WO2019232866A1 - Human eye model training method, human eye recognition method, apparatus, device and medium - Google Patents

Abstract

A human eye model training method, a human eye recognition method, an apparatus, a device and a medium. The human eye model training method comprises: acquiring a face image sample and marking the face image sample so as to obtain face image sample data, and extracting a feature vector of the face image sample, wherein the face image sample data comprises the face image sample and marking data (S10); dividing the face image sample data into training sample data and verification sample data (S20); using the training sample data to train a support vector machine classifier so as to obtain a critical plane of the support vector machine classifier (S30); calculating a vector distance of a feature vector of a verification sample in the verification sample data and the critical plane (S40); acquiring a preset true positive rate or a preset false positive rate, and acquiring a classification threshold according to the vector distance and the marking data corresponding to the verification sample (S50); acquiring a human eye determination model according to the classification threshold (S60). The described method may obtain a human eye determination model that is highly accurate in determining whether a human eye is occluded.

Description

Human eye model training method, human eye recognition method, device, equipment and medium

This application is based on a Chinese invention patent application filed on June 8, 2018 with application number 201810585092.2, entitled "Human Eye Model Training Method, Human Eye Recognition Method, Device, Equipment, and Medium" and claims priority.

Technical field

The present application relates to the field of computer technology, and in particular, to a human eye model training method, a human eye recognition method, a device, a device, and a medium.

Background technique

With the rapid development of artificial intelligence, human eye location recognition has received extensive attention and has become a hot topic in the field of artificial intelligence. Traditionally, in existing facial feature point recognition algorithms, the positions of different organs, such as eyes, ears, mouth, or nose, can be marked from a face picture, even if the corresponding part is blocked (glasses, hair, cover Mouth and other actions), the algorithm can still identify the relative positions of different parts and provide corresponding pictures. However, in some image processing processes, an unobstructed eye image is required. However, the eye pictures identified by the conventional facial feature point recognition algorithm cannot filter the obstructed pictures, which is easy to introduce errors and is not conducive to further development. To deal with needs.

Summary of the Invention

Based on this, it is necessary to provide a human eye model training method, device, computer equipment, and storage medium that can improve the efficiency of model training in response to the above technical problems.

In addition, it is necessary to propose a human eye recognition method. After training according to the human eye model training method, the trained human eye pictures are used for recognition to improve the accuracy of human eye recognition.

A human eye model training method includes:

Obtaining a face image sample, and labeling the face image sample to obtain face image sample data, and extracting a feature vector of the face image sample from the face image sample data, wherein the face image sample data Including face image samples and annotation data;

Dividing the face image sample data into training sample data and verification sample data;

Training the support vector machine classifier using the training sample data to obtain a critical surface of the support vector machine classifier;

Calculating a distance between a feature vector of a verification sample in the verification sample data and a vector of the critical surface;

Obtaining a preset true class rate or a preset false positive class rate, and obtaining a classification threshold according to the vector distance and the labeled data corresponding to the verification sample data;

A human eye judgment model is obtained according to the classification threshold.

A human eye model training device includes:

A facial image sample data acquisition module is configured to acquire a facial image sample, and mark the facial image sample to obtain facial image sample data, and extract a facial image sample from the facial image sample data. Feature vectors, where the face image sample data includes face image samples and annotation data;

A face image sample data division module, configured to divide the face image sample data into training sample data and verification sample data;

A critical surface acquisition module, configured to train a support vector machine classifier using the training sample data to obtain a critical surface of the support vector machine classifier;

A vector distance calculation module, configured to calculate a vector distance between a feature vector of a verification sample and the critical surface in the verification sample data;

A classification threshold obtaining module, configured to obtain a preset true classification rate or a preset false positive classification rate, and obtain a classification threshold according to the vector distance and the labeled data corresponding to the verification sample data;

The human eye judgment model acquisition module is configured to acquire a human eye judgment model according to the classification threshold.

A human eye recognition method includes:

Obtain a face picture to be identified, and use a facial feature point detection algorithm to obtain a positive eye area image;

Performing normalization processing on the forward eye area image to obtain an eye image to be identified;

And inputting the eye image to be identified into a human eye judgment model trained by the human eye model training method to identify and obtain a recognition result.

A human eye recognition device includes:

A face picture acquisition module to be recognized is used to obtain a face picture to be identified, and a facial feature point detection algorithm is used to obtain a positive eye area image;

A to-be-recognized eye image acquisition module, configured to perform normalization processing on the forward eye area image to obtain the to-be-recognized eye image;

A recognition result acquisition module is configured to input the eye image to be recognized into a human eye judgment model trained by the human eye model training method to recognize and obtain a recognition result.

A computer device includes a memory, a processor, and computer-readable instructions stored in the memory and executable on the processor. When the processor executes the computer-readable instructions, the following steps are implemented:

Obtaining a face image sample, and labeling the face image sample to obtain face image sample data, and extracting a feature vector of the face image sample from the face image sample data, wherein the face image sample data Including face image samples and annotation data;

Dividing the face image sample data into training sample data and verification sample data;

Training the support vector machine classifier using the training sample data to obtain a critical surface of the support vector machine classifier;

Calculating a distance between a feature vector of a verification sample in the verification sample data and a vector of the critical surface;

Obtaining a preset true class rate or a preset false positive class rate, and obtaining a classification threshold according to the vector distance and the labeled data corresponding to the verification sample data;

A human eye judgment model is obtained according to the classification threshold.

One or more non-volatile readable storage media storing computer-readable instructions, which when executed by one or more processors, cause the one or more processors to perform the following steps:

Obtaining a face image sample, and labeling the face image sample to obtain face image sample data, and extracting a feature vector of the face image sample from the face image sample data, wherein the face image sample data Including face image samples and annotation data;

Dividing the face image sample data into training sample data and verification sample data;

Training the support vector machine classifier using the training sample data to obtain a critical surface of the support vector machine classifier;

Calculating a distance between a feature vector of a verification sample in the verification sample data and a vector of the critical surface;

Obtaining a preset true class rate or a preset false positive class rate, and obtaining a classification threshold according to the vector distance and the labeled data corresponding to the verification sample data;

A human eye judgment model is obtained according to the classification threshold.

Details of one or more embodiments of the present application are set forth in the accompanying drawings and description below. Other features and advantages of the application will become apparent from the description, the drawings, and the claims.

BRIEF DESCRIPTION OF THE DRAWINGS

In order to explain the technical solutions of the embodiments of the present application more clearly, the drawings used in the description of the embodiments of the application will be briefly introduced below. Obviously, the drawings in the following description are just some embodiments of the application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without paying creative labor.

1 is a schematic diagram of an application environment of a human eye model training method and a human eye recognition method according to an embodiment of the present application;

FIG. 2 is an implementation flowchart of a human eye model training method provided by an embodiment of the present application; FIG.

3 is a flowchart of implementing step S10 in a human eye model training method according to an embodiment of the present application;

FIG. 4 is a flowchart of implementing step S30 in a human eye model training method according to an embodiment of the present application; FIG.

5 is a flowchart of implementing step S15 in a human eye model training method according to an embodiment of the present application;

FIG. 6 is an implementation flowchart of step S50 in a human eye model training method according to an embodiment of the present application;

7 is a schematic diagram of a human eye model training device provided by an embodiment of the present application;

FIG. 8 is a flowchart of implementing a human eye recognition method according to an embodiment of the present application; FIG.

9 is a schematic diagram of a human eye recognition device provided by an embodiment of the present application;

FIG. 10 is a schematic diagram of a computer device according to an embodiment of the present application.

Detailed ways

In the following, the technical solutions in the embodiments of the present application will be clearly and completely described with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, but not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by a person of ordinary skill in the art without creative efforts shall fall within the protection scope of this application.

The human eye model training method provided in this application can be applied in the application environment shown in FIG. 1, in which a client communicates with a server through a network, the server receives training sample data sent by the client, and establishes a human eye judgment model. Furthermore, it receives the verification samples sent by the client, and performs human eye judgment model training. Among them, the client can be, but is not limited to, various personal computers, notebook computers, smart phones, tablet computers, and portable wearable devices. The server can be implemented by an independent server or a server cluster composed of multiple servers.

In one embodiment, as shown in FIG. 2, the method is applied to the server in FIG. 1 as an example for description, and includes the following steps:

S10: Obtain a facial image sample, and mark the facial image sample to obtain facial image sample data, and extract a feature vector of the facial image sample from the facial image sample data, where the facial image sample data includes a person Face image samples and annotation data.

The face image sample data is human eye image data used for model training. The feature vector of a face image sample refers to a vector used to characterize the image information characteristics of each face image sample in the face image sample data, for example: a HOG (Histogram of Oriented Gradient) feature vector, LBP (Local Binary Patterns (Local Binary Patterns) feature vector or PCA (Principal Component Analysis) feature vector. Feature vectors can represent image information with simple data and avoid repeated extraction operations in subsequent training processes.

Preferably, in this embodiment, a HOG feature vector of a face image sample can be extracted. Since the HOG feature vector of the face image sample is described by the gradient of the local information of the face image sample, extracting the HOG feature vector of the face image sample can avoid the influence of factors such as geometric deformation and light changes on the training of the human eye model. . Marking a face image sample refers to dividing the face image sample into a positive sample (unblocked eye image) and a negative sample (blocked eye image) according to the content of the sample, and labeling these two sample data respectively Then, the face image sample data was obtained. The face image samples include positive samples and negative samples. Understandably, the face image sample data includes face image samples and annotation data. Preferably, the number of negative samples is 2-3 times the number of positive samples, which can make the sample information more comprehensive and improve the accuracy of model training.

In this embodiment, the face detection sample data is acquired for subsequent model training, and the occluded eye image is used as the face image sample for training, thereby reducing the false detection rate.

Optionally, the face image sample data includes, but is not limited to, a face image sample collected in advance and a face image sample stored in a commonly used face database in a memory in advance.

S20: Divide the face image sample data into training sample data and verification sample data.

Among them, the training sample data is sample data for learning, and a classifier is established by matching some parameters, that is, using the face image samples in the training sample data to train a machine learning model to determine the parameters of the machine learning model. Validation sample data is sample data used to verify the resolving power (such as recognition rate) of a trained machine learning model. Optionally, the number of 70% -75% of the face image sample data is used as training sample data, and the rest is used as verification sample data. In a specific embodiment, a total of 1000 positive face samples and 700 negative samples are selected to combine 1000 face image samples with adult face image sample data, of which 260 samples are used as verification sample data and 740 samples are used as training sample data.

S30: Use the training sample data to train the support vector machine classifier to obtain the critical surface of the support vector machine classifier.

Support Vector Machine (SVM) classifier is a discriminative classifier defined by the classification critical surface, which is used to classify or regression analysis the data. The critical surface is a classification surface that can correctly separate the two types of samples from the positive sample and the negative sample and maximize the distance between the two types of samples. Specifically, according to the characteristics of the face image sample data, a suitable kernel function is selected, and then the feature vector of the training sample data and the kernel function are used to perform a kernel function operation, so that the feature vector of the training sample data is mapped to a high-dimensional feature space to achieve the The feature vectors are linearly separable in this high-dimensional feature space to obtain a critical surface, and the critical surface is used as a classification surface for classifying training sample data, separating positive samples from negative samples. Specifically, by inputting training sample data, the support vector machine classifier will output a critical face training data to classify. The classification process of the support vector machine classifier is simplified by obtaining the critical surface.

In this embodiment, a support vector machine classifier is trained by using feature vectors of a face image sample to obtain a critical surface, which has a good classification ability and improves the efficiency of human eye model training.

S40: Calculate the distance between the feature vector of the verification sample and the vector of the critical surface in the verification sample data.

The verification sample data is pre-stored face image sample data for verification, which includes positive sample data (unblocked eye images) and negative sample data (blocked eye images). For these two types of sample data, Verification samples were obtained after labeling separately. The feature vector of the verification sample refers to a feature vector obtained by extracting a feature vector from the verification sample.

The feature vectors of the verification samples include, but are not limited to, HOG feature vectors, LBP feature vectors, and PCA feature vectors.

The distance between the feature vector of the verification sample and the vector of the critical plane in the verification sample data refers to the distance between the directed line segment corresponding to the feature vector of the verification sample in mathematical sense and a plane corresponding to the critical plane in mathematical sense. That is, the distance from a line to a surface in a mathematical sense. The distance is a value, and the distance is a vector distance. Suppose the expression of the critical surface is g (x) = wx + b, where w is a multi-dimensional vector, which can be expressed as w = [w ₁ , w ₂ , w ₃ ... w _n ], then the feature vector x arrives at The expression of the vector distance of the interface is

Where || w || represents the norm of w, that is,

By calculating the distance between the feature vector of the verification sample and the vector of the critical surface in the verification sample data, the closeness of each verification sample to the category to which it belongs can be intuitively compared.

S50: Obtain a preset true class rate or a preset false positive class rate, and obtain a classification threshold based on the vector distance and the labeled data corresponding to the verification sample data.

The preset true class rate refers to the preset ratio of the number of positive samples determined to be positive and the result is the total number of positive samples. The preset false positive class rate refers to the preset number of negative samples determined to be negative and the results are incorrect. The ratio of the total number of positive samples. In this embodiment, the true class rate refers to the ratio of the face image samples of the unblocked eye images to the total unblocked eye image face image samples, and the false positive class rate refers to The ratio of the face image samples of the occluded eye image determined to be unoccluded eyes to the total face image samples of the unoccluded eye image. It is easy to understand that the higher the true class rate or the lower the false positive class rate, it means that the classification requirements of the target are more stringent and can be adapted to more applications. Preferably, when the preset true class rate in this embodiment is 95%, or when the preset false positive rate is 5%, a good classification effect can be obtained, which can be adapted to a variety of different application scenarios, and by setting the real class reasonably Rate or false positive class rate, so that the adaptability of the support vector machine classifier is better extended.

It should be understood that the preset true class rate or the preset false positive class rate here is the preferred range of this application, but it can be set according to the needs of the actual application occasion, and there is no limitation here.

The classification threshold is a critical value used to classify samples. Specifically, when samples are classified, a judgment that is lower than the classification threshold is a positive sample, and a judgment that is higher than the classification threshold is a negative sample.

Specifically, the annotation data corresponding to the verification sample data refers to the annotation of the verification sample, for example, a positive sample is marked as 1 and a negative sample is marked as -1. After obtaining the distance between the feature vector of the verification sample and the critical surface and the label data of the verification sample, the classification threshold is calculated according to a preset true class rate or a preset false positive class rate.

For example, the preset false positive rate is 10%, and there are 15 verification samples of S ₁ , S ₂ ... S ₁₅ , among which there are 5 positive samples, 10 negative samples, and the feature vectors and critical surfaces of 10 negative samples. The vector distances of are respectively 1, 2, ... 10, so when the classification threshold is in the interval [1,2], if the classification threshold is 1.5, it can meet the preset false positive class rate of 10%.

S60: Obtain a human eye judgment model according to the classification threshold.

Specifically, the human eye judgment model refers to a model for judging whether an eye position is occluded in a face image sample. After the classification threshold is determined, the feature vector of the face image sample data and the vector distance of the critical surface of the support vector machine classifier are compared with the classification threshold, and the face image sample data is classified according to the comparison result to determine the face The position of the eyes in the image sample is either occluded or unoccluded. Therefore, after the classification threshold is given, the human eye judgment model is established. After inputting the face image to be identified into the human eye judgment model, the classification result of yes or no will be directly given according to the classification threshold, thus avoiding repeated training. To improve the efficiency of human eye model training.

In this embodiment, first obtain a face image sample and mark the face image sample to obtain the face image sample data, extract the feature vector of the face image sample in the face image sample data, and then convert the face image sample The data is divided into training sample data and verification sample data; the training sample data is used to train the support vector machine classifier to obtain the critical surface of the support vector machine classifier, thereby simplifying the classification process, and then calculating the characteristics of the verification samples in the verification sample data The vector distance between the vector and the critical surface of the support vector machine classifier can intuitively compare the closeness of each verification sample to the category to which it belongs, and obtain the preset true class rate or the preset false positive class rate in order to extend the support vector machine classifier. Adaptability. The classification threshold is obtained based on the vector distance and the labeled data corresponding to the verification sample data. Finally, the human eye judgment model is obtained to avoid repeated training and improve the efficiency of human eye model training.

In an embodiment, as shown in FIG. 3, in step S10, the feature vector of the facial image sample in the facial image sample data is extracted, and specifically includes the following steps:

S11: Use facial feature point detection algorithm to obtain facial feature points. The facial feature points include: left eye corner point, right eye corner point, and eyebrow center point; among which, the left eye corner point, right eye corner point, and eyebrow center point belong to the same eye area. Feature points.

Among them, the facial feature point detection algorithm refers to an algorithm for detecting facial features and marking position information. Face feature points refer to points used to mark the contours of the eyes, nose, and mouth, such as corner points, nose points, and mouth corner points. Specifically, the face feature point detection algorithm includes, but is not limited to, a face feature point detection algorithm based on deep learning, a face feature point detection algorithm based on a model, or a face feature point detection algorithm based on cascade shape regression.

Optionally, the facial feature points can be obtained by using the Viola-Jones algorithm based on Harr features that comes with OpenCV. Among them, OpenCV is a cross-platform computer vision library that can run on Linux, Windows, Android, and Mac OS operating systems. It consists of a series of C functions and a small number of C ++ classes. It also provides interfaces for languages such as Python, Ruby, and MATLAB. Many general algorithms in image processing and computer vision have been implemented, and the Viola-Jones algorithm based on Harr features is one of facial feature point detection algorithms. Haar feature is a feature that reflects the gray change of an image, and is a feature that reflects the difference between pixel sub-modules. Haar features are divided into three categories: edge features, linear features, and center-diagonal features. The Viola-Jones algorithm is a method for face detection based on haar feature values of a face.

Specifically, the input facial image sample data is obtained, the facial image sample data is preprocessed, and then the skin color region segmentation, face feature region segmentation, and face feature region classification steps are sequentially performed, and finally according to the Harr feature Viola- The Jones algorithm performs matching calculations on the classification of facial feature regions to obtain the facial feature point information of the facial image.

In this embodiment, the left eye corner point, the right eye corner point, and the eyebrow point of the face image sample are obtained by using a face feature point detection algorithm, so as to determine the area where the eyes of the face image sample are located according to the position information of these feature points. . Understandably, the left eye corner point, right eye corner point, and eyebrow center point mentioned in this step are three feature points belonging to the same eye area, for example, three feature points corresponding to the left eye or three feature points corresponding to the right eye. . In one embodiment, for a face image sample, only an image of one of the eyes (left eye or right eye) can be collected. If there is a need to process two eyes, after collecting the image of one eye, mirroring it can be used as the image of the other eye in a face image sample to save acquisition time and improve data processing efficiency.

S12: Perform forward adjustment on the face image sample according to the left eye corner point and the right eye corner point.

Among them, the forward adjustment is a normalization of the orientation of the feature points of the face and is set to a forward adjustment. In this embodiment, forward adjustment refers to adjusting the left eye corner point and the right eye corner point on the same horizontal line (that is, the vertical coordinates of the left eye corner point and the right eye corner point are equal), thereby normalizing the human eye feature points to the same orientation. In order to avoid the impact of training sample orientation changes on model training. Improve the robustness of face image samples to changes in orientation.

S13: Construct a rectangular area of the eye according to the left corner point, the right corner point, and the eyebrow center point.

The rectangular area of the eye refers to a rectangular area including an eye image. In a specific embodiment, the position coordinates of the left corner point, the right corner point, and the eyebrow center point are located using a facial feature point detection algorithm. The abscissa of the corner of the eye is the left coordinate, the abscissa of the right corner of the eye is the right coordinate, the ordinate of the eyebrow point is the upper coordinate, and the ordinate of the left eye corner point (or the ordinate of the right eye corner point) plus the eyebrow center The distance from the point to the left eye corner point in the vertical direction is the lower side coordinates. The rectangular area formed by these four point coordinates (left side coordinates, right side coordinates, upper side coordinates, and lower side coordinates) is the eye rectangular area.

S14: Perform image normalization processing on the rectangular area of the eyes to obtain a normalized rectangular area of the eyes.

Among them, normalization processing refers to performing a series of transformations on an image to be processed to convert the image to be processed into a corresponding standard form. Such as image size normalization, image grayscale normalization and so on. Preferably, the normalization process refers to size normalization of a rectangular area of the eye. Specifically, the eye rectangular area is set to a fixed size according to the resolution of the face image sample. For example, the eye rectangular area can be set to a Size (48,32) rectangle, that is, a rectangular area with a length of 48 pixels and a width of 32 pixels. By setting the rectangular area of the eyes to a fixed size, the complexity of feature vector extraction is subsequently reduced.

It is easy to understand that the image normalization processing on the rectangular area of the eye is conducive to the subsequent training of the support vector machine model. It can avoid the attribute of the large numerical interval to be over-branched and the attribute of the small numerical interval, and it can also avoid the complex value during the calculation degree.

S15: Extract the HOG feature vector according to the normalized rectangular area of the eyes.

The HOG (Histogram of Oriented Gradient, HOG) feature vector is a vector used to describe the gradient direction information of a local area of the image. This feature is greatly affected by changes in image size and position. The fixed input image range makes the calculated HOG feature vector more accurate. Unified, model training can pay more attention to the difference between unobstructed eye images and obstructed eye images without paying attention to changes in eye position, which is more convenient for training. At the same time, the HOG feature vector itself focuses on image gradient features rather than color features. It is not greatly affected by changes in illumination and changes in geometric shapes. Therefore, extracting HOG feature vectors can conveniently and efficiently extract feature vectors from face image samples. Among them, according to different detection targets, feature extraction is also different. Generally, color, texture, and shape are used as target features. According to the requirements for detecting the accuracy of the human eye image, this embodiment chooses to use the shape feature and the HOG feature vector of the training sample.

In this embodiment, a facial feature point detection algorithm is used to obtain the left eye corner point, the right eye corner point, and the eyebrow center point of the facial feature point; then, the image sample is adjusted forward to improve the robustness of the face image to the direction change. Then, the eye rectangular area is constructed and the eye rectangular area is subjected to image normalization processing to obtain the normalized eye rectangular area, which is conducive to subsequent training of the support vector machine model, and finally extracts the normalized eye rectangular area HOG feature vector, so that It is convenient and efficient to extract feature vectors from the face image samples in the face image sample data.

In an embodiment, as shown in FIG. 4, in step S30, training sample data is used to train a support vector machine classifier to obtain a critical surface of the support vector machine classifier, which specifically includes the following steps:

S31: Obtain the kernel function of the support vector machine classifier and the penalty parameters of the support vector machine classifier, and use the following formula to solve the Lagrangian multiplier

And decision threshold b:

0≤α _i ≤C, i = 1, ... l

In the formula, st is the abbreviation of the constraint condition in the mathematical formula, and min means to replace the number formula under the constraint condition.

K (x _i , x _j ) is the kernel function of the support vector machine classifier, C is the penalty parameter of the support vector machine classifier, C> 0, α _i and Lagrange multiplier

Is the conjugate relationship, x _i is the feature vector of the training sample data, l is the number of feature vectors of the training sample data, and y _i is the label of the training sample data.

The kernel function is a kernel function in a support vector machine classifier, and is used to perform a kernel function operation on the feature vectors of training samples input during the training of the support vector machine classifier. Linear kernel function, polynomial kernel function, Gaussian kernel function, Gaussian kernel function, and radial basis kernel function. Because the support vector machine classifier in this embodiment is linearly separable, it is preferred that a linear kernel The function is a kernel function in the support vector machine classifier, so K (x _i , x _j ) = (x _i , x _j ). The linear kernel parameter has the characteristics of few parameters and fast operation speed, which is suitable for linearly separable cases. . y _i is the labeling of training sample data. Because it is a two-class classification problem of support vector machine classifiers, y _i can be 1 or -1. If the face image sample is a positive sample, y _i = 1. If the image samples are negative samples, y _i = -1.

The penalty parameter C is a parameter used to optimize the support vector machine classifier, and it is a certain value. It can solve the problem of classification of sample skew. Specifically, the number of samples in the two categories (also referred to as multiple categories) that participate in the classification is very different. For example, there are 10,000 positive samples and 100 negative samples. This will cause sample bias. The skew problem. At this time, the distribution of positive samples is wide. To solve the problem of sample skew, specifically, the value of C can be reasonably increased according to the ratio of the number of positive samples to the number of negative samples. The larger C is, the smaller the fault tolerance of the classifier is. The decision threshold b is a real number used to determine the threshold for decision classification in the process of a support vector machine classifier.

Specifically, by obtaining an appropriate kernel function K (x _i , x _j ) and setting an appropriate penalty parameter C, the formula is adopted

After performing the kernel function operation on the feature vectors and kernel functions of the training sample data, the optimal problem is solved, that is, the Lagrange multiplier

Value of the kernel function

Reached the minimum and got

Then, determine the range in the open interval (0, C)

Weight

And according to

Calculate the b value.

Solve the Lagrangian multiplier in the support vector machine classifier

And decision threshold b to obtain better parameters in order to build an efficient support vector machine classifier.

S32: According to the Lagrangian multiplier

And decision threshold b, the critical surface g (x) of the support vector machine classifier is obtained using the following formula:

Lagrangian multiplier obtained by training support vector machine classifier

And the decision threshold b, the Lagrangian multiplier of the training sample is adjusted

And the decision threshold b, and put them into the formula

The critical surface of the support vector machine classifier is obtained.

It is easy to understand that the critical surface is obtained through calculation, so that subsequent face image samples are classified according to the critical face training samples. The training program first extracts and saves the feature vectors of the samples, so that the extracted features can be saved during the continuous adjustment of the training parameters for multiple training processes. Time to get the training parameters that meet the requirements as soon as possible. In this way, the false positive rate and accuracy rate of a certain category can be adjusted without needing to repeatedly train the model, which improves the model training efficiency.

In this embodiment, first obtain an appropriate kernel function K (x _i , x _j ), set an appropriate penalty parameter C, perform a kernel function operation on the feature vector of the training sample data and the kernel function, and solve the support vector machine classifier. Decision threshold b in the algorithm to obtain better parameters, build a support vector machine classifier, and then divide the Lagrangian multiplier

And the decision threshold b are substituted into the formula

In the method, a critical surface g (x) is obtained, so that subsequent face image samples are classified according to the training data of the critical face, without the need to repeatedly train the model, which improves the efficiency of model training.

In an embodiment, as shown in FIG. 5, in step S15, the HOG feature vector is extracted according to the normalized rectangular area of the eyes, and specifically includes the following steps:

S151: Divide the normalized eye rectangular area into cell units, and calculate the size and direction of each pixel gradient of the cell unit.

Specifically, according to the actual needs and requirements of the support vector machine classifier, the manner of dividing the normalized eye rectangular region is also different. The sub-region and the sub-region may or may not overlap. Cell units are connected subregions of the image, that is, each subregion is composed of multiple cell units. For example, a 48 * 32 normalized rectangular area of the eye. Assuming a cell unit is 4 * 4 pixels, 2 * 2 cells make up a sub-region, then this normalized eye rectangular region has 6 * 4 sub-regions. The gradient direction interval of each cell unit from 0 ° to 180 ° is divided into 9 intervals, so a 9-dimensional vector can be used to describe a cell unit.

The specific process of obtaining the magnitude and direction of each pixel gradient of the normalized rectangular area of the eye is: first obtain the gradient of each pixel, if the pixel is (x, y), the gradient calculation formula is as follows:

Where G _x (x, y) is the horizontal gradient of the pixel (x, y), where G _y (x, y) is the vertical gradient of the pixel (x, y), and H (x, y) is the pixel ( x, y). Then use the following formula to calculate the gradient of the pixel:

Among them, G (x, y) is the size of the pixel gradient.

Finally, the direction of the pixel gradient is calculated using the following formula:

Where α (x, y) is the directional angle of the direction of the pixel gradient.

S152: Count the gradient histogram of the magnitude and direction of each pixel gradient of the cell unit.

Among them, the gradient histogram refers to a histogram obtained by statistically calculating the magnitude and direction of the pixel gradient, and is used to characterize the gradient information of each cell unit. Specifically, first divide the gradient direction of each cell unit from 0 ° to 180 ° into 9 direction blocks, that is, 0 ° -20 ° is the first direction block, and 20 ° -40 ° is the second direction block. By analogy, 160 ° -180 ° is the ninth direction block. Then determine the direction block where the direction of the pixel gradient of the cell unit is located, and add the size of the pixel gradient of the direction block. For example, if the direction of a certain pixel of a cell unit falls between 40 ° and 60 °, the pixel value in the third direction of the gradient histogram is added to the magnitude of the pixel gradient in that direction to obtain the gradient histogram of the cell unit. Illustration.

S153: A histogram of gradients in series is obtained to obtain a HOG feature vector.

Among them, tandem refers to merging all gradient histograms of gradient histograms of each cell unit from left to right and top to bottom to obtain a normalized eye rectangular region HOG feature vector.

In this embodiment, the normalized eye rectangle area is divided into several small areas, and then the gradient histograms of the small areas are calculated. Finally, the gradient histograms corresponding to the small areas are connected in series to obtain the entire normalized eye rectangle. The gradient histogram of the region is used to describe the feature vector of the face image sample. At the same time, the HOG feature vector itself focuses on the image gradient feature rather than the color feature, and is not affected by the change of illumination. Extracting HOG feature vectors can easily and efficiently recognize human eye images.

In an embodiment, as shown in FIG. 6, in step S50, a preset true class rate or false positive class rate is obtained, and a classification threshold is obtained according to the vector distance and the labeled data corresponding to the verification sample data, which specifically includes the following steps:

S51: Draw a ROC curve according to the vector distance and the label data corresponding to the verification sample data.

Among them, ROC curve refers to the receiver's operating characteristic curve / receiver operating characteristic curve (receiver operating characteristic curve). It is a comprehensive index reflecting continuous variables of sensitivity and specificity. It is a composition method to reveal the relationship between sensitivity and specificity. . In this embodiment, the ROC curve shows the relationship between the true class rate and the false positive class rate of the support vector machine classifier. The closer the curve is to the upper left corner of the classifier, the higher the accuracy.

In the validation training sample, the samples are classified into positive and negative samples: positive samples (negative) or negative samples (negative). In the process of classifying the face image data in the verification training sample, four situations will occur: if the face image data is a positive sample and is also predicted as a positive sample, it is a true class (TP). Face image data is a negative sample that is predicted to be a positive sample, which is called a false positive (FP). Correspondingly, if the face image data is a negative sample, it is predicted as a negative sample, which is called a true negative (TN), and a positive sample is predicted as a negative sample, which is a false negative (FN).

The true class rate (TPR) characterizes the ratio of positive instances identified by the classifier to all positive instances. The calculation formula is TPR = TP / (TP + FN). The false positive rate (FPR) characterizes the proportion of negative instances that the classifier mistakes for positive samples to all negative instances. The calculation formula is FPR = FP / (FP + TN).

The process of drawing the ROC curve is: according to the feature vector of the verification sample data and the vector distance of the critical surface feature vector and the corresponding verification sample data annotation, the true class rate and false positive class rate of many verification samples are obtained. The ROC curve is false positive class The rate is the horizontal axis, and the true class rate is the vertical axis. Connect the points, that is, the true class rate and false positive class rate of many verification samples, draw a curve, and then calculate the area under the curve. The larger the area, the higher the judgment value.

In a specific implementation manner, the ROC curve drawing tool can be used for drawing. Specifically, the ROC curve is drawn using the plotSVMroc (true_labels, predict_labels, classnumber) function in matlab. Among them, true_labels are correct labels, predict_labels are labels for classification judgment, and classnumber is the number of classification classes. In this embodiment, because it is a binary classification problem of positive and negative samples, classnumber = 2. Specifically, after calculating the vector distance between the feature vector of the verification sample data and the critical surface feature vector, according to the vector distance distribution, that is, the distribution range of the closeness of each verification sample data to the critical surface, and according to the corresponding verification sample data, Annotate the true and false positive rates of the verification sample data, and then draw the ROC curve based on the true and false positive rates of the verification sample data.

S52: Obtain a classification threshold on the horizontal axis of the ROC curve according to a preset true class rate or a preset false positive class rate.

Specifically, the preset true class rate or preset false positive class rate is set according to actual use needs. After the server obtains the preset true class rate or preset false positive class rate, it passes the horizontal axis in the ROC curve. The false positive class rate and the true class rate represented by the vertical axis are compared with the preset true class rate or the preset false positive class rate, that is, the preset true class rate or the preset false positive class rate is used to classify the test sample data. The classification threshold is determined from the horizontal axis of the ROC curve according to the classification criteria, so that in the subsequent model training, different classification thresholds can be selected according to different scenarios through the ROC curve, which avoids the need for repeated training and improves the efficiency of model training.

In this embodiment, the true class rate and false positive class rate of the verification sample data can be obtained after calculating the vector distance between the feature vector of the verification sample data and the critical surface feature vector, and according to the corresponding verification sample data label. Draw ROC curves based on the true and false positive rates of the validation sample data. The classification threshold is obtained from the horizontal axis of the ROC curve by presetting the real class rate or the preset false positive class rate, so that in the subsequent model training, different classification thresholds can be selected according to different scenarios through the ROC curve to avoid the need for repeated training. Improve the efficiency of model training.

It should be understood that the size of the sequence numbers of the steps in the above embodiments does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

FIG. 7 shows a principle block diagram of a human eye model training device corresponding to the human eye model training method in the embodiment. As shown in FIG. 7, the human eye model training device includes a face image sample data acquisition module 10, a face image sample data division module 20, a critical surface acquisition module 30, a vector distance calculation module 40, a classification threshold acquisition module 50, and a person. Eye judgment model acquisition module 60. Among them, the realization function and implementation of the face image sample data acquisition module 10, the face image sample data division module 20, the critical surface acquisition module 30, the vector distance calculation module 40, the classification threshold acquisition module 50, and the human eye judgment model acquisition module 60 In the example, the corresponding steps of the human eye model training method are one-to-one. The detailed description of each functional module is as follows:

A face image sample data obtaining module 10, configured to obtain a face image sample, and mark the face image sample to obtain the face image sample data; and extract a feature vector of the face image sample from the face image sample data. The facial image sample data includes facial image samples and annotation data;

A face image sample data dividing module 20, configured to divide the face image sample data into training sample data and verification sample data;

A critical surface acquisition module 30 is configured to train a support vector machine classifier using training sample data to obtain a critical surface of the support vector machine classifier;

The vector distance calculation module 40 is configured to calculate a vector distance between a feature vector of a verification sample and a critical surface in the verification sample data;

A classification threshold obtaining module 50, configured to obtain a preset true class rate or a preset false positive class rate, and obtain a classification threshold according to a vector distance and labeled data corresponding to the verification sample data;

The human eye judgment model acquisition module 60 is configured to acquire a human eye judgment model according to a classification threshold.

Specifically, the facial image sample data acquisition module 10 includes a facial feature point acquisition unit 11, a forward adjustment unit 12, an eye rectangular region construction unit 13, an eye rectangular region acquisition unit 14, and a feature vector extraction unit 15.

A facial feature point acquisition unit 11 is configured to obtain a facial feature point by using a facial feature point detection algorithm. The facial feature point includes: a left eye corner point, a right eye corner point, and a brow center point; among which, the left eye corner point and the right eye corner point And the eyebrow center point are characteristic points belonging to the same eye area;

A forward adjustment unit 12 configured to perform forward adjustment on a face image sample according to a left eye corner point and a right eye corner point;

The eye rectangular region constructing unit 13 is configured to construct an eye rectangular region according to the left eye corner point, the right eye corner point, and the eyebrow center point;

The eye rectangular area obtaining unit 14 is configured to perform image normalization processing on the eye rectangular area to obtain a normalized eye rectangular area;

A feature vector extraction unit 15 is configured to extract a HOG feature vector according to a normalized rectangular area of the eye.

Specifically, the feature vector extraction unit 15 includes a pixel gradient acquisition subunit 151, a gradient histogram acquisition subunit 152, and a HOG feature vector acquisition subunit 153.

A pixel gradient acquisition subunit 151, configured to divide a normalized eye rectangular area into cell units, and calculate the size and direction of each pixel gradient of the cell unit;

The gradient histogram acquisition subunit 152 is used to count the gradient histogram of the magnitude and direction of each pixel gradient of the cell unit;

The HOG feature vector acquisition subunit 153 is used to concatenate gradient histograms to obtain a HOG feature vector.

Specifically, the critical surface acquisition module 30 includes a parameter acquisition unit 31 and a critical surface acquisition unit 32.

A parameter obtaining unit 31 is used for obtaining a kernel function of the support vector machine classifier and a penalty parameter of the support vector machine classifier, and solving the Lagrange multiplier by using the following formula

And decision threshold b:

0≤α _i ≤C, i = 1, ... l

In the formula, st is the abbreviation of the constraint condition in the mathematical formula, and min means to replace the number formula under the constraint condition.

K (x _i , x _j ) is the kernel function of the support vector machine classifier, C is the penalty parameter of the support vector machine classifier, C> 0, α _i and Lagrange multiplier

Is the conjugate relationship, x _i is the feature vector of the training sample data, l is the number of feature vectors of the training sample data, and y _i is the label of the training sample data;

Critical plane acquisition unit 32: used to obtain a Lagrangian multiplier

And decision threshold b, the critical surface g (x) of the support vector machine classifier is obtained using the following formula:

Specifically, the classification threshold acquisition module 50 includes a ROC curve drawing unit 51 and a classification threshold acquisition unit 52.

The ROC curve drawing unit 51 is configured to draw an ROC curve according to the vector distance and the labeled data corresponding to the verification sample data;

The classification threshold acquiring unit 52 is configured to acquire a classification threshold on a horizontal axis of the ROC curve according to a preset true class rate or a preset false positive class rate.

For specific limitations on the human eye model training device, reference may be made to the foregoing limitation on the human eye model training method, and details are not described herein again. Each module in the above-mentioned human eye model training device may be implemented in whole or in part by software, hardware, and a combination thereof. The above-mentioned modules may be embedded in the hardware in or independent of the processor in the computer device, or may be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

In one embodiment, a human eye recognition method is provided. The human eye recognition method can also be applied in the application environment as shown in FIG. 1, where a computer device communicates with a server through a network. The client communicates with the server through the network, and the server receives the face picture to be identified sent by the client for human eye recognition. Among them, the client can be, but is not limited to, various personal computers, notebook computers, smart phones, tablet computers, and portable wearable devices. The server can be implemented by an independent server or a server cluster composed of multiple servers.

In one embodiment, as shown in FIG. 8, the method is applied to the server in FIG. 1 as an example for description, and includes the following steps:

S70: Obtain a face picture to be identified, and use a facial feature point detection algorithm to obtain a positive eye area image.

The face picture to be identified refers to a face picture that needs to be recognized by human eyes. Specifically, the face image can be obtained by collecting a face picture in advance, or directly obtaining a face picture from a face database, such as an AR face database.

In this embodiment, the face pictures to be identified include unoccluded eye pictures and occluded eye pictures, and a facial feature point detection algorithm is used to obtain a positive eye area image. The implementation process of using the facial feature point detection algorithm to obtain a positive eye area image is the same as the method in steps S11 to S13, and details are not described herein again.

S80: Perform normalization processing on the forward eye area image to obtain an eye image to be identified.

The to-be-recognized eye image refers to a forward-looking eye area image after the normalization process is performed. By normalizing the forward-looking eye area image, the recognition efficiency can be improved. Specifically, the normalized to-be-recognized eye image is transformed to a unified standard form, thereby avoiding the attribute of the large-value interval in the support vector machine classifier from being over-branched with the attribute of the small-value interval, and also avoiding calculation Numerical complexity in the process. Optionally, the implementation process of normalizing the forward eye area image is the same as step S14, and details are not described herein again.

S90: Input the eye image to be identified into a human eye judgment model trained by the human eye model training method in steps S10 to S60 to perform recognition, and obtain a recognition result.

The recognition result refers to a result obtained by using a human eye judgment model for recognition of an eye image to be identified, including two cases: the eye image to be identified is an unobstructed eye image and the eye image to be identified is an obstructed eye image. Specifically, an eye image to be recognized is input to a human eye judgment model for recognition, so as to obtain a recognition result.

In this embodiment, first obtain a face picture to be identified, and perform normalization processing on the forward eye area image to obtain the to-be-recognized eye image, so as to input the normalized face image to be identified into the human eye judgment model. Perform recognition, obtain recognition results, quickly recognize whether the eyes of the face picture are occluded, and improve the recognition efficiency, thereby avoiding affecting the subsequent image processing process.

It should be understood that the size of the sequence numbers of the steps in the above embodiments does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

FIG. 9 shows a principle block diagram of a human eye recognition device that corresponds to the human eye recognition method in a one-to-one manner in the embodiment. As shown in FIG. 9, the human eye recognition device includes a to-be-recognized eye image acquisition module 70, an to-be-recognized eye image acquisition module 80, and a recognition result acquisition module 90. The functions of the eye image acquisition module 70, the eye image acquisition module 80, and the recognition result acquisition module 90 to be identified correspond to the steps corresponding to the human eye recognition method in the embodiment, and each functional module is described in detail as follows:

A face picture to be identified module 70 is configured to obtain a face picture to be identified, and a facial feature point detection algorithm is used to obtain a positive eye area image;

The eye image to be identified module 80 is configured to perform normalization processing on the forward eye area image to obtain the eye image to be identified;

The recognition result acquisition module 90 is configured to input an eye image to be recognized into a human eye judgment model trained by a human eye model training method to recognize and obtain a recognition result.

For the specific limitation of the human eye model training device, refer to the limitation on the human eye recognition method described above, and details are not described herein again. Each module in the above-mentioned human eye recognition device may be implemented in whole or in part by software, hardware, and a combination thereof. The above-mentioned modules may be embedded in the hardware in or independent of the processor in the computer device, or may be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

In one embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as shown in FIG. 10. The computer device includes a processor, a memory, a network interface, and a database connected through a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, computer-readable instructions, and a database. The internal memory provides an environment for the operation of the operating system and computer-readable instructions in a non-volatile storage medium. The database of the computer equipment is used to store the feature vector of the human face image sample data and the human eye model training data in the human eye model training method. The network interface of the computer device is used to communicate with an external terminal through a network connection. The computer-readable instructions are executed by a processor to implement a human eye model training method. Alternatively, when the computer-readable instructions are executed by a processor, the functions of each module / unit in the human eye recognition device in the embodiment are realized.

In one embodiment, a computer device is provided, including a memory, a processor, and computer-readable instructions stored on the memory and executable on the processor. When the processor executes the computer-readable instructions, the human eyes of the foregoing embodiments are implemented. The steps of the model training method are, for example, steps S10 to S60 shown in FIG. 2. Alternatively, when the processor executes the computer-readable instructions, the steps of the human eye recognition method of the foregoing embodiment are implemented, for example, steps S70 to S90 shown in FIG. 7. Alternatively, when the processor executes the computer-readable instructions, the functions of the modules / units of the human eye model training device of the foregoing embodiment are implemented, for example, modules 10 to 60 shown in FIG. 7. Alternatively, when the processor executes the computer-readable instructions, the functions of the modules / units of the human eye recognition device in the foregoing embodiment are implemented, for example, modules 70 to 90 shown in FIG. 9. To avoid repetition, we will not repeat them here.

One or more non-volatile readable storage media storing computer-readable instructions, and when the computer-readable instructions are executed by one or more processors, the one or more processors cause the human eye model training of the foregoing embodiment to be performed The steps of the method, or the steps of the human eye recognition method of the foregoing embodiment are implemented when the computer readable instructions are executed by one or more processors, or the human eyes of the above embodiments are implemented when the computer readable instructions are executed by one or more processors. The functions of the modules / units of the model training device, or the functions of the modules / units of the human eye recognition device of the above embodiment when computer-readable instructions are executed by one or more processors, to avoid repetition, details are not repeated here. .

A person of ordinary skill in the art can understand that all or part of the processes in the methods of the foregoing embodiments can be implemented by using computer-readable instructions to instruct related hardware. The computer-readable instructions can be stored in a non-volatile computer. In the readable storage medium, the computer-readable instructions, when executed, may include the processes of the embodiments of the methods described above.

The above embodiments are only used to illustrate the technical solutions of the present application, and are not limited thereto. Although the present application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still apply the foregoing embodiments. The recorded technical solutions are modified, or some of the technical features are equivalently replaced; and these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of this application, and shall be included in this application. Within the scope of protection.

Claims (20)

1. A human eye model training method, comprising:

   Obtaining a face image sample, and labeling the face image sample to obtain face image sample data, and extracting a feature vector of the face image sample from the face image sample data, wherein the face image sample data Including face image samples and annotation data;

   Dividing the face image sample data into training sample data and verification sample data;

   Training the support vector machine classifier using the training sample data to obtain a critical surface of the support vector machine classifier;

   Calculating a distance between a feature vector of a verification sample in the verification sample data and a vector of the critical surface;

   Obtaining a preset true class rate or a preset false positive class rate, and obtaining a classification threshold according to the vector distance and the labeled data corresponding to the verification sample data;

   A human eye judgment model is obtained according to the classification threshold.
2. The human eye model training method according to claim 1, wherein the extracting a feature vector of a face image sample in the face image sample data specifically comprises:

   A facial feature point detection algorithm is used to obtain a facial feature point, the facial feature points include: a left eye corner point, a right eye corner point, and an eyebrow point; wherein the left eye corner point, the right eye corner point, and the eyebrow point Are characteristic points belonging to the same eye area;

   Perform forward adjustment on the face image sample according to the left eye corner point and the right eye corner point;

   Constructing a rectangular area of the eye according to the left eye corner point, the right eye corner point, and the eyebrow center point;

   Performing image normalization processing on the eye rectangular area to obtain a normalized eye rectangular area;

   A HOG feature vector is extracted according to the normalized eye rectangular area.
3. The human eye model training method according to claim 1, wherein the training of a support vector machine classifier with training sample data to obtain a critical surface of the support vector machine classifier specifically includes:

   Obtain the kernel function of the support vector machine classifier and the penalty parameters of the support vector machine classifier, and use the following formula to solve the Lagrange multiplier

   

   And decision threshold b:

   

   In the formula, st is the abbreviation of the constraint condition in the mathematical formula, and min means to replace the number formula under the constraint condition.

   

   K (x _i , x _j ) is a kernel function of the support vector machine classifier, C is a penalty parameter of the support vector machine classifier, C> 0, α _i and the Lagrange Multiplier

   

   Is a conjugate relationship, x _i is a feature vector of the training sample data, l is the number of feature vectors of the training sample data, and y _i is a label of the training sample data;

   According to the Lagrange multiplier

   

   And the decision threshold b, use the following formula to obtain the critical surface g (x) of the support vector machine classifier:

   
4. The human eye model training method according to claim 2, wherein the extracting a HOG feature vector based on the normalized rectangular area of the eye specifically comprises:

   Divide the normalized eye rectangular area into cell units, and calculate the size and direction of each pixel gradient of the cell unit;

   A gradient histogram of the magnitude and direction of each pixel gradient of the cell unit;

   The gradient histograms are connected in series to obtain the HOG feature vector.
5. The human eye model training method according to claim 1, wherein the preset true class rate or the preset false positive class rate is obtained, and a classification threshold is obtained according to the vector distance and the label data corresponding to the verification sample data. , Including:

   Draw a ROC curve according to the vector distance and labeled data corresponding to the verification sample data;

   A classification threshold is obtained on the horizontal axis of the ROC curve according to the preset true class rate or the preset false positive class rate.
6. A human eye recognition method, comprising:

   Obtain a face picture to be identified, and use a facial feature point detection algorithm to obtain a positive eye area image;

   Performing normalization processing on the forward eye area image to obtain an eye image to be identified;

   The eye image to be identified is input to a human eye judgment model trained by the human eye model training method according to any one of claims 1-5 to perform recognition, and obtain a recognition result.
7. A human eye model training device, comprising:

   A facial image sample data acquisition module is configured to acquire a facial image sample, and mark the facial image sample to obtain facial image sample data, and extract a facial image sample from the facial image sample data. Feature vectors, where the face image sample data includes face image samples and annotation data;

   A face image sample data division module, configured to divide the face image sample data into training sample data and verification sample data;

   A critical surface acquisition module, configured to train a support vector machine classifier using the training sample data to obtain a critical surface of the support vector machine classifier;

   A vector distance calculation module, configured to calculate a vector distance between a feature vector of a verification sample and the critical surface in the verification sample data;

   A classification threshold obtaining module, configured to obtain a preset true classification rate or a preset false positive classification rate, and obtain a classification threshold according to the vector distance and the labeled data corresponding to the verification sample data;

   The human eye judgment model acquisition module is configured to acquire a human eye judgment model according to the classification threshold.
8. The human eye model training device according to claim 7, wherein the face image sample data acquisition module specifically comprises:

   A facial feature point acquisition unit is configured to obtain a facial feature point by using a facial feature point detection algorithm, and the facial feature point includes a left eye corner point, a right eye corner point, and an eyebrow center point, wherein the left eye corner point, the The right eye corner point and the eyebrow center point are characteristic points belonging to the same eye area;

   A forward adjustment unit, configured to perform forward adjustment on the face image sample according to the left eye corner point and the right eye corner point;

   An eye rectangular region constructing unit, configured to construct an eye rectangular region according to the left eye corner point, the right eye corner point, and the eyebrow center point;

   A feature vector extraction unit for obtaining a rectangular area of eyes, which is the same as performing image normalization processing on the rectangular area of eyes to obtain a normalized rectangular area of eyes;

   A HOG feature vector is extracted according to the normalized eye rectangular area.
9. The human eye model training device according to claim 7, wherein the critical surface acquisition module specifically comprises:

   A parameter obtaining unit, configured to obtain a kernel function of the support vector machine classifier and a penalty parameter of the support vector machine classifier, and solve the Lagrange multiplier by using the following formula

   

   And decision threshold b:

   

   In the formula, st is the abbreviation of the constraint condition in the mathematical formula, and min means to replace the number formula under the constraint condition.

   

   K (x _i , x _j ) is a kernel function of the support vector machine classifier, C is a penalty parameter of the support vector machine classifier, C> 0, α _i and the Lagrange Multiplier

   

   Is a conjugate relationship, x _i is a feature vector of the training sample data, l is the number of feature vectors of the training sample data, and y _i is a label of the training sample data;

   A critical surface acquisition unit, according to the Lagrangian multiplier

   

   And the decision threshold b, use the following formula to obtain the critical surface g (x) of the support vector machine classifier:

   
10. A human eye recognition device, comprising:

    A face picture acquisition module to be recognized is used to obtain a face picture to be identified, and a facial feature point detection algorithm is used to obtain a positive eye area image;

    A to-be-recognized eye image acquisition module, configured to perform normalization processing on the forward eye area image to obtain the to-be-recognized eye image;

    The eye image to be identified is input to a human eye judgment model for recognition, and a recognition result is obtained, wherein the human eye judgment model is obtained by using the following training method:

    Obtaining a face image sample, and labeling the face image sample to obtain face image sample data, and extracting a feature vector of the face image sample from the face image sample data, wherein the face image sample data Including face image samples and annotation data;

    Dividing the face image sample data into training sample data and verification sample data;

    Training the support vector machine classifier using the training sample data to obtain a critical surface of the support vector machine classifier;

    Calculating a distance between a feature vector of a verification sample in the verification sample data and a vector of the critical surface;

    Obtaining a preset true class rate or a preset false positive class rate, and obtaining a classification threshold according to the vector distance and the labeled data corresponding to the verification sample data;

    A human eye judgment model is obtained according to the classification threshold.
11. A computer device includes a memory, a processor, and computer-readable instructions stored in the memory and executable on the processor, and is characterized in that the processor implements the computer-readable instructions as follows step:

    Obtaining a face image sample, and labeling the face image sample to obtain face image sample data, and extracting a feature vector of the face image sample from the face image sample data, wherein the face image sample data Including face image samples and annotation data;

    Dividing the face image sample data into training sample data and verification sample data;

    Training the support vector machine classifier using the training sample data to obtain a critical surface of the support vector machine classifier;

    Calculating a distance between a feature vector of a verification sample in the verification sample data and a vector of the critical surface;

    Obtaining a preset true class rate or a preset false positive class rate, and obtaining a classification threshold according to the vector distance and the labeled data corresponding to the verification sample data;

    A human eye judgment model is obtained according to the classification threshold.
12. The computer device according to claim 11, wherein the extracting a feature vector of a face image sample in the face image sample data specifically comprises:

    A facial feature point detection algorithm is used to obtain a facial feature point, the facial feature points include: a left eye corner point, a right eye corner point, and an eyebrow point; wherein the left eye corner point, the right eye corner point, and the eyebrow point Are characteristic points belonging to the same eye area;

    Perform forward adjustment on the face image sample according to the left eye corner point and the right eye corner point;

    Constructing a rectangular area of the eye according to the left eye corner point, the right eye corner point, and the eyebrow center point;

    Performing image normalization processing on the eye rectangular area to obtain a normalized eye rectangular area;

    A HOG feature vector is extracted according to the normalized eye rectangular area.
13. The computer device according to claim 11, wherein the training of a support vector machine classifier with training sample data to obtain a critical surface of the support vector machine classifier specifically comprises:

    Obtain the kernel function of the support vector machine classifier and the penalty parameters of the support vector machine classifier, and use the following formula to solve the Lagrange multiplier

    

    And decision threshold b:

    

    In the formula, st is the abbreviation of the constraint condition in the mathematical formula, and min means to replace the number formula under the constraint condition.

    

    K (x _i , x _j ) is a kernel function of the support vector machine classifier, C is a penalty parameter of the support vector machine classifier, C> 0, α _i and the Lagrange Multiplier

    

    Is a conjugate relationship, x _i is a feature vector of the training sample data, l is the number of feature vectors of the training sample data, and y _i is a label of the training sample data;

    According to the Lagrange multiplier

    

    And the decision threshold b, use the following formula to obtain the critical surface g (x) of the support vector machine classifier:

    
14. The computer device according to claim 12, wherein the extracting the HOG feature vector based on the normalized rectangular area of the eye specifically comprises:

    Divide the normalized eye rectangular area into cell units, and calculate the size and direction of each pixel gradient of the cell unit;

    A gradient histogram of the magnitude and direction of each pixel gradient of the cell unit;

    The gradient histograms are connected in series to obtain the HOG feature vector.
15. A computer device includes a memory, a processor, and computer-readable instructions stored in the memory and executable on the processor, and is characterized in that the processor implements the computer-readable instructions as follows Step: Obtain a face picture to be identified, and use a facial feature point detection algorithm to obtain a positive eye area image;

    Performing normalization processing on the forward eye area image to obtain an eye image to be identified;

    The eye image to be identified is input to a human eye judgment model for recognition, and a recognition result is obtained, wherein the human eye judgment model is obtained by using the following training method:

    Obtaining a face image sample, and labeling the face image sample to obtain face image sample data, and extracting a feature vector of the face image sample from the face image sample data, wherein the face image sample data Including face image samples and annotation data;

    Dividing the face image sample data into training sample data and verification sample data;

    Training the support vector machine classifier using the training sample data to obtain a critical surface of the support vector machine classifier;

    Calculating a distance between a feature vector of a verification sample in the verification sample data and a vector of the critical surface;

    Obtaining a preset true class rate or a preset false positive class rate, and obtaining a classification threshold according to the vector distance and the labeled data corresponding to the verification sample data;

    A human eye judgment model is obtained according to the classification threshold.
16. One or more non-volatile readable storage media storing computer readable instructions, characterized in that when the computer readable instructions are executed by one or more processors, the one or more processors are caused to execute The following steps:

    Obtaining a face image sample, and labeling the face image sample to obtain face image sample data, and extracting a feature vector of the face image sample from the face image sample data, wherein the face image sample data Including face image samples and annotation data;

    Dividing the face image sample data into training sample data and verification sample data;

    Training the support vector machine classifier using the training sample data to obtain a critical surface of the support vector machine classifier;

    Calculating a distance between a feature vector of a verification sample in the verification sample data and a vector of the critical surface;

    Obtaining a preset true class rate or a preset false positive class rate, and obtaining a classification threshold according to the vector distance and the labeled data corresponding to the verification sample data;

    A human eye judgment model is obtained according to the classification threshold.
17. The non-volatile readable storage medium according to claim 16, wherein the extracting a feature vector of a face image sample in the face image sample data specifically comprises:

    A facial feature point detection algorithm is used to obtain a facial feature point, the facial feature points include: a left eye corner point, a right eye corner point, and an eyebrow point; wherein the left eye corner point, the right eye corner point, and the eyebrow point Are characteristic points belonging to the same eye area;

    Perform forward adjustment on the face image sample according to the left eye corner point and the right eye corner point;

    Constructing a rectangular area of the eye according to the left eye corner point, the right eye corner point, and the eyebrow center point;

    Performing image normalization processing on the eye rectangular area to obtain a normalized eye rectangular area;

    A HOG feature vector is extracted according to the normalized eye rectangular area.
18. The non-volatile readable storage medium according to claim 16, wherein the training a support vector machine classifier using training sample data to obtain a critical surface of the support vector machine classifier specifically comprises:

    Obtain the kernel function of the support vector machine classifier and the penalty parameters of the support vector machine classifier, and use the following formula to solve the Lagrangian multiplier

    

    And decision threshold b:

    

    In the formula, st is the abbreviation of the constraint condition in the mathematical formula, and min means to replace the number formula under the constraint condition.

    

    K (x _i , x _j ) is a kernel function of the support vector machine classifier, C is a penalty parameter of the support vector machine classifier, C> 0, α _i and the Lagrange Multiplier

    

    Is a conjugate relationship, x _i is a feature vector of the training sample data, l is the number of feature vectors of the training sample data, and y _i is a label of the training sample data;

    According to the Lagrange multiplier

    

    And the decision threshold b, use the following formula to obtain the critical surface g (x) of the support vector machine classifier:

    
19. The non-volatile readable storage medium according to claim 16, wherein the obtaining a preset true class rate or a preset false positive class rate is based on the vector distance and a label corresponding to the verification sample data Data acquisition classification threshold, including:

    Draw a ROC curve according to the vector distance and labeled data corresponding to the verification sample data;

    A classification threshold is obtained on the horizontal axis of the ROC curve according to the preset true class rate or the preset false positive class rate.
20. One or more non-volatile readable storage media storing computer readable instructions, characterized in that when the computer readable instructions are executed by one or more processors, the one or more processors are caused to execute The following steps:

    Obtain a face picture to be identified, and use a facial feature point detection algorithm to obtain a positive eye area image;

    Performing normalization processing on the forward eye area image to obtain an eye image to be identified;

    The eye image to be identified is input to a human eye judgment model for recognition, and a recognition result is obtained, wherein the human eye judgment model is obtained by using the following training method:

    Obtaining a face image sample, and labeling the face image sample to obtain face image sample data, and extracting a feature vector of the face image sample from the face image sample data, wherein the face image sample data Including face image samples and annotation data;

    Dividing the face image sample data into training sample data and verification sample data;

    Training the support vector machine classifier using the training sample data to obtain a critical surface of the support vector machine classifier;

    Calculating a distance between a feature vector of a verification sample in the verification sample data and a vector of the critical surface;

    Obtaining a preset true class rate or a preset false positive class rate, and obtaining a classification threshold according to the vector distance and the labeled data corresponding to the verification sample data;

    A human eye judgment model is obtained according to the classification threshold.

Images