WO2019033570A1

WO2019033570A1 - Lip movement analysis method, apparatus and storage medium

Info

Publication number: WO2019033570A1
Application number: PCT/CN2017/108749
Authority: WO
Inventors: 陈林; 张国辉
Original assignee: 平安科技（深圳）有限公司
Priority date: 2017-08-17
Filing date: 2017-10-31
Publication date: 2019-02-21
Also published as: CN107633205B; CN107633205A

Abstract

A lip movement analysis method, an apparatus and a storage medium. The method comprises: obtaining a real-time image captured by an imaging apparatus, and extracting a real-time facial image from the real-time image (S10); inputting the real-time facial image to a pre-trained lip average model, and identifying t lip feature points representing lip positions in the real-time facial image (S20); determining a lip area according to the t lip feature points, inputting the lip area to a pre-trained lip classification model, and determining whether the lip area is a lip area of a person (S30); if yes, calculating a lip movement direction and movement distance in the real-time facial image according to x and y coordinates of the t lip feature points in the real-time facial image (S40). Lip movement information in the real-time facial image is calculated according to the coordinates of the lip feature points, so as to implement lip area analysis and real-time lip movement capturing.

Description

Lip movement analysis method, device and storage medium

Priority claim

This application is based on the priority of the Chinese Patent Application entitled "Palm Action Analysis Method, Apparatus and Storage Medium" filed on August 17, 2017, with the application number of CN 201710708364.9, which is filed on August 17, 2017, the entire contents of which are The manner of reference is incorporated in the present application.

Technical field

The present application relates to the field of computer vision processing technologies, and in particular, to a lip motion analysis method, apparatus, and computer readable storage medium.

Background technique

Lip motion capture is a biometric recognition technique that performs user lip motion recognition based on human facial feature information. At present, the application of lip motion capture is very extensive, and plays a very important role in many fields such as access control attendance and identity recognition, which brings great convenience to people's lives. The capture of lip movements, the general product approach is to use the deep learning method to train the classification model of the lip features through deep learning, and then use the classification model to judge the characteristics of the lips.

However, using deep learning methods to train lip features, the number of lip features depends entirely on the type of lip sample, such as judging mouth opening, closing mouth, then at least need to take a mouth, shut a large number of samples, if you want to judge the mouth, then It takes a lot of samples to pout and then retrain. This is not only time consuming, but also impossible to capture in real time. In addition, the lip feature is judged based on the classification model of the lip feature, and it is not possible to analyze whether or not the recognized lip region is a human lip region.

Summary of the invention

The present application provides a lip motion analysis method, device and computer readable storage medium, the main purpose of which is to calculate the motion information of the lips in the real-time facial image according to the coordinates of the lip feature points, and realize the analysis of the lip region and the action on the lips. Capture in real time.

In order to achieve the above object, the present application provides an electronic device, including: a memory, a processor, and an imaging device, wherein the memory includes a lip motion analysis program, and the lip motion analysis program is executed by the processor to implement the following step:

a real-time facial image acquisition step: acquiring a real-time image captured by the camera device, and extracting a real-time facial image from the real-time image by using a face recognition algorithm;

a feature point recognition step: inputting the real-time facial image into a pre-trained lip average model, and using the lip average model to identify t lip feature points representing the position of the lips in the real-time facial image;

a lip region recognizing step: determining a lip region according to the t lip feature points, inputting the lip region into a pre-trained lip classification model, and determining whether the lip region is a human lip region;

Lip movement judging step: if the lip area is a human lip area, according to the real-time face The x and y coordinates of t lip feature points in the image are calculated, and the moving direction and moving distance of the lips in the real-time facial image are calculated.

Preferably, when the lip motion analysis program is executed by the processor, the following steps are further implemented:

Prompting step: When the lip classification model judges that the lip region is not the human lip region, the prompt does not detect the human lip region from the current live image, cannot determine the lip motion, and returns to the real-time facial image acquisition step.

Preferably, the lip movement determining step comprises:

Calculating the distance between the central feature point of the inner side of the upper lip and the central feature point of the inner side of the lower lip in the real-time facial image, and determining the degree of opening of the lip;

The characteristic points of the left outer lip corner feature points and the feature points closest to the left outer lip corner feature points on the outer contour lines of the upper and lower lips are respectively connected to form a vector

Calculation vector

The angle between the left side of the lips; and

The feature points of the right outer lip corner point and the feature points closest to the right outer lip corner feature point on the outer contour line of the upper and lower lips are respectively connected to form a vector

Calculation vector

The angle between the right side of the lips is obtained.

In addition, in order to achieve the above object, the present application further provides a lip motion analysis method, the method comprising:

Lip motion judging step: If the lip region is a human lip region, the moving direction and the moving distance of the lip in the real-time facial image are calculated according to the x and y coordinates of the t lip feature points in the real-time facial image.

Preferably, the lip movement determining step comprises:

Calculation vector

The angle between the left corners of the lips; and

Calculation vector

The angle between the right side of the lips is obtained.

In addition, in order to achieve the above object, the present application further provides a computer readable storage medium including a lip motion analysis program, when the lip motion analysis program is executed by a processor, implementing the above Any step in the lip motion analysis method.

The lip motion analysis method, apparatus and computer readable storage medium provided by the present application determine whether a region composed of lip feature points is a human lip region by recognizing a lip feature point from a real-time facial image, and if so, according to the lip feature point The coordinate calculation calculates the motion information of the lips, and the deep analysis of the samples of the various movements of the lips is not required, so that the analysis of the lip region and the real-time capture of the lip motion can be realized.

DRAWINGS

1 is a schematic diagram of a preferred embodiment of an electronic device of the present application;

2 is a block diagram of the lip motion analysis program of FIG. 1;

3 is a flow chart of a preferred embodiment of a lip motion analysis method of the present application;

FIG. 4 is a schematic diagram showing the refinement of the step S40 of the lip motion analysis method of the present application.

The implementation, functional features and advantages of the present application will be further described with reference to the accompanying drawings.

Detailed ways

It is understood that the specific embodiments described herein are merely illustrative of the application and are not intended to be limiting.

The application provides an electronic device 1 . Referring to FIG. 1 , it is a schematic diagram of a preferred embodiment of the electronic device 1 of the present application.

In this embodiment, the electronic device 1 may be a terminal device having a computing function, such as a server, a smart phone, a tablet computer, a portable computer, or a desktop computer.

The electronic device 1 includes a processor 12, a memory 11, an imaging device 13, a network interface 14, and a communication bus 15. The camera device 13 is installed in a specific place, such as an office place and a monitoring area, and real-time images are taken in real time for the target entering the specific place, and the captured real-time image is transmitted to the processor 12 through the network. Network interface 14 may optionally include a standard wired interface, a wireless interface (such as a WI-FI interface). Communication bus 15 is used to implement connection communication between these components.

The memory 11 includes at least one type of readable storage medium. The at least one type of readable storage medium may be a non-volatile storage medium such as a flash memory, a hard disk, a multimedia card, a card type memory, or the like. In some embodiments, the readable storage medium may be an internal storage unit of the electronic device 1, For example, the hard disk of the electronic device 1. In other embodiments, the readable storage medium may also be an external memory of the electronic device 1, such as a plug-in hard disk equipped on the electronic device 1, a smart memory card (SMC), Secure Digital (SD) card, Flash Card, etc.

In this embodiment, the readable storage medium of the memory 11 is generally used for storing a lip motion analysis program 10 installed on the electronic device 1, a face image sample library, a human lip sample library, and being constructed and trained. Lip average model and lip classification model. The memory 11 can also be used to temporarily store data that has been output or is about to be output.

The processor 12, in some embodiments, may be a Central Processing Unit (CPU), microprocessor or other data processing chip for running program code or processing data stored in the memory 11, such as performing lip motion analysis. Program 10 and so on.

1 shows only the electronic device 1 having the components 11-15 and the lip motion analysis program 10, but it should be understood that not all illustrated components may be implemented, and more or fewer components may be implemented instead.

Optionally, the electronic device 1 may further include a user interface, and the user interface may include an input unit such as a keyboard, a voice input device such as a microphone, a device with a voice recognition function, a voice output device such as an audio, a headphone, and the like. Optionally, the user interface may also include a standard wired interface and a wireless interface.

Optionally, the electronic device 1 may further include a display, which may also be appropriately referred to as a display screen or a display unit. In some embodiments, it may be an LED display, a liquid crystal display, a touch liquid crystal display, and an Organic Light-Emitting Diode (OLED) touch sensor. The display is used to display information processed in the electronic device 1 and a user interface for displaying visualizations.

Optionally, the electronic device 1 further comprises a touch sensor. The area provided by the touch sensor for the user to perform a touch operation is referred to as a touch area. Further, the touch sensor described herein may be a resistive touch sensor, a capacitive touch sensor, or the like. Moreover, the touch sensor includes not only a contact type touch sensor but also a proximity type touch sensor or the like. Furthermore, the touch sensor may be a single sensor or a plurality of sensors arranged, for example, in an array.

In addition, the area of the display of the electronic device 1 may be the same as or different from the area of the touch sensor. Optionally, a display is stacked with the touch sensor to form a touch display. The device detects a user-triggered touch operation based on a touch screen display.

Optionally, the electronic device 1 may further include a radio frequency (RF) circuit, a sensor, an audio circuit, and the like, and details are not described herein.

In the apparatus embodiment shown in FIG. 1, an operating system and a lip motion analysis program 10 may be included in the memory 11 as a computer storage medium; when the processor 12 executes the lip motion analysis program 10 stored in the memory 11, the following is realized as follows step:

The real-time facial image acquisition step: acquiring a real-time image captured by the camera device, and extracting a real-time facial image from the real-time image by using a face recognition algorithm.

When the camera 13 captures a real-time image, the camera 13 transmits the real-time image to the processor 12. When the processor 12 receives the real-time image, it first acquires the size of the image to create a grayscale image of the same size. Converting the acquired color image into a grayscale image and creating a memory space; equalizing the grayscale image histogram, reducing the amount of grayscale image information, speeding up the detection speed, and then loading the training library to detect the person in the image Face, and return an object containing face information, obtain the data of the location of the face, and record the number; finally obtain the area of the avatar and save it, thus completing a real-time facial image extraction process.

Specifically, the face recognition algorithm for extracting the real-time facial image from the real-time image may be a geometric feature-based method, a local feature analysis method, a feature face method, an elastic model-based method, a neural network method, or the like.

Feature point recognition step: input the real-time facial image into a pre-trained lip average model, and use the lip average model to identify t lip feature points representing the position of the lips in the real-time facial image.

Establishing a first sample library with n face images, and manually marking t feature points in the lip portion of each face image in the first sample library, the t feature points are evenly distributed on the Lower lip and left and right lip corners.

The face feature recognition model is trained using the face image marking the lip feature point to obtain a lip average model for the face. The face feature recognition model is an Ensemble of Regression Tress (ERT) algorithm. The ERT algorithm is expressed as follows:

Where t represents the cascading sequence number and τ _t (·, ·) represents the regression of the current stage. Each regression is composed of a number of regression trees, and the purpose of training is to obtain these regression trees.

Where S(t) is the shape estimate of the current model; each regression τ _t (·, ·) predicts an increment based on the input images I and S(t)

Add this increment to the current shape estimate to improve the current model. Each level of regression is based on feature points for prediction. The training data set is: (I1, S1), ..., (In, Sn) where I is the input sample image and S is the shape feature vector composed of the feature points in the sample image.

In the process of model training, the number of face images in the sample library is n, assuming t=20, that is, each sample picture has 20 feature points, and some feature points of all sample pictures are taken (for example, in each sample picture) The 15 feature points are randomly selected from the 20 feature points. The first regression tree is trained, and the predicted value of the first regression tree and the true value of the partial feature points (15 features taken from each sample image) The residual of the weighted average of the points is used to train the second tree... and so on, until the predicted value of the Nth tree is trained and the true value of the part of the feature points is close to 0, and all the regression trees of the ERT algorithm are obtained. According to these regression trees, the lip average model of the face is obtained, and the model file and the sample library are saved in the memory 11. Since the sample image of the training model marks 20 lip feature points, the trained lip average model of the face can be used to identify 20 lip feature points from the face image.

After the trained lip average model is called from the memory 11, the real-time facial image is aligned with the lip average model, and then the feature extraction algorithm is used to search the real-time facial image to match the 20 lip feature points of the lip average model. 20 lip feature points. It is assumed that the 20 lip feature points recognized from the real-time facial image are still recorded as P1 to P20, and the coordinates of the 20 lip feature points are: (x ₁ , y ₁ ), (x ₂ , y ₂ ) , (x ₃ , y ₃ ), ..., (x ₂₀ , y ₂₀ ).

Among them, as shown in Fig. 2, the upper and lower lips of the lip have eight feature points (respectively labeled as P1 to P8, P9 to P16), and the left and right lip angles respectively have two feature points (respectively labeled as P17 to P18). , P19 ~ P20). Of the 8 feature points of the upper lip, 5 are located on the outer contour line of the upper lip (P1 to P5), 3 are located on the contour line of the upper lip (P6 to P8, and P7 is the central feature point on the inner side of the upper lip); 8 of the lower lip Of the feature points, 5 are located on the outer contour line of the lower lip (P9 to P13), and 3 are located in the outline of the lower lip (P14 to P16, and P15 is the central feature point on the inner side of the lower lip). Of the two feature points of the left and right lip angles, one is located outside the lip contour (for example, P18, P20, hereinafter referred to as the outer lip feature point), and one is located outside the lip contour (for example, P17, P19, hereinafter referred to as Inner lip corner feature point). In this embodiment, the feature extraction algorithm is a SIFT (scale-invariant feature transform) algorithm. The SIFT algorithm extracts the local features of each lip feature point from the lip model of the face, selects a lip feature point as the reference feature point, and finds the same or similar feature in the real-time face image as the local feature of the reference feature point. The point (for example, the difference of the local features of the two feature points is within a preset range), according to this principle until all the lip feature points are found in the real-time face image. In other embodiments, the feature extraction algorithm may also be a SURF (Speeded Up Robust Features) algorithm, an LBP (Local Binary Patterns) algorithm, a HOG (Histogram of Oriented Gridients) algorithm, or the like.

Lip region recognizing step: determining a lip region based on the t lip feature points, inputting the lip region into a pre-trained lip classification model, and determining whether the lip region is a human lip region.

The m lip positive sample image and the k lip negative sample image are collected to form a second sample library. The lip positive sample image refers to an image containing human lips, and the lip portion can be extracted from the face image sample library as a positive lip sample image. A negative sample image of a lip refers to an image of a person's lip region being defective, or a lip in the image is not a human (eg, animal) lip, and a plurality of lips positive sample images and negative sample images form a second sample bank.

The local features of the positive sample image of each lip and the negative sample image of the lips are extracted. The feature extraction algorithm is used to extract the Histogram of Oriented Gradient (HOG) feature of the lip sample image. Since the color information in the lip sample image is not very effective, it is usually converted into a grayscale image, and the entire image is normalized, the gradient of the horizontal and vertical directions of the image is calculated, and the gradient of each pixel position is calculated accordingly. Direction values to capture outlines, silhouettes, and some texture information, and further weaken the effects of lighting. Then the whole image is divided into individual Cell cells, and a gradient direction histogram is constructed for each Cell cell to calculate the local image gradient information and quantize to obtain the feature description vector of the local image region. Then the Cell cells are combined into a large block. Due to the change of local illumination and the change of the foreground-background contrast, the gradient intensity varies greatly. This requires normalization of the gradient intensity, further Light, shadow, and edges are compressed. Finally, all "block" HOG descriptors are combined to form the final HOG feature description vector.

Using the positive sample image of the lips, the negative sample image of the lips, and the extracted HOG feature pair support vector machine The (Support Vector Machine, SVM) classifier performs training to obtain a lip classification model of the face.

After identifying 20 lip feature points from the real-time facial image, a lip region can be determined according to the 20 lip feature points, and then the determined lip region is input into the trained lip classification model, and the result is determined according to the result of the model. Whether the determined lip area is the human lip area.

Specifically, the lip motion determining step includes:

Calculation vector

The angle between the left side of the lips; and

Calculation vector

The angle between the right side of the lips is obtained.

In the real-time facial image, the coordinates of the central feature point P7 on the inner side of the upper lip are (x ₇ , y ₇ ), the coordinates of the central feature point P15 on the inner side of the lower lip are (x ₁₅ , y ₁₅ ), and the lip region is human The lip area, then, the distance between the two points is as follows:

If d=0, it means that P7 and P15 are coincident, that is, the lips are in a closed state; if d>0, the degree of opening of the lips is judged according to the size of d, and the larger d is, the degree of opening of the lips. The bigger.

The coordinates of the left outer lip corner feature point P18 are (x ₁₈ , y ₁₈ ), and the coordinates of the feature points P1 and P9 closest to P18 on the outer contour lines of the upper and lower lips are (x ₁ , y ₁ ), respectively. x ₉ , y ₉ ), connect P18 with P1 and P9 to form vectors respectively

Calculation vector

The angle α between the calculation formula is as follows:

among them,

α represents a vector

The angle between the angles can be calculated by calculating the angle of the angle. The smaller the angle, the greater the degree of left ankle.

Similarly, the coordinates of the right outer lip corner feature point P20 are (x ₂₀ , y ₂₀ ), and the coordinates of the feature points P5 and P13 closest to P20 on the outer contour lines of the upper and lower lips are (x ₅ , y ₅ , respectively). ), (x ₁₃ , y ₁₃ ), connect P20 with P5 and P13 to form vectors

Calculation vector

The angle between the calculations is as follows:

among them,

表示 indicates a vector

The angle between the angles of the lips can be judged by calculating the angle of the angle; the smaller the angle, the greater the degree of right-handedness of the lips.

Prompting step: When the lip classification model determines that the lip region is not the human lip region, the prompt does not detect the human lip region from the current real-time image, and the lip motion cannot be determined, and the flow returns to the real-time image capturing step to capture the next real-time. image. After inputting the lip region determined by the 20 lip feature points into the lip classification model, it is determined according to the model result that the lip region is not the human lip region, and the lip region of the person is not recognized, and the next lip motion judging step cannot be performed. The real-time image taken by the camera 13 is reacquired and the subsequent steps are performed.

The electronic device 1 of the present embodiment extracts a real-time facial image from a real-time image, recognizes a lip feature point in the real-time facial image by using a lip average model, and analyzes a lip region determined by a lip feature point using a lip classification model. If the lip region is a human lip region, the motion information of the lip in the real-time facial image is calculated according to the coordinates of the lip feature point, and the analysis of the lip region and the real-time capture of the lip motion are realized.

In other embodiments, the lip motion analysis program 10 can also be partitioned into one or more modules, one or more modules being stored in the memory 11 and executed by the processor 12 to complete the application. A module as referred to in this application refers to a series of computer program instructions that are capable of performing a particular function. Referring to FIG. 2, it is a block diagram of the lip motion analysis program 10 of FIG. In this embodiment, the lip motion analysis program 10 can be divided into: an acquisition module 110, an identification module 120, a determination module 130, a calculation module 140, and a prompt module 150. The functions or operational steps implemented by the modules 110-150 are similar to the above, and are not described in detail herein, by way of example, for example:

The acquiring module 110 is configured to acquire a real-time image captured by the camera device 13 and extract a real-time face image from the real-time image by using a face recognition algorithm;

The recognition module 120 is configured to input the real-time facial image into a pre-trained lip average model, and use the lip average model to identify t lip feature points representing the lip position in the real-time facial image;

The determining module 130 is configured to determine a lip region according to the t lip feature points, input the lip region into a pre-trained lip classification model, and determine whether the lip region is a human lip region;

The calculating module 140 is configured to: when the lip region is a human lip region, calculate a moving direction and a moving distance of the lip in the real-time facial image according to the x and y coordinates of the t-lip feature points in the real-time facial image; and

The prompting module 150 is configured to: when the lip classification model determines that the lip region is not a human lip region, prompting that the human lip region is not detected from the current real-time image, and the lip motion cannot be determined, the flow returns to the real-time image capturing step, and the capturing is performed. A live image.

In addition, the present application also provides a lip motion analysis method. Referring to FIG. 3, a flow chart of a preferred embodiment of the lip motion analysis method of the present application is shown. The method can be performed by a device that can be implemented by software and/or hardware.

In the present embodiment, the lip motion analysis method includes steps S10 to S50.

Step S10: Acquire a real-time image captured by the camera device, and extract a real-time face image from the real-time image by using a face recognition algorithm.

When the camera captures a real-time image, the camera transmits the real-time image to the processor. When the processor receives the real-time image, first acquires the size of the image, and creates a grayscale image of the same size; Color image, converted into gray image, and create a memory space; equalize the gray image histogram, reduce the amount of gray image information, speed up the detection, then load the training library, detect the face in the picture, and return An object containing face information, obtains the data of the location of the face, and records the number; finally obtains the area of the avatar and saves it, thus completing a real-time facial image extraction process. Specifically, the face recognition algorithm for extracting the real-time facial image from the real-time image may also be a geometric feature-based method, a local feature analysis method, a feature face method, an elastic model-based method, a neural network method, or the like.

In step S20, the real-time facial image is input into the pre-trained lip average model, and the lip average model is used to identify t lip feature points representing the lip position in the real-time facial image.

The face feature recognition model is trained using the face image marking the lip feature point to obtain a lip average model for the face. The face feature recognition model is an ERT algorithm. The ERT algorithm is expressed as follows:

In the process of model training, the number of face images in the sample library is n, assuming t=20, that is, each sample picture has 20 feature points, and some feature points of all sample pictures are taken (for example, in each sample picture) The 15 feature points are randomly selected from the 20 feature points. The first regression tree is trained, and the predicted value of the first regression tree and the true value of the partial feature points (15 features taken from each sample image) The residual of the weighted average of the points is used to train the second tree... and so on, until the predicted value of the Nth tree is trained and the true value of the part of the feature points is close to 0, and all the regression trees of the ERT algorithm are obtained. According to these regression trees, the average model of the face's lips is obtained, and the model file and the sample library are saved. To the memory. Since the sample image of the training model marks 20 lip feature points, the trained lip average model of the face can be used to identify 20 lip feature points from the face image.

After calling the trained lip average model from the memory, the real-time facial image is aligned with the lip average model, and then the feature extraction algorithm is used to search the real-time facial image for matching the 20 lip feature points of the lip average model. 20 lip feature points. It is assumed that the 20 lip feature points recognized from the real-time facial image are still recorded as P1 to P20, and the coordinates of the 20 lip feature points are: (x ₁ , y ₁ ), (x ₂ , y ₂ ) , (x ₃ , y ₃ ), ..., (x ₂₀ , y ₂₀ ).

Among them, as shown in Fig. 2, the upper and lower lips of the lip have eight feature points (respectively labeled as P1 to P8, P9 to P16), and the left and right lip angles respectively have two feature points (respectively labeled as P17 to P18). , P19 ~ P20). Of the 8 feature points of the upper lip, 5 are located on the outer contour line of the upper lip (P1 to P5), 3 are located on the contour line of the upper lip (P6 to P8, and P7 is the central feature point on the inner side of the upper lip); 8 of the lower lip Of the feature points, 5 are located on the outer contour line of the lower lip (P9 to P13), and 3 are located in the outline of the lower lip (P14 to P16, and P15 is the central feature point on the inner side of the lower lip). Of the two feature points of the left and right lip angles, one is located outside the lip contour (for example, P18, P20, hereinafter referred to as the outer lip feature point), and one is located outside the lip contour (for example, P17, P19, hereinafter referred to as Inner lip corner feature point).

Specifically, the feature extraction algorithm may also be a SIFT algorithm, a SURF algorithm, an LBP algorithm, an HOG algorithm, or the like.

Step S30, determining a lip region according to the t lip feature points, inputting the lip region into a pre-trained lip classification model, and determining whether the lip region is a human lip region.

The local features of the positive sample image of each lip and the negative sample image of the lips are extracted. A feature extraction algorithm is used to extract a direction gradient histogram (HOG) feature of the lip sample image. Since the color information in the lip sample image is not very effective, it is usually converted into a grayscale image, and the entire image is normalized, the gradient of the horizontal and vertical directions of the image is calculated, and the gradient of each pixel position is calculated accordingly. Direction values to capture outlines, silhouettes, and some texture information, and further weaken the effects of lighting. Then the whole image is divided into individual Cell cells, and a gradient direction histogram is constructed for each Cell cell to calculate the local image gradient information and quantize to obtain the feature description vector of the local image region. Then the Cell cells are combined into a large block. Due to the change of local illumination and the change of the foreground-background contrast, the gradient intensity varies greatly. This requires normalization of the gradient intensity, further Light, shadow, and edges are compressed. Finally, all "block" HOG descriptors are combined to form the final HOG feature description vector.

The support vector machine classifier is trained by using the positive sample image of the lips, the negative sample image of the lips, and the extracted HOG feature to obtain a lip classification model of the face.

After recognizing 20 lip feature points from the real-time facial image, the 20 lips can be based on The feature point determines a lip region, and then inputs the determined lip region into the trained lip classification model, and judges whether the determined lip region is a human lip region based on the result obtained by the model.

Step S40, if the lip region is a human lip region, the moving direction and the moving distance of the lip in the real-time facial image are calculated according to the x and y coordinates of the t lip feature points in the real-time facial image.

Referring to FIG. 4, it is a schematic flowchart of the refinement of step S40 in the lip motion analysis method of the present application. Specifically, step S40 includes:

Step S41, calculating a distance between a central feature point of the inner side of the upper lip and a central feature point of the inner side of the lower lip in the real-time facial image, and determining the degree of opening of the lip;

Step S42, connecting the left outer lip feature point and the feature points closest to the left outer lip feature point on the outer contour of the upper and lower lips respectively to form a vector

Calculation vector

The angle between the left side of the lips; and

Step S43, connecting the feature points of the right outer lip corner with the feature points closest to the feature points of the right outer lip corner on the outer contour lines of the upper and lower lips respectively to form a vector

Calculation vector

The angle between the right side of the lips is obtained.

Calculation vector

The angle α between the calculation formula is as follows:

among them,

α represents a vector

Calculation vector

The angle between the calculations is as follows:

among them,

表示 indicates a vector

Step S50, when the lip classification model determines that the lip region is not the human lip region, the prompt does not detect the human lip region from the current real-time image, and the lip motion cannot be determined, and the flow returns to the real-time image capturing step to capture the next real-time. image. After inputting the lip region determined by the 20 lip feature points into the lip classification model, it is determined according to the model result that the lip region is not the human lip region, and the lip region of the person is not recognized, and the next lip motion judging step cannot be performed. Re-acquire the live image captured by the camera and follow the next steps.

The lip motion analysis method of the present embodiment uses the lip average model to identify the lip feature points in the real-time facial image, and uses the lip classification model to analyze the lip region determined by the lip feature point, if the lip region is a human lip region Then, according to the coordinates of the lip feature points, the motion information of the lips in the real-time facial image is calculated, and the analysis of the lip region and the real-time capture of the lip motion are realized.

In addition, the embodiment of the present application further provides a computer readable storage medium, where the computer readable storage medium includes a lip motion analysis program, and when the lip motion analysis program is executed by the processor, the following operations are implemented:

Model construction steps: construct and train a facial feature recognition model, obtain a lip average model on the face, and use the lip sample image to train the SVM to obtain a lip classification model;

Optionally, when the lip motion analysis program is executed by the processor, the following operations are also implemented:

Optionally, the lip motion determining step includes:

Calculation vector

The angle between the left side of the lips; and

Calculation vector

The angle between the right side of the lips is obtained.

The specific implementation manner of the computer readable storage medium of the present application is substantially the same as the specific embodiment of the lip motion analysis method described above, and details are not described herein again.

It is to be understood that the term "comprises", "comprising", or any other variants thereof, is intended to encompass a non-exclusive inclusion, such that a process, apparatus, article, or method that comprises a series of elements includes those elements. It also includes other elements not explicitly listed, or elements that are inherent to such a process, device, item, or method. An element that is defined by the phrase "comprising a ..." does not exclude the presence of additional equivalent elements in the process, the device, the item, or the method that comprises the element.

The serial numbers of the embodiments of the present application are merely for the description, and do not represent the advantages and disadvantages of the embodiments. Through the description of the above embodiments, those skilled in the art can clearly understand that the foregoing embodiment method can be implemented by means of software plus a necessary general hardware platform, and of course, can also be through hardware, but in many cases, the former is better. Implementation. Based on such understanding, the technical solution of the present application, which is essential or contributes to the prior art, may be embodied in the form of a software product stored in a storage medium (such as ROM/RAM as described above). , a disk, an optical disk, including a number of instructions for causing a terminal device (which may be a mobile phone, a computer, a server, or a network device, etc.) to perform the methods described in the various embodiments of the present application.

The above is only a preferred embodiment of the present application, and is not intended to limit the scope of the patent application, and the equivalent structure or equivalent process transformations made by the specification and the drawings of the present application, or directly or indirectly applied to other related technical fields. The same is included in the scope of patent protection of this application.

Claims

An electronic device, comprising: a memory, a processor, and an imaging device, wherein the memory includes a lip motion analysis program, and when the lip motion analysis program is executed by the processor, the following steps are implemented:

a real-time facial image acquisition step: acquiring a real-time image captured by the camera device, and extracting a real-time facial image from the real-time image by using a face recognition algorithm;

a feature point recognition step: inputting the real-time facial image into a pre-trained lip average model, and using the lip average model to identify t lip feature points representing the position of the lips in the real-time facial image;

a lip region recognizing step: determining a lip region according to the t lip feature points, inputting the lip region into a pre-trained lip classification model, and determining whether the lip region is a human lip region;

Lip motion judging step: If the lip region is a human lip region, the moving direction and the moving distance of the lip in the real-time facial image are calculated according to the x and y coordinates of the t lip feature points in the real-time facial image.
The electronic device according to claim 1, wherein when the lip motion analysis program is executed by the processor, the following steps are further implemented:

Prompting step: When the lip classification model judges that the lip region is not the human lip region, the prompt does not detect the human lip region from the current live image, cannot determine the lip motion, and returns to the real-time facial image acquisition step.
The electronic device according to claim 2, wherein the lip movement determining step comprises:

Calculating the distance between the central feature point of the inner side of the upper lip and the central feature point of the inner side of the lower lip in the real-time facial image, and determining the degree of opening of the lip;

The characteristic points of the left outer lip corner feature points and the feature points closest to the left outer lip corner feature points on the outer contour lines of the upper and lower lips are respectively connected to form a vector
Calculation vector
The angle between the left side of the lips; and

The feature points of the right outer lip corner point and the feature points closest to the right outer lip corner feature point on the outer contour line of the upper and lower lips are respectively connected to form a vector
Calculation vector
The angle between the right side of the lips is obtained.
The electronic device according to claim 2, wherein the training step of the lip classification model comprises:

Collecting m lip positive sample images and k lips negative sample images to form a second sample library;

Extracting the local features of the positive sample image of each lip and the negative sample image of the lips;

The support vector machine classifier is trained by using the positive sample image of the lips, the negative sample image of the lips and its local features to obtain a lip classification model of the face.
The electronic device according to claim 1, wherein the training step of the lip average model comprises:

A first sample library having n face images is created, and t feature points are marked in the lip portion of each face image in the first sample library, and the t feature points are evenly distributed on the upper and lower sides. Lips and left and right lip angles; and

The face feature recognition model is trained using the face image marking the lip feature point to obtain a lip average model for the face.
The electronic device according to claim 5, wherein the face feature recognition model is an ERT algorithm, which is formulated as follows:

Where t represents the cascading sequence number, τ t (·, ·) represents the current class of the regression, S(t) is the shape estimate of the current model, and each regression τ t (·, ·) is based on the input current image I and S(t) to predict an increment
In the process of model training, some feature points of all sample images are trained to train the first regression tree, and the predicted value of the first regression tree and the residual of the true value of the partial feature points are used to train the second tree. The tree... and so on, until the predicted value of the Nth tree is trained and the true value of the partial feature points is close to 0, all the regression trees of the ERT algorithm are obtained, and the mouth average model of the face is obtained according to the regression trees.
The electronic device according to claim 1, wherein the face recognition algorithm comprises: a geometric feature based method, a local feature analyzing method, a feature face method, an elastic model based method, and a neural network method.
A lip motion analysis method is applied to an electronic device, characterized in that the method comprises:

a real-time facial image acquisition step: acquiring a real-time image captured by the camera device, and extracting a real-time facial image from the real-time image by using a face recognition algorithm;

a feature point recognition step: inputting the real-time facial image into a pre-trained lip average model, and using the lip average model to identify t lip feature points representing the position of the lips in the real-time facial image;

a lip region recognizing step: determining a lip region according to the t lip feature points, inputting the lip region into a pre-trained lip classification model, and determining whether the lip region is a human lip region;

Lip motion judging step: If the lip region is a human lip region, the moving direction and the moving distance of the lip in the real-time facial image are calculated according to the x and y coordinates of the t lip feature points in the real-time facial image.
The lip motion analysis method according to claim 8, wherein the method further comprises:

Prompting step: When the lip classification model judges that the lip region is not the human lip region, the prompt does not detect the human lip region from the current live image, cannot determine the lip motion, and returns to the real-time facial image acquisition step.
The lip motion analysis method according to claim 9, wherein the lip motion determining step comprises:

Calculating the distance between the central feature point of the inner side of the upper lip and the central feature point of the inner side of the lower lip in the real-time facial image, and determining the degree of opening of the lip;

The characteristic points of the left outer lip corner feature points and the feature points closest to the left outer lip corner feature points on the outer contour lines of the upper and lower lips are respectively connected to form a vector
Calculation vector
The angle between the left side of the lips; and

The feature points of the right outer lip corner point and the feature points closest to the right outer lip corner feature point on the outer contour line of the upper and lower lips are respectively connected to form a vector
Calculation vector
The angle between the right side of the lips is obtained.
The lip motion analysis method according to claim 9, wherein the training step of the lip classification model comprises:

Collecting m lip positive sample images and k lips negative sample images to form a second sample library;

Extracting the local features of the positive sample image of each lip and the negative sample image of the lips;

The support vector machine classifier is trained by using the positive sample image of the lips, the negative sample image of the lips and its local features to obtain a lip classification model of the face.
The lip motion analysis method according to claim 8, wherein the training step of the lip average model comprises:

A first sample library having n face images is created, and t feature points are marked in the lip portion of each face image in the first sample library, and the t feature points are evenly distributed on the upper and lower sides. Lips and left and right lip angles; and

The face feature recognition model is trained using the face image marking the lip feature point to obtain a lip average model for the face.
The lip motion analysis method according to claim 12, wherein the face feature recognition model is an ERT algorithm, which is expressed by the following formula:

Where t represents the cascading sequence number, τ t (·, ·) represents the current class of the regression, S(t) is the shape estimate of the current model, and each regression τ t (·, ·) is based on the input current image I and S(t) to predict an increment
In the process of model training, some feature points of all sample images are trained to train the first regression tree, and the predicted value of the first regression tree and the residual of the true value of the partial feature points are used to train the second tree. The tree... and so on, until the predicted value of the Nth tree is trained and the true value of the partial feature points is close to 0, all the regression trees of the ERT algorithm are obtained, and the mouth average model of the face is obtained according to the regression trees.
The lip motion analysis method according to claim 8, wherein the face recognition algorithm comprises: a geometric feature based method, a local feature analysis method, a feature face method, an elastic model based method, and a neural network method.
A computer readable storage medium, comprising: a lip motion analysis program, wherein the lip motion analysis program is executed by a processor to implement the following steps:

a real-time facial image acquisition step: acquiring a real-time image captured by the camera device, and extracting a real-time facial image from the real-time image by using a face recognition algorithm;

a feature point recognition step: inputting the real-time facial image into a pre-trained lip average model, and using the lip average model to identify t lip feature points representing the position of the lips in the real-time facial image;

a lip region recognizing step: determining a lip region according to the t lip feature points, inputting the lip region into a pre-trained lip classification model, and determining whether the lip region is a human lip region;

Lip motion judging step: If the lip region is a human lip region, the moving direction and the moving distance of the lip in the real-time facial image are calculated according to the x and y coordinates of the t lip feature points in the real-time facial image.
The computer readable storage medium according to claim 15, wherein when said lip motion analysis program is executed by a processor, the following steps are further implemented:

Prompting step: When the lip classification model judges that the lip region is not the human lip region, the prompt does not detect the human lip region from the current live image, cannot determine the lip motion, and returns to the real-time facial image acquisition step.
The computer readable storage medium of claim 16, wherein the lip motion determining step comprises:

Calculating the distance between the central feature point of the inner side of the upper lip and the central feature point of the inner side of the lower lip in the real-time facial image, and determining the degree of opening of the lip;

The characteristic points of the left outer lip corner feature points and the feature points closest to the left outer lip corner feature points on the outer contour lines of the upper and lower lips are respectively connected to form a vector
Calculation vector
The angle between the left side of the lips; and

The characteristic points of the right outer lip corner point and the feature points closest to the right outer lip corner feature point on the outer contour lines of the upper and lower lips are respectively connected to form a vector.
Calculation vector
The angle between the right side of the lips is obtained.
The computer readable storage medium of claim 16 wherein the training step of the lip classification model comprises:

Collecting m lip positive sample images and k lips negative sample images to form a second sample library;

Extracting the local features of the positive sample image of each lip and the negative sample image of the lips;

The support vector machine classifier is trained by using the positive sample image of the lips, the negative sample image of the lips and its local features to obtain a lip classification model of the face.
The computer readable storage medium of claim 15 wherein the training step of the lip average model comprises:

A first sample library having n face images is created, and t feature points are marked in the lip portion of each face image in the first sample library, and the t feature points are evenly distributed on the upper and lower sides. Lips and left and right lip angles; and

The face feature recognition model is trained using the face image marking the lip feature point to obtain a lip average model for the face.
The computer readable storage medium according to claim 19, wherein the face feature recognition model is an ERT algorithm, which is formulated as follows:

Where t represents the cascading sequence number, τ t (·, ·) represents the current class of the regression, S(t) is the shape estimate of the current model, and each regression τ t (·, ·) is based on the input current image I and S(t) to predict an increment
In the process of model training, some feature points of all sample images are trained to train the first regression tree, and the predicted value of the first regression tree and the residual of the true value of the partial feature points are used to train the second tree. The tree... and so on, until the predicted value of the Nth tree is trained and the true value of the partial feature points is close to 0, all the regression trees of the ERT algorithm are obtained, and the mouth average model of the face is obtained according to the regression trees.