CN106022223A - High-dimensional local-binary-pattern face identification algorithm and system - Google Patents
High-dimensional local-binary-pattern face identification algorithm and system Download PDFInfo
- Publication number
- CN106022223A CN106022223A CN201610305175.2A CN201610305175A CN106022223A CN 106022223 A CN106022223 A CN 106022223A CN 201610305175 A CN201610305175 A CN 201610305175A CN 106022223 A CN106022223 A CN 106022223A
- Authority
- CN
- China
- Prior art keywords
- image
- characteristic
- feature
- gray
- gray level
- 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.)
- Granted
Links
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
-
- 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
- G06V40/168—Feature extraction; Face representation
Landscapes
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Oral & Maxillofacial Surgery (AREA)
- General Health & Medical Sciences (AREA)
- Human Computer Interaction (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
The invention discloses a high-dimensional local-binary-pattern face identification algorithm and system. The algorithm is implemented by the following steps: S1, a face image is obtained and pretreatment is carried out on the face image to obtain a grayscale image having a same size; S2, HDLBP feature extraction is carried out on the grayscale image after pretreatment, thereby obtaining a corresponding feature image; S3, a histogram of the feature image is extracted to obtain a corresponding feature vector; and S4, according to the feature vector, comparison with information in a feature database is carried out to obtain an identification result. According to the invention, a local feature and a global feature of an image are extracted, so that the identification rate of the algorithm is improved substantially; and the accuracy of the image identification is increased on the premise that the algorithm complexity is low.
Description
Technical field
The present invention relates to field of face identification, particularly relate to a kind of higher-dimension local binary patterns face identification method and be
System.
Background technology
Local binary patterns (Local Binary pattern, LBP) algorithm as the one of face recognition algorithms, be by
A kind of algorithm depending on local grain description that Ojala, Ahonen proposed equal to 1996, is used for describing pixel adjacent with it
Pixel relation numerically in territory, because its computational methods are simple and clear, has the most descriptive, to light to the local feature of image
According to the feature such as insensitivity and used widely in field of face identification.Simultaneously as LBP describes son has only focused on figure
The description of the local feature of picture, have ignored the description to image overall feature, result in LBP algorithm on global characteristics extracts
Not enough.In order to effectively solve this problem, it is studied by numerous scholars, and proposes improvement and the optimization of many
Method.
Zhou Xi etc. propose the method that piecemeal processes, and describe son to LBP to be solved and are extracting not enough asking on global characteristics
Topic.Dividing the fast core concept processed is according to point artwork such as certain sizes, or divides according to the position at the face place of people
Artwork, extracts LBP feature respectively, all of characteristic vector is concatenated together subgraph, obtains some characteristics in the overall situation, logical
Cross the performance of LBP algorithm after experiment proves point fast process be better than untreated before LBP algorithm, but to artwork according to one
But neither one answer certainly on this point which type of standard carrying out dividing.Can not obtain very in the process to artwork
Good effect, then Wang Hong etc. describe sub itself research from LBP, and its thinking is to describe son by the amplification LBP of equimultiple, makes
It can extract characteristics of image in larger scope.Compare is no longer some pixel, but includes certain picture
The average of vegetarian refreshments pixel in interior fixing neighborhood calculates, it is ensured that local characteristics embodies the most to a certain extent
Global property, but the selection of Size of Neighborhood is become again to the emphasis of research.It is to describe son from BLP equally,
Being different from the principle that equimultiple is amplified, Wang Cheng etc. is special by the feature instantiation overall situation in the multiple dimensioned weighted center multi-level neighborhood of point
Levying, the some distance center point in neighborhood is the nearest, and weights when being weighted are the heaviest, and the level of weighting is the most, the overall situation of embodiment
Property is the best, and computation complexity is the highest simultaneously.
Summary of the invention
The technical problem to be solved in the present invention is for the defect that can not extract global characteristics in prior art, it is provided that one
Plant higher-dimension local binary patterns face identification method and the system that can be greatly improved algorithm discrimination.
The technical solution adopted for the present invention to solve the technical problems is:
The present invention provides a kind of higher-dimension local binary patterns face recognition algorithms, comprises the following steps:
S1, obtain facial image, and it is carried out pretreatment obtain the gray level image of same size;
S2, pretreated gray level image is carried out HDLBP feature extraction, obtain characteristic of correspondence image;
S3, the rectangular histogram of extraction characteristic image, obtain characteristic of correspondence vector;
S4, compare with the information in property data base according to characteristic vector, be identified result.
Further, in step S1 of the present invention pretreatment obtain gray level image method particularly as follows:
If the local grain V of face gray level image is distributed as:
V=v (gc g0 … gp-1 g)
Wherein, gcRepresent the central threshold of window, gi(i=0,2...p-1) represents the gray value of each field pixel, p table
Showing field point number, g represents the gray average of face gray level image, and computing formula is:
Wherein, m × n is the size of gray level image, and g (i j) is the gray value of each pixel in image.
Further, step S2 of the present invention carries out the method for HDLBP feature extraction particularly as follows:
HDLBP describes sub when calculating, and the LBP first having adopted classics describes son computational methods in window, it is ensured that
Local feature;Then the gray average of window center pixel gray value and face gray level image is used same calculating side
Method, it is ensured that global characteristics;Finally local feature is combined as low-dimensional, global characteristics as higher-dimension, the result of calculating
It it is exactly the eigenvalue of central pixel point in this window.
Further, step S2 of the present invention carries out method that higher-dimension and low-dimensional merge particularly as follows:
Central feature is joined in the binary sequence of edge feature as the component of higher-dimension so that characteristic sequence to
High one-dimensional stretching, extension, expands the quantity of information that characteristic sequence comprises;According to equation below, low dimensional feature and high dimensional feature are merged one
Rise, make two row characteristic sequences become string characteristic sequence, and turn metric method according to binary system and carry out calculating the most available
Characteristic of correspondence value;Computing formula is:
Wherein, gcRepresent the central threshold of window, gi(i=0,2...p-1) represents the gray value of each field pixel, p table
Showing field point number, g represents the gray average of face gray level image, and computing formula is:
Wherein, m × n is the size of gray level image, and g (i j) is the gray value of each pixel in image.
Further, the formula carrying out HDLBP feature extraction in step S2 of the present invention is:
Wherein, s function is as follows:
Further, step S3 of the present invention is extracted the histogrammic method of characteristic image particularly as follows:
According to the characteristic image of input, pixel all of in image is carried out ascending order row according to the size of its gray value
Sequence, then statistics has the number of times that the pixel of same grayscale value occurs, obtains n*1 sequence, i.e. characteristic vector, wherein a n
Representing the number of different gray values in characteristic image, histogrammic computing formula is as follows:
H (i)=gi i∈(1,n)
Wherein h (i) represents that gray value is giThe number of pixel.
Further, step S4 of the present invention draws the method for recognition result particularly as follows:
Compare with the information in property data base according to characteristic vector, with Euclidean distance for weighing, utilize closest
Classification method is identified.
The present invention provides a kind of higher-dimension local binary patterns face identification system, including:
Image pre-processing unit, is used for obtaining facial image, and it is carried out pretreatment obtains the gray-scale map of same size
Picture;
HDLBP feature extraction unit, for pretreated gray level image is carried out HDLBP feature extraction, obtains correspondence
Characteristic image;
Characteristic vector pickup unit, for extracting the rectangular histogram of characteristic image, obtains characteristic of correspondence vector;
Image identification unit, for comparing with the information in property data base according to characteristic vector, is identified knot
Really.
The beneficial effect comprise that: the higher-dimension local binary patterns face recognition algorithms of the present invention, by adopting
Classical LBP describes son computational methods in window, it is ensured that local feature;Then to window center pixel gray value and
The gray average of face gray level image uses same computational methods, it is ensured that global characteristics;Finally using local feature as low-dimensional,
Global characteristics combines as higher-dimension, as the eigenvalue of central pixel point in this window;The algorithm of the present invention can carry
Take local feature and the global characteristics of image, the discrimination of algorithm is greatly improved;And at the base ensureing that algorithm complex is the highest
On plinth, add the accuracy rate of image recognition.
Accompanying drawing explanation
Below in conjunction with drawings and Examples, the invention will be further described, in accompanying drawing:
Fig. 1 is the flow chart of the higher-dimension local binary patterns face recognition algorithms of the embodiment of the present invention;
Fig. 2 is the fusion process of the higher-dimension local binary patterns face recognition algorithms of the embodiment of the present invention;
Fig. 3 is the block diagram of the higher-dimension local binary patterns face recognition algorithms of the embodiment of the present invention;
Fig. 4 is the calculating procedure chart of the higher-dimension local binary patterns face recognition algorithms of the embodiment of the present invention;
Fig. 5 is the block diagram of the higher-dimension local binary patterns face identification system of the embodiment of the present invention.
Detailed description of the invention
In order to make the purpose of the present invention, technical scheme and advantage clearer, below in conjunction with drawings and Examples, right
The present invention is further elaborated.Should be appreciated that specific embodiment described herein only in order to explain the present invention, not
For limiting the present invention.
As it is shown in figure 1, the higher-dimension local binary patterns face recognition algorithms of the embodiment of the present invention, comprise the following steps:
S1, obtain facial image, and it is carried out pretreatment obtain the gray level image of same size;Pretreatment obtains gray scale
The method of image particularly as follows:
If the local grain V of face gray level image is distributed as:
V=v (gc g0 … gp-1 g)
Wherein, gcRepresent the central threshold of window, gi(i=0,2...p-1) represents the gray value of each field pixel, p table
Showing field point number, g represents the gray average of face gray level image, and computing formula is:
Wherein, m × n is the size of gray level image, and g (i j) is the gray value of each pixel in image.
S2, pretreated gray level image is carried out HDLBP feature extraction, obtain characteristic of correspondence image;Carry out HDLBP
The formula of feature extraction is:
Wherein, s function is as follows:
S3, the rectangular histogram of extraction characteristic image, obtain characteristic of correspondence vector;
S4, compare with the information in property data base according to characteristic vector, be identified result.
HDLBP describes sub when calculating, and the LBP first having adopted classics describes son computational methods in window, it is ensured that
Local feature;Then the gray average of window center pixel gray value and face gray level image is used same calculating side
Method, it is ensured that global characteristics;Finally local feature is combined as low-dimensional, global characteristics as higher-dimension, the result of calculating
It it is exactly the eigenvalue of central pixel point in this window.
As in figure 2 it is shown, merge process particularly as follows:
Central feature joins as the component of higher-dimension in the binary sequence of edge feature so that characteristic sequence is to height
One-dimensional stretching, extension, the quantity of information that characteristic sequence comprises is bigger.Owing to low dimensional feature and high dimensional feature merge so that Liang Liete
Levy sequence and become string characteristic sequence, when therefore characteristic sequence being changed into eigenvalue, it is only necessary to turn decimal scale according to binary system
Method carry out calculating and i.e. can get characteristic of correspondence value, computing formula is:
As it is shown on figure 3, in another specific embodiment of the present invention,
The concrete calculating process of algorithm is as follows:
The first step: input facial image to be identified;
Second step: the face picture of input is carried out pretreatment, obtains the gray level image of same size;
3rd step: extract the HDLBP feature of facial image to be identified according to the following formula, obtains characteristic of correspondence image;
According to the gray level image of input, the formula being calculated as follows g calculates the value of g, then utilizes HDLBP feature
Computing formula to gray level image from left to right, be scanned from top to bottom, for i each time, the value of j has a correspondence
Eigenvalue, the image being made up of eigenvalue is i.e. characteristic image.
4th step: extract the rectangular histogram of characteristic image, obtains characteristic of correspondence vector;
According to the characteristic image of input, pixel all of in image is carried out ascending order row according to the size of its gray value
Sequence, then statistics has the number of times that the pixel of same grayscale value occurs, obtains n*1 sequence, i.e. characteristic vector, wherein a n
Representing the number of different gray values in characteristic image, histogrammic computing formula is as follows:
H (i)=gi i∈(1,n)
Wherein h (i) represents that gray value is giThe number of pixel.
5th step: on the property data base set up, with Euclidean distance for weighing, utilizes nearest neighbour classification to know
Not;
Euclidean distance computing formula in n-dimensional space is as follows:
Wherein x1kAnd x2kIt is n-dimensional vector x respectively1And x2Kth be component.Nearest neighbor classification: by data to be sorted, press
Range is from the apoplexy due to endogenous wind being divided into the sample place closest with it.
The formula utilizing Euclidean distance calculates characteristic vector and the Euclidean distance of all samples in property data base, wherein
The sample with minimum range is i.e. the result identified.Wherein, dmIt it is characteristic vector xkWith m-th sample y in data basekEurope
Family name's distance.
6th step: the result that output identifies.
As shown in Figure 4, utilizing ORL face database and YALE face database to test, ORL face database is by Cambridge University's media
AT&T establishment of laboratory.This storehouse includes 40 class faces, every class 10 width, and size is the gray level image of 112 × 92.All of image
Having similar dark background, the different images of same people is in different time, different illumination, different head attitude, different expression
With shooting under different details;YALE face database is the face database that Yale sets up, and this storehouse comprises
15 people, everyone 11 width, including different expressions, different direction of illumination and variations in detail etc..It is divided into test by unified for face database
Sample and experiment sample, for ORL face database, test sample: experiment sample=5:5;To YALE face database, test sample: real
Test sample=6:5.
Experimental result is as shown in the table.
Therefore: whether at YALE face database still on ORL face database, the knowledge of higher-dimension local binary patterns algorithm
Not rate will be higher than tradition local binary patterns.
The shortcoming that can not extract global characteristics for the face recognition algorithms of tradition local binary patterns, it is proposed that Yi Zhongji
Face recognition algorithms in higher-dimension local binary patterns.This algorithm, relative to traditional local binary patterns algorithm, greatly carries
The high discrimination of algorithm.
As it is shown in figure 5, the higher-dimension local binary patterns face identification system of the embodiment of the present invention, it is used for realizing the present invention real
Execute the higher-dimension local binary patterns face recognition algorithms of example, including:
Image pre-processing unit, is used for obtaining facial image, and it is carried out pretreatment obtains the gray-scale map of same size
Picture;
HDLBP feature extraction unit, for pretreated gray level image is carried out HDLBP feature extraction, obtains correspondence
Characteristic image;
Characteristic vector pickup unit, for extracting the rectangular histogram of characteristic image, obtains characteristic of correspondence vector;
Image identification unit, for comparing with the information in property data base according to characteristic vector, is identified knot
Really.
It should be appreciated that for those of ordinary skills, can be improved according to the above description or be converted,
And all these modifications and variations all should belong to the protection domain of claims of the present invention.
Claims (8)
1. a higher-dimension local binary patterns face recognition algorithms, it is characterised in that comprise the following steps:
S1, obtain facial image, and it is carried out pretreatment obtain the gray level image of same size;
S2, pretreated gray level image is carried out HDLBP feature extraction, obtain characteristic of correspondence image;
S3, the rectangular histogram of extraction characteristic image, obtain characteristic of correspondence vector;
S4, compare with the information in property data base according to characteristic vector, be identified result.
Higher-dimension local binary patterns face recognition algorithms the most according to claim 1, it is characterised in that locate in advance in step S1
Reason obtain gray level image method particularly as follows:
If the local grain V of face gray level image is distributed as:
V=v (gc g0 … gp-1 g)
Wherein, gcRepresent the central threshold of window, gi(i=0,2...p-1) represents the gray value of each field pixel, and p represents neck
Territory point number, g represents the gray average of face gray level image, and computing formula is:
Wherein, m × n is the size of gray level image, and g (i j) is the gray value of each pixel in image.
Higher-dimension local binary patterns face recognition algorithms the most according to claim 1, it is characterised in that carry out in step S2
The method of HDLBP feature extraction particularly as follows:
HDLBP describes sub when calculating, and the LBP first having adopted classics describes son computational methods in window, it is ensured that office
Portion's feature;Then the gray average of window center pixel gray value and face gray level image is used same computational methods,
Ensure global characteristics;Finally being combined as low-dimensional, global characteristics as higher-dimension by local feature, the result of calculating is exactly
The eigenvalue of central pixel point in this window.
Higher-dimension local binary patterns face recognition algorithms the most according to claim 3, it is characterised in that carry out in step S2
Method that higher-dimension and low-dimensional merge particularly as follows:
Central feature is joined as the component of higher-dimension in the binary sequence of edge feature so that characteristic sequence is to high by one
Dimension stretches, and expands the quantity of information that characteristic sequence comprises;According to equation below, low dimensional feature and high dimensional feature are merged, make
Two row characteristic sequences become string characteristic sequence, and turn metric method according to binary system and carry out calculating i.e. available corresponding
Eigenvalue;Computing formula is:
Wherein, gcRepresent the central threshold of window, gi(i=0,2...p-1) represents the gray value of each field pixel, and p represents neck
Territory point number, g represents the gray average of face gray level image, and computing formula is:
Wherein, m × n is the size of gray level image, and g (i j) is the gray value of each pixel in image.
Higher-dimension local binary patterns face recognition algorithms the most according to claim 2, it is characterised in that carry out in step S2
The formula of HDLBP feature extraction is:
Wherein, s function is as follows:
Higher-dimension local binary patterns face recognition algorithms the most according to claim 1, it is characterised in that extract in step S3
The histogrammic method of characteristic image particularly as follows:
According to the characteristic image of input, pixel all of in image is carried out ascending sort according to the size of its gray value, so
Rear statistics has the number of times that the pixel of same grayscale value occurs, obtains n*1 sequence, i.e. a characteristic vector, and wherein n represents special
Levying the number of different gray values in image, histogrammic computing formula is as follows:
H (i)=gi i∈(1,n)
Wherein h (i) represents that gray value is giThe number of pixel.
Higher-dimension local binary patterns face recognition algorithms the most according to claim 1, it is characterised in that draw in step S4
The method of recognition result particularly as follows:
Compare with the information in property data base according to characteristic vector, with Euclidean distance for weighing, utilize nearest neighbour classification
Method is identified.
8. a higher-dimension local binary patterns face identification system, it is characterised in that including:
Image pre-processing unit, is used for obtaining facial image, and it is carried out pretreatment obtains the gray level image of same size;
HDLBP feature extraction unit, for pretreated gray level image is carried out HDLBP feature extraction, obtains the spy of correspondence
Levy image;
Characteristic vector pickup unit, for extracting the rectangular histogram of characteristic image, obtains characteristic of correspondence vector;
Image identification unit, for comparing with the information in property data base according to characteristic vector, is identified result.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610305175.2A CN106022223B (en) | 2016-05-10 | 2016-05-10 | A kind of higher-dimension local binary patterns face identification method and system |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610305175.2A CN106022223B (en) | 2016-05-10 | 2016-05-10 | A kind of higher-dimension local binary patterns face identification method and system |
Publications (2)
Publication Number | Publication Date |
---|---|
CN106022223A true CN106022223A (en) | 2016-10-12 |
CN106022223B CN106022223B (en) | 2019-10-25 |
Family
ID=57100152
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610305175.2A Expired - Fee Related CN106022223B (en) | 2016-05-10 | 2016-05-10 | A kind of higher-dimension local binary patterns face identification method and system |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106022223B (en) |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110309692A (en) * | 2018-03-27 | 2019-10-08 | 杭州海康威视数字技术股份有限公司 | Face identification method, apparatus and system, model training method and device |
CN110543853A (en) * | 2019-09-04 | 2019-12-06 | 上海观安信息技术股份有限公司 | Method for realizing face recognition processing based on image processing technology |
CN110956090A (en) * | 2019-11-04 | 2020-04-03 | 南京邮电大学 | Face feature extraction method based on positive and negative illumination invariant unit fusion |
CN111709312A (en) * | 2020-05-26 | 2020-09-25 | 上海海事大学 | Local feature face recognition method based on joint main mode |
CN111832639A (en) * | 2020-06-30 | 2020-10-27 | 山西大学 | Drawing emotion prediction method based on transfer learning |
CN116719966A (en) * | 2023-05-25 | 2023-09-08 | 中国医学科学院皮肤病医院(中国医学科学院皮肤病研究所) | Hospital patient information management system |
CN117373100A (en) * | 2023-12-08 | 2024-01-09 | 成都乐超人科技有限公司 | Face recognition method and system based on differential quantization local binary pattern |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20080166026A1 (en) * | 2007-01-10 | 2008-07-10 | Samsung Electronics Co., Ltd. | Method and apparatus for generating face descriptor using extended local binary patterns, and method and apparatus for face recognition using extended local binary patterns |
CN101447021A (en) * | 2008-12-30 | 2009-06-03 | 爱德威软件开发(上海)有限公司 | Face fast recognition system and recognition method thereof |
CN102663400A (en) * | 2012-04-16 | 2012-09-12 | 北京博研新创数码科技有限公司 | LBP (length between perpendiculars) characteristic extraction method combined with preprocessing |
US20140212044A1 (en) * | 2012-09-26 | 2014-07-31 | Carnegie Mellon University | Image Matching Using Subspace-Based Discrete Transform Encoded Local Binary Patterns |
CN104318219A (en) * | 2014-10-31 | 2015-01-28 | 上海交通大学 | Face recognition method based on combination of local features and global features |
CN105550658A (en) * | 2015-12-24 | 2016-05-04 | 蔡叶荷 | Face comparison method based on high-dimensional LBP (Local Binary Patterns) and convolutional neural network feature fusion |
-
2016
- 2016-05-10 CN CN201610305175.2A patent/CN106022223B/en not_active Expired - Fee Related
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20080166026A1 (en) * | 2007-01-10 | 2008-07-10 | Samsung Electronics Co., Ltd. | Method and apparatus for generating face descriptor using extended local binary patterns, and method and apparatus for face recognition using extended local binary patterns |
CN101447021A (en) * | 2008-12-30 | 2009-06-03 | 爱德威软件开发(上海)有限公司 | Face fast recognition system and recognition method thereof |
CN102663400A (en) * | 2012-04-16 | 2012-09-12 | 北京博研新创数码科技有限公司 | LBP (length between perpendiculars) characteristic extraction method combined with preprocessing |
US20140212044A1 (en) * | 2012-09-26 | 2014-07-31 | Carnegie Mellon University | Image Matching Using Subspace-Based Discrete Transform Encoded Local Binary Patterns |
CN104318219A (en) * | 2014-10-31 | 2015-01-28 | 上海交通大学 | Face recognition method based on combination of local features and global features |
CN105550658A (en) * | 2015-12-24 | 2016-05-04 | 蔡叶荷 | Face comparison method based on high-dimensional LBP (Local Binary Patterns) and convolutional neural network feature fusion |
Non-Patent Citations (1)
Title |
---|
DONG CHEN等: "Blessing of Dimensionality:High-dimensional Feature and Its Efficient Compression for Face Verification", 《2013 IEEE CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION》 * |
Cited By (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110309692A (en) * | 2018-03-27 | 2019-10-08 | 杭州海康威视数字技术股份有限公司 | Face identification method, apparatus and system, model training method and device |
CN110543853A (en) * | 2019-09-04 | 2019-12-06 | 上海观安信息技术股份有限公司 | Method for realizing face recognition processing based on image processing technology |
CN110956090A (en) * | 2019-11-04 | 2020-04-03 | 南京邮电大学 | Face feature extraction method based on positive and negative illumination invariant unit fusion |
CN110956090B (en) * | 2019-11-04 | 2022-08-30 | 南京邮电大学 | Face feature extraction method based on fusion of positive and negative illumination invariant units |
CN111709312A (en) * | 2020-05-26 | 2020-09-25 | 上海海事大学 | Local feature face recognition method based on joint main mode |
CN111709312B (en) * | 2020-05-26 | 2023-09-22 | 上海海事大学 | Local feature face recognition method based on combined main mode |
CN111832639A (en) * | 2020-06-30 | 2020-10-27 | 山西大学 | Drawing emotion prediction method based on transfer learning |
CN111832639B (en) * | 2020-06-30 | 2022-05-31 | 山西大学 | Drawing emotion prediction method based on transfer learning |
CN116719966A (en) * | 2023-05-25 | 2023-09-08 | 中国医学科学院皮肤病医院(中国医学科学院皮肤病研究所) | Hospital patient information management system |
CN116719966B (en) * | 2023-05-25 | 2024-03-08 | 中国医学科学院皮肤病医院(中国医学科学院皮肤病研究所) | Hospital patient information management system |
CN117373100A (en) * | 2023-12-08 | 2024-01-09 | 成都乐超人科技有限公司 | Face recognition method and system based on differential quantization local binary pattern |
CN117373100B (en) * | 2023-12-08 | 2024-02-23 | 成都乐超人科技有限公司 | Face recognition method and system based on differential quantization local binary pattern |
Also Published As
Publication number | Publication date |
---|---|
CN106022223B (en) | 2019-10-25 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108460356B (en) | Face image automatic processing system based on monitoring system | |
CN107256246B (en) | printed fabric image retrieval method based on convolutional neural network | |
CN106022223A (en) | High-dimensional local-binary-pattern face identification algorithm and system | |
CN107316036B (en) | Insect pest identification method based on cascade classifier | |
CN111767882A (en) | Multi-mode pedestrian detection method based on improved YOLO model | |
CN110929593B (en) | Real-time significance pedestrian detection method based on detail discrimination | |
Cao et al. | A low-cost pedestrian-detection system with a single optical camera | |
CN105718889B (en) | Based on GB (2D)2The face personal identification method of PCANet depth convolution model | |
CN112488229B (en) | Domain self-adaptive unsupervised target detection method based on feature separation and alignment | |
JP2008310796A (en) | Computer implemented method for constructing classifier from training data detecting moving object in test data using classifier | |
KR101178333B1 (en) | A method for detecting human using wavelet-based cs-lbp and a cascade of random forests | |
Cai et al. | Traffic sign recognition algorithm based on shape signature and dual-tree complex wavelet transform | |
CN109165658B (en) | Strong negative sample underwater target detection method based on fast-RCNN | |
Yuan et al. | A combined pedestrian detection method based on haar-like features and hog features | |
Han et al. | Multilinear supervised neighborhood embedding of a local descriptor tensor for scene/object recognition | |
Meng et al. | An extended HOG model: SCHOG for human hand detection | |
CN104050460B (en) | The pedestrian detection method of multiple features fusion | |
CA3056026A1 (en) | A method and system for tracking objects | |
CN112580480A (en) | Hyperspectral remote sensing image classification method and device | |
Minhas et al. | A fast recognition framework based on extreme learning machine using hybrid object information | |
Zhong et al. | Improving pedestrian attribute recognition with multi-scale spatial calibration | |
Qu et al. | Vehicle detection in satellite images by incorporating objectness and convolutional neural network | |
Yang et al. | Pedestrian detection in thermal infrared image using extreme learning machine | |
CN116580289A (en) | Fine granularity image recognition method based on attention | |
Jubair et al. | A simplified method for handwritten character recognition from document image |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant | ||
CF01 | Termination of patent right due to non-payment of annual fee | ||
CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20191025 Termination date: 20200510 |