CN117523642A - Face recognition method based on optimal-spacing Bayesian classification model - Google Patents
Face recognition method based on optimal-spacing Bayesian classification model Download PDFInfo
- Publication number
- CN117523642A CN117523642A CN202311633669.XA CN202311633669A CN117523642A CN 117523642 A CN117523642 A CN 117523642A CN 202311633669 A CN202311633669 A CN 202311633669A CN 117523642 A CN117523642 A CN 117523642A
- Authority
- CN
- China
- Prior art keywords
- face
- classification model
- optimal
- face recognition
- bayesian classification
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000013145 classification model Methods 0.000 title claims abstract description 34
- 238000000034 method Methods 0.000 title claims abstract description 34
- 238000012545 processing Methods 0.000 claims abstract description 10
- 230000009467 reduction Effects 0.000 claims abstract description 6
- 238000000605 extraction Methods 0.000 claims abstract description 5
- 238000007781 pre-processing Methods 0.000 claims abstract description 5
- 238000005259 measurement Methods 0.000 claims abstract description 3
- 238000012549 training Methods 0.000 claims description 25
- 239000011159 matrix material Substances 0.000 claims description 22
- 238000005516 engineering process Methods 0.000 claims description 12
- 238000000513 principal component analysis Methods 0.000 claims description 10
- 239000013598 vector Substances 0.000 claims description 10
- 238000001514 detection method Methods 0.000 claims description 8
- 238000005286 illumination Methods 0.000 claims description 5
- 238000004364 calculation method Methods 0.000 claims description 4
- 230000001186 cumulative effect Effects 0.000 claims description 4
- 238000005315 distribution function Methods 0.000 claims description 4
- 238000011156 evaluation Methods 0.000 claims description 4
- 230000006870 function Effects 0.000 claims description 4
- 239000002245 particle Substances 0.000 claims description 4
- 230000008569 process Effects 0.000 claims description 3
- 101000941926 Saccharomyces cerevisiae (strain ATCC 204508 / S288c) Carboxypeptidase Y inhibitor Proteins 0.000 claims description 2
- 238000000354 decomposition reaction Methods 0.000 claims description 2
- 238000013507 mapping Methods 0.000 claims description 2
- 238000010606 normalization Methods 0.000 claims description 2
- 238000012795 verification Methods 0.000 claims description 2
- 230000006872 improvement Effects 0.000 description 3
- 241000209094 Oryza Species 0.000 description 2
- 235000007164 Oryza sativa Nutrition 0.000 description 2
- 239000008280 blood Substances 0.000 description 2
- 210000004369 blood Anatomy 0.000 description 2
- 230000014509 gene expression Effects 0.000 description 2
- 239000011521 glass Substances 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 235000009566 rice Nutrition 0.000 description 2
- 238000012360 testing method Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 210000000887 face Anatomy 0.000 description 1
- 230000001815 facial effect Effects 0.000 description 1
- 230000008921 facial expression Effects 0.000 description 1
- 239000005433 ionosphere Substances 0.000 description 1
- 210000000554 iris Anatomy 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/172—Classification, e.g. identification
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/004—Artificial life, i.e. computing arrangements simulating life
- G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/46—Descriptors for shape, contour or point-related descriptors, e.g. scale invariant feature transform [SIFT] or bags of words [BoW]; Salient regional features
- G06V10/467—Encoded features or binary features, e.g. local binary patterns [LBP]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/50—Extraction of image or video features by performing operations within image blocks; by using histograms, e.g. histogram of oriented gradients [HoG]; by summing image-intensity values; Projection analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/764—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using classification, e.g. of video objects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/77—Processing image or video features in feature spaces; using data integration or data reduction, e.g. principal component analysis [PCA] or independent component analysis [ICA] or self-organising maps [SOM]; Blind source separation
- G06V10/7715—Feature extraction, e.g. by transforming the feature space, e.g. multi-dimensional scaling [MDS]; Mappings, e.g. subspace methods
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Software Systems (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Computing Systems (AREA)
- Databases & Information Systems (AREA)
- Medical Informatics (AREA)
- Human Computer Interaction (AREA)
- Oral & Maxillofacial Surgery (AREA)
- Life Sciences & Earth Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Molecular Biology (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Image Analysis (AREA)
Abstract
The invention provides a novel method for an optimal-spacing Bayesian classification model, which is applied to the field of face recognition. The face recognition method is different from the traditional naive Bayesian classification method, the proposed optimal interval Bayesian core concept is to utilize probability inference capability of naive Bayesian classification, and consider optimal interval measurement between samples at the same time so as to solve the problem of possible category overlapping, so that a classifier can better distinguish categories. The method comprises the following steps: firstly, capturing a target face in real time by using a camera, and preprocessing an image by adopting histogram equalization; then, extracting the face features by adopting a feature extraction method of a local binary pattern, and performing dimension reduction processing by using PCA; and finally, classifying the extracted features by using the proposed optimal interval Bayesian classification model, thereby outputting a final face recognition result. Experimental results show that the method provided by the invention has more excellent performance in the aspect of face recognition.
Description
Technical Field
The invention relates to the technical field of biological feature recognition, in particular to a face recognition method based on an optimal-spacing Bayesian classification model.
Background
In the current digital age, the biological recognition technology is a revolutionary technology, and makes a great contribution to the safety and convenience of human society. Biometric methods are distinguished by the unique dependence of individual biometric features, such as authentication and identification of faces, fingerprints, irises, sounds, etc. Face recognition is popular as a biometric solution with its excellent performance and wide range of applications. However, the conventional face recognition method still has some challenges in terms of processing image complexity, illumination variation, posture variation, expression variation, and the like. Naive bayes are widely used in face recognition, but there are some limitations to the traditional naive bayes classification model. One of the main limitations is its strong assumption of feature independence, which means that the model assumes that all features are conditional independent, and in practical applications, there may be a correlation between features. And also, category overlap problems may be encountered when processing complex data sets, resulting in reduced classifier accuracy. Although algorithms exist to solve the problem of the spacing between categories, there is still room for some potential drawbacks and improvements in practical applications.
Disclosure of Invention
The invention aims to overcome the technical defects and shortcomings, and provides a face recognition method based on an optimal-spacing Bayesian classification model.
The technical scheme adopted by the invention is as follows:
a face recognition method based on an optimal interval Bayesian classification model comprises the following steps:
step 1, capturing face images in real time through a camera or image acquisition equipment;
step 2, preprocessing the captured face image, and eliminating uneven illumination or other image quality problems by using a histogram equalization technology;
step 3, extracting the most representative and robust features in the face image by adopting a Local Binary Pattern (LBP) feature extraction method;
step 4, performing feature dimension reduction processing by using a Principal Component Analysis (PCA) technology to reduce feature dimensions, improve calculation efficiency and reserve main feature information;
step 5, training and identifying classification by using an optimal interval Bayesian classification model;
and step 6, outputting a face recognition result.
Step 1: capturing a face image, and detecting a face area in real time by using a Haar classifier to capture the face image in real time, wherein the method comprises the following specific steps of:
step 1.1, calculating Haar characteristics, and using pixels of a difference sum of adjacent rectangles at a specified position of a detection range;
step 1.2, accelerating to obtain Haar characteristics, and accelerating to calculate the Haar characteristics by using an integral graph:
0 sum ＝I(A)+I(B)+I(C)+I(D)
where O represents the detection window, A, B, C, D represents four vertices of the detection window, and I (x, y) represents the pixel value of the coordinate point;
step 1.3, normalizing the Haar characteristic values, and normalizing the characteristic values to ensure that the characteristic values have uniform scale or range.
The method comprises the following steps:
step 1.3.1, calculating the gray value and the square sum of the gray values of the images in the detection window:
sum＝ΣI(x,y)
sq sum ＝ΣI 2 (x,y)
step 1.3.2, calculating the average value:
step 1.3.3, normalization factor:
step 1.3.4, normalized eigenvalues:
finally, comparing the normalized characteristic value with a threshold value;
step 1.4, the training of the AdaBoost algorithm is used for distinguishing the human face from the non-human face, and the training of the AdaBoost algorithm comprises the following steps:
step 1.4.1, initializing sample weights. Assigning an equal weight value to each sample;
step 1.4.2, iteratively training a plurality of weak classifiers. Training a weak classifier, and classifying the data according to the current sample weight learning. And then, the error rate of the classifier on the weighted data set is calculated, the weight of the sample is adjusted, the weight of the error classification sample is improved, and the weight of the correct classification sample is reduced. Finally, calculating the weight of the weak classifier according to the classification accuracy of the weak classifier under the current weight, and generating a plurality of weak classifiers by continuously iterating the process;
and 1.4.3, constructing an integrated strong classifier. And (3) weighting and combining all the weak classifiers according to the weights of the weak classifiers to construct an integrated strong classifier.
Step 2: preprocessing the captured face image by using a histogram equalization technology, wherein the steps are as follows:
step 2.1, converting the captured face image into a gray image, and calculating a gray histogram:
wherein N is γ The number of pixels with gamma is the gray scale, and N is the imageTotal number of elements, gray value of gamma i P for frequency of (2) γ (γ i ) Representing, and the gray level number of the image is represented by k;
step 2.2, calculating a cumulative distribution function based on the gray level histogram:
where n is the total number of pixels in the frame, k is the number of gray levels corresponding to the image, k=0, 1,2, …, L-1.L is the total number of gray levels, n i The number of pixels of the current gray level;
step 2.3, carrying out histogram equalization. For each pixel value, mapping the pixel value of the original image by using a cumulative distribution function to obtain an equalized pixel value.
Step 3: the method adopts a Local Binary Pattern (LBP) to extract the face characteristics, and comprises the following specific steps:
step 3.1, calculate the LBP value for each pixel. Comparing each pixel point in the image with surrounding pixels, comparing the gray values of the surrounding pixels with the gray values of the central pixels, and encoding the result into binary numbers, wherein finally, a binary number, namely LBP code, can be obtained for each pixel;
step 3.2, calculating LBP characteristic vector. And counting the LBP codes of all pixels to obtain an LBP characteristic vector.
Step 4: the feature dimension reduction processing is carried out by using a Principal Component Analysis (PCA) technology, and the specific steps are as follows:
step 4.1, calculating the mean value of the feature matrix, and carrying out centering treatment:
where n is the number of samples, x j For each feature;
step 4.2, calculating a covariance matrix. Solving a covariance matrix C for the standardized data matrix:
wherein n is the number of samples, and X' is the standardized data matrix;
step 4.3, calculating the eigenvalues and eigenvectors and selecting the principal component quantity. Performing eigenvalue decomposition on the covariance matrix to obtain eigenvalues and corresponding eigenvectors, and selecting the number k of principal components to be reserved according to the magnitude of the eigenvalues;
step 4.4, constructing a projection matrix and projecting data. The first k eigenvectors with the largest eigenvalues are selected to form a projection matrix W, and the projection matrix W is used for projecting the standardized data matrix X' into a new k-dimensional space:
X n ＝X′·W
wherein X is n Is a data matrix in the new k-dimensional space.
Step 5: the optimal pitch bayesian classification model consists of four steps. The method comprises the following specific steps:
step 5.1, calculation of prior probability and conditional probability of Gaussian naive Bayes:
wherein y is i Representing the category of the ith data in the training data set, y representing the category label, n representing the number of samples of the training data, s representing the total category number, x j Represents the j-th attribute, x ij Is the j-th attribute, μ of the i-th sample j Is the mean value of the j-th attribute,is the variance of the jth attribute;
step 5.2, defining the fitness evaluation function. And in the optimal interval Bayesian classification model, the proposed optimal interval measurement and the accuracy of the training set are used as fitness evaluation functions. The optimal spacing is defined as the difference between the joint probabilities of all samples of one class and all samples of another class:
for the multi-classification problem, a one-to-many strategy is adopted to construct a multi-classification model;
and 5.3, training and optimizing the Bayesian classification model with the optimal distance. And (3) maximizing the optimal distance and the accuracy of the training set by adopting a multi-target particle swarm algorithm, and outputting a Pareto solution set through iterative updating to obtain a plurality of groups of attribute weight vectors. Evaluating the obtained multiple groups of attribute weights by using cross verification, and finally obtaining an optimal global attribute weight vector W;
step 5.4, calculating the posterior probability of each instance with weight under each category by using the attribute weight vector and the class conditional probability, and predicting the category to which the data belongs by maximizing the posterior probability:
wherein Y represents class labels to which the data belongs, Y represents a set of all class labels in the training data set, n represents the number of all attributes in the data set, and w i Indicating the weight given to the ith attribute.
Step 6: and outputting a face recognition result.
And classifying the target face by adopting an optimal interval Bayesian classification model, and taking the face with the highest probability as a final recognition result.
Advantages and beneficial effects of the invention
The invention provides a face recognition method based on an optimal spacing Bayesian classification model. The boundaries between different categories can be effectively increased, and the accurate recognition capability of the classifier on different face features is improved. Meanwhile, the invention can complete the capture of the face image under various conditions of changing expression, posture and illumination, and adopts a histogram equalization technology to improve the image quality. In the aspect of feature extraction, a local binary pattern is adopted to extract more robust and distinguishable facial features, and dimension disasters are avoided through PCA. And finally, classifying by adopting an optimal interval Bayesian classification method, so that the performance of face recognition classification is obviously improved, and obvious improvement is brought to the aspects of safety, user experience and the like.
Drawings
FIG. 1 is a general flow chart of an implementation of the face recognition method of the present invention;
FIG. 2 is an overall flow diagram of an optimal pitch Bayesian classification model;
FIG. 3 is a flow chart of optimizing attribute weights for multiple target particle swarms;
Detailed Description
The invention is described in detail below with reference to the accompanying drawings and examples.
A face recognition method based on an optimal interval Bayesian classification model is implemented with the general flow shown in figure 1; according to the invention, face image data of 8 persons are collected, each person has 26 images, and training and testing are carried out by using an optimal interval Bayesian classification model. The specific implementation method comprises the following steps:
step 1, capturing face images in real time by using a camera, wherein the acquired images are stored in a designated folder and are used in the subsequent face feature extraction and test model training process;
step 2, carrying out gray processing on the obtained face image, and carrying out histogram equalization processing on the gray face image;
step 3, extracting face features of the grey face image by adopting a local binary pattern algorithm;
step 4, performing characteristic dimension reduction processing by using a principal component analysis technology;
step 5, training and identifying face data by using an optimal interval Bayesian classification model, wherein a specific flow chart is shown in fig. 2, and a multi-target particle swarm optimization attribute weight flow chart is shown in fig. 3;
and step 6, outputting a face recognition result. Taking the recognition result of the 8 # face image as an example, the comparison result is that the probability of being similar to the 1 # face is 0, the probability of being similar to the 2 # face is 0, the probability of being similar to the 3 # face is 0, the probability of being similar to the 4 # face is 0, the probability of being similar to the 5 # face is 0, the probability of being similar to the 6 # face is 0, the probability of being similar to the 7 # face is 0, and the probability of being similar to the 8 # face is a percentage, so the final recognition result is the 8 # face.
The optimal interval Bayesian classification model provided by the invention is repeatedly tested for a plurality of times by using the UCI standard data set, and is compared with the traditional algorithm. The 6 data sets Inonosphere, glass, spambase, kr-vs-kp, blood, rice were selected. The experimental results are shown in graph 1.
Table 1 comparison of experimental results (Classification accuracy)
| KNN | Traditional Bayes algorithm | Optimum spacing Bayes | |
| Ionosphere | 86.79％ | 83.81％ | 90.48％ |
| Glass | 66.155 | 51.16％ | 76.92％ |
| Spambase | 79.65％ | 80.94％ | 91.23％ |
| kr-vs-kp | 63.98％ | 60.17％ | 73.93％ |
| Blood | 72.00％ | 74.11％ | 80.00％ |
| Rice | 88.45％ | 91.69％ | 92.40％ |
The optimal interval Bayesian classification model provided by the invention is repeatedly tested for a plurality of times by using the ORL face database, and is compared with other common face recognition methods. The ORL face database consists of face images of 40 persons, researchers acquire face images of each person under different environments such as facial expressions, illumination conditions and the like, and each person acquires 10 photos under different environments. We performed experiments on the ORL face database and compared with algorithms such as EDLPP, LDENP, 2DJLNDA, OR-NMF, pca+svm, etc. The experimental results are shown in graph 2.
Table 2 ORL face database experimental results
The above description is only a preferred example of the present invention and is not intended to limit the present invention, and various modifications and variations of the present invention will be apparent to those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention should be included in the scope of the claims of the present invention.
Claims (7)
1. The face recognition method based on the optimal interval Bayesian classification model is characterized by comprising the following steps of:
step 1, capturing face images in real time through a camera or image acquisition equipment;
step 2, preprocessing the captured face image, and eliminating uneven illumination or other image quality problems by using a histogram equalization technology;
step 3, extracting the most representative and robust features in the face image by adopting a Local Binary Pattern (LBP) feature extraction method;
step 4, performing feature dimension reduction processing by using a Principal Component Analysis (PCA) technology to reduce feature dimensions, improve calculation efficiency and reserve main feature information;
step 5, training and identifying classification by using an optimal interval Bayesian classification model;
and step 6, outputting a face recognition result.
2. The face recognition method based on the optimal pitch bayesian classification model according to claim 1, wherein: the step 1 capturing a face image in real time through a camera, and using a Haar classifier, wherein the specific steps are as follows:
step 1.1, calculating Haar characteristics, and using pixels of a difference sum of adjacent rectangles at a specified position of a detection range;
step 1.2, accelerating to obtain Haar characteristics, and accelerating to calculate the Haar characteristics by using an integral graph:
O sum ＝I(A)+I(B)+I(C)+I(D)
where O represents the detection window, A, B, C, D represents four vertices of the detection window, and I (x, y) represents the pixel value of the coordinate point;
step 1.3, normalizing the Haar characteristic values, and normalizing the characteristic values to ensure that the characteristic values have uniform scale or range. The method comprises the following steps:
step 1.3.1, calculating the gray value and the square sum of the gray values of the images in the detection window:
sum＝∑I(x,y)
sq sum ＝∑I 2 (x,y)
step 1.3.2, calculating the average value:
step 1.3.3, normalization factor:
step 1.3.4, normalized eigenvalues:
finally, comparing the normalized characteristic value with a threshold value;
step 1.4, the training of the AdaBoost algorithm is used for distinguishing the human face from the non-human face, and the training of the AdaBoost algorithm comprises the following steps:
step 1.4.1, initializing sample weights. Assigning an equal weight value to each sample;
step 1.4.2, iteratively training a plurality of weak classifiers. Training a weak classifier, and classifying the data according to the current sample weight learning. And then, the error rate of the classifier on the weighted data set is calculated, the weight of the sample is adjusted, the weight of the error classification sample is improved, and the weight of the correct classification sample is reduced. Finally, calculating the weight of the weak classifier according to the classification accuracy of the weak classifier under the current weight, and generating a plurality of weak classifiers by continuously iterating the process;
and 1.4.3, constructing an integrated strong classifier. And (3) weighting and combining all the weak classifiers according to the weights of the weak classifiers to construct an integrated strong classifier.
3. The face recognition method based on the optimal pitch bayesian classification model according to claim 1, wherein: the step 2 of preprocessing the captured face image by using a histogram equalization technology, specifically comprises the following steps:
step 2.1, converting the captured face image into a gray image, and calculating a gray histogram:
wherein N is γ The number of pixels with the gray scale of gamma is N, the total number of pixels of the image, and the gray scale value of gamma i P for frequency of (2) γ (γ i ) Representing, and the gray level number of the image is represented by k;
step 2.2, calculating a cumulative distribution function based on the gray level histogram:
where n is the total number of pixels in the frame, k is the number of gray levels corresponding to the image, k=0, 1,2, …, L-1.L is the total number of gray levels, n i The number of pixels of the current gray level;
step 2.3, carrying out histogram equalization. For each pixel value, mapping the pixel value of the original image by using a cumulative distribution function to obtain an equalized pixel value.
4. The face recognition method based on the optimal pitch bayesian classification model according to claim 1, wherein: the step 3 of extracting the face features by adopting a Local Binary Pattern (LBP) specifically comprises the following steps:
step 3.1, calculate the LBP value for each pixel. Comparing each pixel point in the image with surrounding pixels, comparing the gray values of the surrounding pixels with the gray values of the central pixels, and encoding the result into binary numbers, wherein finally, a binary number, namely LBP code, can be obtained for each pixel;
step 3.2, calculating LBP characteristic vector. And counting the LBP codes of all pixels to obtain an LBP characteristic vector.
5. The face recognition method based on the optimal pitch bayesian classification model according to claim 1, wherein: the feature dimension reduction processing is performed by using Principal Component Analysis (PCA) technology, and the specific steps are as follows:
step 4.1, calculating the mean value of the feature matrix, and carrying out centering treatment:
where n is the number of samples, x j For each feature;
step 4.2, calculating a covariance matrix. Solving a covariance matrix C for the standardized data matrix:
wherein n is the number of samples, and X' is the standardized data matrix;
step 4.3, calculating the eigenvalues and eigenvectors and selecting the principal component quantity. Performing eigenvalue decomposition on the covariance matrix to obtain eigenvalues and corresponding eigenvectors, and selecting the number k of principal components to be reserved according to the magnitude of the eigenvalues;
step 4.4, constructing a projection matrix and projecting data. The first k eigenvectors with the largest eigenvalues are selected to form a projection matrix W, and the projection matrix W is used for projecting the standardized data matrix X' into a new k-dimensional space:
X n ＝X′·W
wherein X is n Is a data matrix in the new k-dimensional space.
6. The face recognition method based on the optimal pitch bayesian classification model according to claim 1, wherein: the optimal interval Bayesian classification model in the step 5 is trained and identified for classification, and the specific steps are as follows:
step 5.1, calculation of prior probability and conditional probability of Gaussian naive Bayes:
wherein y is i Representing the category of the ith data in the training data set, y representing the category label, n representing the number of samples of the training data, s representing the total category number, x j Represents the j-th attribute, x ij Is the j-th attribute, μ of the i-th sample j Is the mean value of the j-th attribute,is the variance of the jth attribute;
step 5.2, defining the fitness evaluation function. And in the optimal interval Bayesian classification model, the proposed optimal interval measurement and the accuracy of the training set are used as fitness evaluation functions. The optimal spacing is defined as the difference between the joint probabilities of all samples of one class and all samples of another class:
for the multi-classification problem, a one-to-many strategy is adopted to construct a multi-classification model;
and 5.3, training and optimizing the Bayesian classification model with the optimal distance. And (3) maximizing the optimal distance and the accuracy of the training set by adopting a multi-target particle swarm algorithm, and outputting a Pareto solution set through iterative updating to obtain a plurality of groups of attribute weight vectors. Evaluating the obtained multiple groups of attribute weights by using cross verification, and finally obtaining an optimal global attribute weight vector W;
step 5.4, calculating the posterior probability of each instance with weight under each category by using the attribute weight vector and the class conditional probability, and predicting the category to which the data belongs by maximizing the posterior probability:
wherein Y represents class labels to which the data belongs, Y represents a set of all class labels in the training data set, n represents the number of all attributes in the data set, and w i Indicating the weight given to the ith attribute.
7. The face recognition method based on the optimal pitch bayesian classification model according to claim 1, wherein: the method for outputting the face recognition result in the step 6 is as follows: and classifying the target face by adopting an optimal interval Bayesian classification model, and taking the face with the highest probability as a final recognition result.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311633669.XA CN117523642A (en) | 2023-12-01 | 2023-12-01 | Face recognition method based on optimal-spacing Bayesian classification model |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311633669.XA CN117523642A (en) | 2023-12-01 | 2023-12-01 | Face recognition method based on optimal-spacing Bayesian classification model |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117523642A true CN117523642A (en) | 2024-02-06 |
Family
ID=89756617
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311633669.XA Pending CN117523642A (en) | 2023-12-01 | 2023-12-01 | Face recognition method based on optimal-spacing Bayesian classification model |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117523642A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117932474A (en) * | 2024-03-22 | 2024-04-26 | 山东核电有限公司 | Training method, device, equipment and storage medium of communication missing data determination model |
-
2023
- 2023-12-01 CN CN202311633669.XA patent/CN117523642A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117932474A (en) * | 2024-03-22 | 2024-04-26 | 山东核电有限公司 | Training method, device, equipment and storage medium of communication missing data determination model |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| JP4543423B2 (en) | Method and apparatus for automatic object recognition and collation | |
| CN106295124B (en) | The method of a variety of image detecting technique comprehensive analysis gene subgraph likelihood probability amounts | |
| Stuhlsatz et al. | Feature extraction with deep neural networks by a generalized discriminant analysis | |
| CN110659665B (en) | Model construction method of different-dimension characteristics and image recognition method and device | |
| CN106529504B (en) | A kind of bimodal video feeling recognition methods of compound space-time characteristic | |
| CN113239839B (en) | Expression recognition method based on DCA face feature fusion | |
| CN108108760A (en) | A kind of fast human face recognition | |
| CN117523642A (en) | Face recognition method based on optimal-spacing Bayesian classification model | |
| Mu et al. | Hyperspectral image classification based on active learning and spectral-spatial feature fusion using spatial coordinates | |
| Zuobin et al. | Feature regrouping for cca-based feature fusion and extraction through normalized cut | |
| CN112132117A (en) | Fusion identity authentication system assisting coercion detection | |
| CN112560710A (en) | Method for constructing finger vein recognition system and finger vein recognition system | |
| CN109948662B (en) | Face image depth clustering method based on K-means and MMD | |
| JP6166981B2 (en) | Facial expression analyzer and facial expression analysis program | |
| CN109002770B (en) | Face recognition method under low-resolution condition | |
| CN112395901A (en) | Improved face detection, positioning and recognition method in complex environment | |
| CN106709442B (en) | Face recognition method | |
| WO2009047561A1 (en) | Value determination | |
| CN115310491A (en) | Class-imbalance magnetic resonance whole brain data classification method based on deep learning | |
| Türkmen et al. | Global feature based female facial beauty decision system | |
| CN112241680A (en) | Multi-mode identity authentication method based on vein similar image knowledge migration network | |
| CN107341485B (en) | Face recognition method and device | |
| CN111860529A (en) | Image preprocessing method, system, device and medium | |
| CN110751023B (en) | Series pedestrian detection method for video stream | |
| CN114693964B (en) | MRI data feature extraction and classification identification method based on artificial neural network |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |