by using a two-dimensional image matrix, the eigenvalue and eigenvector of the covariance matrix are solved, a coordinate system is constructed by using eigenvectors corresponding to the largest eigenvalue, and then each image matrix is projected on the coordinate system, so that the characteristics of the image are obtained, and the characteristics are less influenced by the number of samples. Compared with the covariance matrix constructed by PCA, the covariance matrix using 2DPCA is much smaller, and its main advantages are: directly calculating a writing difference matrix of the training sample; the time required to compute the eigenvalue eigenvectors is relatively small.

(5) And (4) mode classification: the embodiment of the present application may adopt a 2DLDA (Two-Dimensional Linear discriminant Analysis) method.

The 2DLDA is also a method directly based on a two-dimensional image matrix, and is used for respectively calculating the intra-class divergence matrix and the inter-class divergence matrix of the two-dimensional image and determining an optimal projection coordinate system under a certain optimal criterion. The 2DLDA characteristic of the face image is obtained by projecting the original image to a coordinate system, the 2DLDA method has small operation amount, and the spatial structure information of the face image is effectively utilized.

(6) Determining the size of the five sense organs contribution weight and fusing:

obtaining the accuracy rate Pi (i is 1,2, … 6) of each sub-classifier, and preliminarily determining the weight Qi of each sub-classifier to be Pi/sigma Pi; a simpler fusion method is addition and multiplication.

Optionally, the gender identification may be combined with other manners, for example, the target feature in the image may be identified and obtained, and may be one or more of a beard, an ornament, a hairstyle, a laryngeal knot, and an eyebrow, for example, when the target feature is a beard, the corresponding target feature information may be a beard (feature value of 1) or an eyebrow (feature value of 0), the gender corresponding to the beard is determined as a male, and when the target feature is an ornament, such as a hair clip, an ornament, a laryngeal knot, an eyebrow, and the like, the gender of the target figure is further determined according to other target features, and when the target feature is an ornament, such as a hair clip or an ear, the corresponding target feature information may be an ornament, the gender corresponding to the ornament is determined as a female, and when the ornament is not present, the gender of the target figure is further determined according to other target features; when the target feature is a laryngeal knot, the corresponding target feature information may be a laryngeal knot or no laryngeal knot, when the target feature information is a laryngeal knot, the target gender of the target portrait is determined to be male, and when the target feature information is a laryngeal knot, the target gender of the target portrait is determined to be female, which is not limited in the embodiment of the invention.

The step of acquiring age information of the target portrait may include: and acquiring skin surface layer characteristic information of the target portrait, and determining age information of the target portrait according to the skin surface layer characteristic information.

Alternatively, for the identification of the age, the clustering classification training can be performed according to the samples by using a regression equation or a multi-class classification algorithm with reference to a two-class classification algorithm of gender identification.

In the process of growing a person from a young to a young, the face skeleton grows continuously, so that the face shape becomes longer and bigger, the appearance changes, in the process of changing the young to an old, the skin color and wrinkles change, and therefore the age recognition rate is reduced rapidly along with the change of the age of the person. The mainstream age recognition method today is a method based on a face model and an age function, wherein the age function is obtained by training face images of a plurality of groups of people with known ages, and based on this method, the following four types can be used:

(1) constructing a global age function: a global age function is constructed through a large number of training images, and the age of the target face image is calculated through directly using the global age function.

(2) Appearance specific age function: this age function divides the input by facial features. Under this method, we consider that the aging process is very similar for individuals with similar facial features, by inputting a face image, looking for an age function with similar features, and then calculating the age.

(3) Weighted age function: the weighted age function is similar to the appearance-specific age function, but is an age function calculated by inputting an image, and finding an image with similar characteristics, and then the age of the face image is estimated by using all the age functions and then weighted, thereby obtaining an estimation result.

(4) Weighted personal special age function: the method has high input requirement, and not only needs the face image, but also needs factors such as gender, health, living standard, economic condition, stress level, working condition, marital condition, living place, exposed weather condition and the like, and the calculation process comprises a weighted age function and an appearance-specific age function and carries out weighted calculation on the influence of non-image information.

The step of obtaining the race information of the target portrait may include:

and matching in a preset human species information base according to the skin color and/or the eyeball color and/or the head outline and/or the face outline so as to determine the human species information of the target human figure.

And S24, acquiring a face image of the target portrait, and extracting facial morphological features and/or muscle morphological features from the face image.

And S25, determining the expression information of the target portrait according to the morphological characteristics of the five sense organs and/or the muscle morphological characteristics.

Optionally, after the facial image of the target portrait is acquired, the facial image may be identified and a feature point set in the facial image may be acquired, and then facial morphological features and/or muscle morphological features may be determined according to the feature point set.

Referring to fig. 6, fig. 6 is a diagram illustrating an exemplary feature point set in an image processing method according to an embodiment of the present invention. For example, first, 9 feature points of the human face may be selected, and the distribution of the feature points has an angular invariance, which is 2 eyeball center points, 4 eye corner points, a middle point of two nostrils, and 2 mouth corner points. Then, further selecting more feature points according to the 9 feature points, so as to obtain 16 feature points of the eye contour, 8 feature points of the nose contour, 16 feature points of the lips, 18 feature points of the face contour and the like, thereby obtaining a more complete feature point set of the face image.

In addition, the accuracy of feature point extraction in the lip region is greatly affected because the lip shape may be greatly changed due to the difference in facial expression, and the lip region is relatively easily interfered by factors such as beard and the like. Because the relative change of the positions of the mouth corner points is less influenced by expressions and the like, and the positions of the corner points are more accurate, the important characteristic points of the lip region are adopted as the positioning modes of the two mouth corner points.

With reference to fig. 7, fig. 7 is a diagram illustrating another example of a feature point set in an image processing method according to an embodiment of the invention. The area circled by the

vertexes

19, 20, 21, 22, 23, 24, 25 and 26 as paths in sequence and the area circled by the

vertexes

27, 28, 29, 30, 31, 32, 33 and 34 as paths in sequence can be determined as eyebrows through the collection of the feature points. The area circled with the

vertices

35, 36, 37, 38, 39, 40, 41, 42, 43 as paths in order and the area circled with the

vertices

44, 45, 46, 47, 48, 49, 50, 51, 52 as paths in order are eyes. The area encircled by the

vertices

54, 56, 58, 59, 60, 61, 62, 64 in this order is the nose. The area circled by taking the

vertexes

65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81 and 82 as paths in sequence is taken as a mouth.

After the morphological characteristics of the five sense organs are obtained, matching can be carried out in an expression database, and the preset expression with the highest similarity is determined as the expression information of the target portrait. Expression types may include, but are not limited to, serious, smiling, silly, sold, funny, and the like. Optionally, the expression type is used to determine a beauty parameter, the beauty parameter corresponds to the expression type information, and different beauty parameters correspond to different beauty styles.

And S26, extracting eyebrow state and/or eyelid color and/or cheek color and/or lip color and/or skin brightness difference value from the face image.

And S27, determining the makeup information of the target portrait according to the eyebrow state and/or the color of the eyelid and/or the color of the cheek and/or the color of the lip and/or the value of the skin brightness difference.

For example, the makeup information of the target portrait may be determined by makeup detection technology, and optionally, the eyebrow shape state may be judged: whether eyebrows exist or not and whether double eyebrows are symmetrical or not; the eyelid state: whether the upper eyelid area is non-skin tone color (with an eye shadow effect); cheek state: whether the two sides of the cheek are skin color (non-skin color, blush is present); lip state: whether the lip area belongs to a lipstick color range (if the lip area is located in a preset color interval, lipstick is considered to be in the lip area); whether the skin brightness difference value exceeds a threshold value (if the skin brightness difference value exceeds the threshold value, the skin brightness difference value is determined to be a corrected skin); so as to comprehensively judge the makeup information of the target portrait, and the makeup types can include but are not limited to no makeup, light makeup, heavy makeup and the like. Optionally, the makeup type is used to determine a beauty parameter, the beauty parameter corresponds to the makeup type information, and different beauty parameters correspond to different beauty styles.

And S28, determining the beauty parameters of the target portrait according to the gender information and/or the age information and/or the race information, the expression information and/or the makeup information.

Optionally, in the embodiment of the present application, the gender information and/or the age information and/or the race information, the expression information and/or the makeup information may be input into a preset fast RCNN network for classification training, so as to obtain the beauty prediction model, and the output result is a predicted value of a beauty parameter.

In the step, the Faster RCNN is an arithmetic mathematical network model which simulates animal neural network behavior characteristics and performs distributed parallel information processing, and has self-learning and self-adaptive capabilities. It is essentially an input-to-output mapping that is able to learn a large number of input-to-output mappings without requiring any precise expression between the input person and the output, and the network has the ability to map between inputs and outputs simply by training the network with known patterns.

The feedback network formed by the adjusted face difference characteristic data can be input into a fast RCNN for training in advance, the difference change condition of the face characteristic between the sample image and the characteristic image is learned, namely, how to learn the face beautifying processing method of a certain face characteristic is learned, and the face beautifying prediction model is obtained according to the learned face beautifying processing method.

And S29, processing the initial image according to the beauty parameters to obtain a target image.

Optionally, the step of processing the initial image according to the beauty parameters may include skin polishing according to parameters corresponding to skin color and skin type, color matching according to parameters corresponding to eyebrow color, hair color, lip color and pupil color, face thinning, ear trimming, eyebrow trimming and the like according to parameters corresponding to face shape, ear shape and eyebrow shape. The mode of processing the target portrait in the initial image can also comprise brightness optimization, definition improvement, denoising processing, obstacle processing and the like besides the beautifying so as to ensure that the original human face three-dimensional model is accurate.

As can be seen from the above, the embodiment of the present application may detect at least one portrait in an initial image, respectively obtain a portrait position and/or a face area of each portrait, determine a passerby portrait in the at least one portrait according to the portrait position and/or the face area, remove the passerby portrait to obtain at least one target portrait, obtain gender information and/or age information and/or race information of the target portrait, obtain a facial image of the target portrait, extract facial morphological features and/or muscle morphological features in the facial image, determine expression information of the target portrait according to the morphological features and/or muscle morphological features, extract an eyebrow state and/or an eyelid color and/or a cheek color and/or a lip color and/or a skin darkness difference value in the facial image, and extract an eyebrow state and/or an eyelid color and/or a cheek color and/or a lip color Or determining the makeup information of the target portrait according to the skin brightness difference value, determining the beauty parameters of the target portrait according to the gender information and/or the age information and/or the race information, the expression information and/or the makeup information, and processing the initial image according to the beauty parameters to obtain the target image. According to the method and the device, personalized beauty can be automatically performed according to the gender information and/or the age information and/or the race information, the expression information and/or the makeup information of the user, so that the diversity and the use efficiency of image processing are improved.

Referring to fig. 8, fig. 8 is a third flowchart illustrating an image processing method according to an embodiment of the present application. The flow of the image processing method may include:

and S31, detecting at least one target portrait in the initial image and acquiring the appearance characteristic information of the target portrait.

And S32, carrying out facial analysis on the target portrait to obtain the expression information and/or the makeup information of the target portrait.

And S33, determining the beauty parameters of the target portrait according to at least one of the appearance feature information, the expression information and the makeup information.

Optionally, the relevant contents of the above steps S31 to S33 may refer to the description of the steps S11 to S13, which is not further described in this embodiment.

And S34, segmenting the initial image to respectively obtain a portrait image and a scene image.

Alternatively, the embodiment may divide the picture of the portrait and the non-portrait area by using a picture division technology. As shown in fig. 9, fig. 9 is a scene schematic diagram of image segmentation in the image processing method according to the embodiment of the present application. Alternatively, the target portrait in the initial image may be determined through face recognition, and then the portrait image and the scene image may be segmented. When determining the target portrait, the portrait of the passerby in the scene may be removed, so that neither the segmented portrait image nor the scene image has the portrait of the passerby.

And S35, recognizing the scene image to obtain a scene type, recognizing the human image to obtain a light ray type, and matching the target filter in a preset filter library according to the scene type and the light ray type.

Optionally, the scene types may include indoor, park, party, building, and the like. The light type comprises angles of illumination light on the face, namely front light, back light and side light, and can be judged specifically through the brightness of the face area and the brightness of the background area, for example, if the brightness difference of the left side, the right side, the forehead, the chin and other areas of the face is smaller than a first brightness threshold, the brightness of each area is larger than a second brightness threshold, and the mean value of the brightness of the portrait area is higher than the mean value of the brightness of the background area, the portrait area is front light; and/or if the brightness difference of the left and right sides of the face, the forehead, the chin and other areas is smaller than a first brightness threshold, the brightness of each area is larger than a second brightness threshold, and the brightness mean value of the portrait area is lower than that of the background area, the portrait area is backlight; and/or, if the brightness difference of the left and right side areas of the face is larger than the third brightness threshold value, the face is sidelight.

After the scene type and the light type are obtained, a scene sample with the highest similarity to the scene type and the light type and a target filter corresponding to the sample can be matched in a preset filter library.

And S36, performing first processing on the scene image according to the target filter, performing second processing on the portrait image according to the beauty parameters, and synthesizing the images subjected to the first processing and the second processing to obtain a target image.

Optionally, after the filter is added to the scene image and the portrait image is processed according to the beauty parameters, the processed scene image and the portrait image can be synthesized to generate a final target image.

Optionally, the method may further include:

For example, when the memory occupancy of the background of the intelligent terminal exceeds or 60%, the CPU temperature exceeds 60 ° and/or the power consumption per unit time exceeds 200mA per minute, the number of the image-to-face skin-care fits can be reduced, for example, a preview interface can be used to remind the user that 2 people in the screen support the skin-care effect, or the skin-care strength is automatically reduced, for example, the skin-care effect is automatically reduced from deep skin-care to light skin-care, or the fit of only opening 3 facial-features is opened, and the like, so that the stability of power consumption is ensured, and the bad user experience of hot hair and fever is reduced.

From the above, the embodiment of the application can detect at least one target portrait in the initial image, acquire the appearance feature information of the target portrait, performing facial analysis on the target portrait to obtain expression information and/or makeup information of the target portrait, determining beauty parameters of the target portrait based on at least one of the appearance feature information, the expression information and the makeup information, segmenting the initial image to obtain a portrait image and a scene image, identifying the scene image to obtain a scene type, identifying the portrait image to obtain a light ray type, matching a target filter in a preset filter library according to the scene type and the light ray type, and carrying out first processing on the scene image according to the target filter, carrying out second processing on the portrait image according to the beauty parameter, and synthesizing the images subjected to the first processing and the second processing to obtain a target image. According to the method and the device, personalized beauty can be automatically performed according to at least one of appearance characteristics, expressions and makeup of the user, and the filter is added according to the scene, so that the diversity and the use efficiency of image processing are improved.

Fig. 10 is a schematic structural diagram of an image processing apparatus according to an embodiment of the present application. The image processing apparatus may be provided in an intelligent terminal. Referring to fig. 10, the image processing apparatus 30 includes:

the detection module 301 is configured to detect at least one target portrait in an initial image, and acquire appearance feature information of the target portrait;

an analysis module 302, configured to perform facial analysis on the target portrait to obtain expression information and/or makeup information of the target portrait;

a determining module 303, configured to determine a beauty parameter of the target portrait according to at least one of the appearance feature information, the expression information, and the makeup information;

and the processing module 304 is configured to process the initial image according to the beauty parameter to obtain a target image.

Optionally, please continue to refer to fig. 11, where fig. 11 is a schematic structural diagram of another image processing apparatus provided in the embodiment of the present application, and optionally, the detection module 301 may include:

a detection sub-module 3011, configured to detect at least one portrait in the initial image;

an obtaining sub-module 3012, configured to obtain a portrait position and/or a face area of each portrait respectively;

the determining submodule 3013 is configured to determine a road portrait from the at least one portrait according to the position of the portrait and/or the area of the face, and remove the road portrait to obtain the at least one target portrait.

Optionally, the determining module 303 may include:

an input sub-module 3031, configured to input at least one of the appearance feature information, the expression information, and the makeup information to a preset network model, where optionally, the preset network model is a neural network model trained by using a user history target image as a sample set;

and the predicting submodule 3032 is configured to obtain the predicted beauty parameters output by the preset network model.

The image processing device provided by the embodiment of the application can detect at least one target portrait in an initial image, acquire appearance feature information of the target portrait, perform face analysis on the target portrait to obtain expression information and/or makeup information of the target portrait, determine beauty parameters of the target portrait according to at least one of the appearance feature information, the expression information and the makeup information, and process the initial image according to the beauty parameters to obtain the target image. The method and the device can automatically perform personalized beauty according to at least one of appearance characteristics, expressions and makeup of the user, so that the diversity and the use efficiency of image processing are improved.

The embodiment of the present application further provides an intelligent terminal, where the intelligent terminal includes a memory and a processor, and the memory stores an image processing program, and the image processing program is executed by the processor to implement the steps of the image processing method in any of the above embodiments.

The embodiment of the present application further provides a computer storage medium, where a computer program is stored on the computer storage medium, and when the computer program is executed by a processor, the steps of the image processing method in any of the above embodiments are implemented.

In the embodiments of the intelligent terminal and the computer storage medium provided in the present application, all technical features of any one of the embodiments of the image processing method may be included, and the expanding and explaining contents of the specification are basically the same as those of the embodiments of the method, and are not described herein again.

Embodiments of the present application also provide a computer program product, which includes computer program code, when the computer program code runs on a computer, the computer is caused to execute the method in the above various possible embodiments.

Embodiments of the present application further provide a chip, which includes a memory and a processor, where the memory is used to store a computer program, and the processor is used to call and run the computer program from the memory, so that a device in which the chip is installed executes the method in the above various possible embodiments.

It is to be understood that the foregoing scenarios are only examples, and do not constitute a limitation on application scenarios of the technical solutions provided in the embodiments of the present application, and the technical solutions of the present application may also be applied to other scenarios. For example, as can be known by those skilled in the art, with the evolution of system architecture and the emergence of new service scenarios, the technical solution provided in the embodiments of the present application is also applicable to similar technical problems.

The above-mentioned serial numbers of the embodiments of the present application are merely for description and do not represent the merits of the embodiments.

The steps in the method of the embodiment of the application can be sequentially adjusted, combined and deleted according to actual needs.

The modules or units in the device in the embodiment of the application can be combined, divided and deleted according to actual needs.

In the present application, the same or similar term concepts, technical solutions and/or application scenario descriptions will be generally described only in detail at the first occurrence, and when the description is repeated later, the detailed description will not be repeated in general for brevity, and when understanding the technical solutions and the like of the present application, reference may be made to the related detailed description before the description for the same or similar term concepts, technical solutions and/or application scenario descriptions and the like which are not described in detail later.

In the present application, each embodiment is described with emphasis, and reference may be made to the description of other embodiments for parts that are not described or illustrated in any embodiment.

The technical features of the technical solution of the present application may be arbitrarily combined, and for brevity of description, all possible combinations of the technical features in the embodiments are not described, however, as long as there is no contradiction between the combinations of the technical features, the scope of the present application should be considered as being described in the present application.

Through the above description of the embodiments, those skilled in the art will clearly understand that the method of the above embodiments can be implemented by software plus a necessary general hardware platform, and certainly can also be implemented by hardware, but in many cases, the former is a better implementation manner. Based on such understanding, the technical solutions of the present application may be embodied in the form of a software product, which is stored in a storage medium (e.g., ROM/RAM, magnetic disk, optical disk) and includes instructions for enabling a terminal device (e.g., a mobile phone, a computer, a server, a controlled terminal, or a network device) to execute the method of each embodiment of the present application.

In the above embodiments, the implementation may be wholly or partially realized by software, hardware, firmware, or any combination thereof. When implemented in software, may be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. The procedures or functions according to the embodiments of the present application are all or partially generated when the computer program instructions are loaded and executed on a computer. The computer may be a general purpose computer, a special purpose computer, a network of computers, or other programmable device. The computer instructions may be stored on a computer readable storage medium or transmitted from one computer readable storage medium to another, for example, the computer instructions may be transmitted from one website, computer, server, or data center to another website, computer, server, or data center by wire (e.g., coaxial cable, fiber optic, digital subscriber line) or wirelessly (e.g., infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device, such as a server, a data center, etc., that incorporates one or more of the available media. The usable medium may be a magnetic medium (e.g., floppy Disk, memory Disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., Solid State Disk (SSD)), among others.

The above description is only a preferred embodiment of the present application, and not intended to limit the scope of the present application, and all modifications of equivalent structures and equivalent processes, which are made by the contents of the specification and the drawings of the present application, or which are directly or indirectly applied to other related technical fields, are included in the scope of the present application.

Claims

1. An image processing method, characterized by comprising the steps of:

s10: acquiring at least one target portrait in the initial image;

2. The method of claim 1, wherein the step of S10 includes:

detecting at least one portrait in the initial image;

3. The method according to claim 1, wherein the key information of the target portrait includes gender information and/or age information, and the step of S20 includes:

4. The method according to claim 1, wherein the key information of the target portrait includes race information, and the step S20 includes:

5. The method according to any one of claims 1 to 4, wherein the step S20 includes:

6. The method according to any one of claims 1 to 4, wherein the step S20 includes:

7. The method according to any one of claims 1 to 4, wherein the step S30 further comprises:

inputting the key information into a preset network model;

8. The method according to any one of claims 1 to 4, wherein the step S30 includes:

9. The method according to any one of claims 1 to 4, further comprising:

10. An intelligent terminal, characterized in that, intelligent terminal includes: memory, a processor, wherein the memory has stored thereon an image processing program which, when executed by the processor, implements the steps of the image processing method of any of claims 1 to 9.

11. A computer-readable storage medium, characterized in that the storage medium has stored thereon a computer program which, when being executed by a processor, carries out the steps of the image processing method according to any one of claims 1 to 9.