CN109002799A - Face identification method - Google Patents
Face identification method Download PDFInfo
- Publication number
- CN109002799A CN109002799A CN201810796508.5A CN201810796508A CN109002799A CN 109002799 A CN109002799 A CN 109002799A CN 201810796508 A CN201810796508 A CN 201810796508A CN 109002799 A CN109002799 A CN 109002799A
- Authority
- CN
- China
- Prior art keywords
- human face
- region
- rotation process
- face region
- textural characteristics
- 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
- G06V40/168—Feature extraction; Face representation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- 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
- G06V10/44—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
-
- 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
- G06V40/161—Detection; Localisation; Normalisation
Abstract
The invention discloses a kind of face identification methods, comprising: obtains images to be recognized；According to edge detection algorithm, the first human face region is obtained, according to benchmark rgb value range, obtains the second human face region, the region that the first human face region is overlapped with the second human face region is as third human face region；Characteristic area is obtained from third human face region, a rectangular area is determined with the center of characteristic area, rectangular area is successively rotated 0~180 ° around center, after each rotation process, rectangular area is divided into multiple zonules, extracts the textural characteristics of each zonule respectively；Textural characteristics set corresponding with each rotation process is compared with database, calculates the corresponding similarity of each rotation process, if the corresponding similarity of each rotation process is above given threshold, using face corresponding in database as the face recognized.The present invention, which can be avoided shelter, influences identification, and increases characteristic information amount, improves recognition accuracy.
Description
Technical field
The present invention relates to technical field of image processing.It is more particularly related to a kind of face identification method.
Background technique
Recognition of face has been obtained in video monitoring, identification field and is widely applied.But head part inclines
Gradient, shelter, shooting angle, light can impact recognition accuracy.Also, in some occasions, people's offer of putting on record
Facial image it is less, many times only one, the characteristic information that can be obtained is limited, also results in centainly to recognition accuracy
Influence.Therefore, it needs to design a kind of face identification method for being able to solve the above problem.
Summary of the invention
It is an object of the present invention to provide a kind of face identification methods, utilize edge algorithms and benchmark rgb value range
The first human face region and the second human face region are obtained respectively, and are avoided using the overlapping region of the two as the region of feature extraction
Shelter influences feature extraction, and increases characteristic information amount using rotation and zoom operations, improves recognition accuracy.
In order to realize these purposes and other advantages according to the present invention, face identification method is provided, comprising:
Step 1: obtaining images to be recognized；
Step 2: the first human face region is obtained from images to be recognized, from images to be recognized according to edge detection algorithm
The region to get colors within the scope of benchmark rgb value, as the second human face region, by the first human face region and the second human face region
The region of coincidence is as third human face region；
Step 3: obtaining characteristic area from third human face region, a rectangular area is determined with the center of characteristic area, it will
Rectangular area successively rotates 0~180 ° around center, and after each rotation process, rectangular area is divided into multiple zonules, respectively
The textural characteristics of each zonule are extracted, textural characteristics set corresponding with each rotation process is obtained；
Step 4: textural characteristics set corresponding with each rotation process is compared with database, each rotation is calculated
Turn the corresponding similarity of operation, it, will be corresponding in database if the corresponding similarity of each rotation process is above given threshold
Face as the face recognized；
Step 5: if the corresponding similarity of wherein one or more rotation process is lower than given threshold, by third face
0.1~10 times of area zoom, the operation of step 3 and step 4 is then repeated using the third human face region after scaling as object.
Preferably, the face identification method, benchmark rgb value range include the rgb value range of skin of face, lip
Rgb value range and eyes rgb value range.
Preferably, the face identification method, characteristic area are eyes or mouth.
Preferably, it is identical to be uniformly divided into multiple shapes, and is rectangle by the face identification method for rectangular area
Zonule.
Preferably, the face identification method compares the textural characteristics in corresponding zonule, Mei Gexuan respectively
Turn the ratio that the corresponding similarity of operation is textural characteristics identical zonule and zonule total quantity.
Preferably, the face identification method, databases contain the people that puts on record facial image and with put on record
The corresponding textural characteristics set of the rotation process of the facial image of people, the acquisition modes and step of the textural characteristics set in database
Rapid three is identical.
The present invention is include at least the following beneficial effects:
The present invention obtains the first human face region and the second human face region using edge algorithms and benchmark rgb value range respectively,
And search for obtain characteristic area from the overlapping region of the two, characteristic area and its neighboring area are put into rectangular area, it will
Rectangular area is divided into multiple zonules, extracts the textural characteristics of zonule, compares, fills with database as textural characteristics set
Divide and local feature is utilized, improves recognition accuracy, and reduce influence of the shelter to recognition effect.The present invention is to rectangle
Region carries out multiple rotary and zooms in and out operation to overlapping region, obtains textural characteristics set respectively, enhances to part
The use of feature, the recognition accuracy further increased.
Further advantage, target and feature of the invention will be partially reflected by the following instructions, and part will also be by this
The research and practice of invention and be understood by the person skilled in the art.
Specific embodiment
The present invention will be further described in detail below with reference to the embodiments, to enable those skilled in the art referring to specification
Text can be implemented accordingly.
It should be appreciated that such as " having ", "comprising" and " comprising " term used herein do not allot one or more
The presence or addition of a other elements or combinations thereof.
The present invention provides face identification methods, comprising:
Step 1: obtaining images to be recognized；
Step 2: the first human face region is obtained from images to be recognized, from images to be recognized according to edge detection algorithm
The region to get colors within the scope of benchmark rgb value, as the second human face region, by the first human face region and the second human face region
The region of coincidence is as third human face region；
Step 3: obtaining characteristic area from third human face region, a rectangular area is determined with the center of characteristic area, it will
Rectangular area successively rotates 0~180 ° around center, and after each rotation process, rectangular area is divided into multiple zonules, respectively
The textural characteristics of each zonule are extracted, textural characteristics set corresponding with each rotation process is obtained；
Step 4: textural characteristics set corresponding with each rotation process is compared with database, each rotation is calculated
Turn the corresponding similarity of operation, it, will be corresponding in database if the corresponding similarity of each rotation process is above given threshold
Face as the face recognized；
Step 5: if the corresponding similarity of wherein one or more rotation process is lower than given threshold, by third face
0.1~10 times of area zoom, the operation of step 3 and step 4 is then repeated using the third human face region after scaling as object.
In the above-mentioned technical solutions, database, identity information, facial image comprising the people that puts on record in database are pre-established
And the face characteristic information obtained according to facial image, the acquisition methods of face characteristic information specifically: obtain facial image
In human face region, characteristic area is obtained from human face region, characteristic area and its neighboring area are put into rectangular area,
Rotation process is carried out, after each rotation process, rectangular area is divided into multiple zonules, extracts the textural characteristics of zonule,
Textural characteristics set corresponding with each rotation process is obtained, and human face region is scaled 0.1~10 times, and carry out rotation behaviour
Make, divide and extract, obtains textural characteristics set corresponding with each scaling-rotation process.That is, containing people in database
The local grain information and the local grain information after scaling of the multiple angles of face characteristic area and neighboring area, it is thus possible to fill
Divide and reflect people's face characteristic of putting on record, improves recognition accuracy.In identification process, images to be recognized is acquired first, utilizes side
Edge algorithm obtains the first human face region, then obtains the second facial image using benchmark rgb value range, i.e., will be located at benchmark RGB
Pixel within the scope of value is included into the second human face region, and the first face is due to being obtained according to edge algorithms, possibly can not only sentence
It is disconnected whether to there is veil (such as cap, mask) to cover face, and be also possible to for the object in background being judged as using color
Face, and use both coincidence region as third human face region be used for handle, can avoid to a certain degree veil or
Environment interferes identification.To third human face region take in database obtain the method for face characteristic information it is identical, see
Step 3 and step 4, characteristic area are eyes, nose, mouth etc., are easy to get using the prior art, and rotation angle is preferably
0 °, 60 °, 120 ° and 180 °, scaling multiple is preferably 0.5 times and 2 times.Then by the face characteristic information of acquisition and database into
Row compares, i.e., compares the textural characteristics of each zonule one by one, judge the corresponding texture feature set of each rotation process
The similarity of conjunction then judges personnel to be identified for the people that puts on record if being above given threshold.If there is one or more to be less than
Given threshold is then rotated, is divided and extraction operation, and and data then by first multiple of third face area zoom
Corresponding data in library is compared, if similarity not yet all higher than given threshold, then by third face area zoom
Two multiples, are also rotated, are divided, extracted and are compared, if similarity is above given threshold, judge people to be identified
Member is the people that puts on record, if similarity repeats scaling, rotation, separates, extracts and compare not yet all higher than given threshold, until
Similarity is above given threshold, if similarity cannot be met the requirements, then judge personnel to be identified under all scaling multiples
It is not the people that puts on record.As can be seen that the present invention obtains human face region using two methods of edge algorithms and color, avoids and block
Interference of the object to identification process, and using repeatedly scaling and the texture information of multiple angles, it avoids leakage and knows, improve knowledge
Other accuracy rate.
In another technical solution, the face identification method, benchmark rgb value range includes the RGB of skin of face
It is worth the rgb value range of range, the rgb value range of lip and eyes.Here there is provided preferred benchmark rgb value range, bases
The skin of face, lip and eye color that quasi- rgb value range can acquire all people that put on record obtain, for example can be averaged
Value.
In another technical solution, the face identification method, characteristic area is eyes or mouth.Here, it provides
Preferred feature region, eyes and mouth have more textural characteristics, everyone lip texture is also unique, eyes and mouth
Bar comparison is suitble to recognition methods of the invention.
In another technical solution, the face identification method, it is identical that rectangular area is uniformly divided into multiple shapes,
It and is the zonule of rectangle.Here, it is easier to compare using the zonule of rectangle, the zonule overlapping that will be compared
It is compared, improves calculating speed.
In another technical solution, the face identification method distinguishes the textural characteristics in corresponding zonule
It compares, the corresponding similarity of each rotation process is the ratio of textural characteristics identical zonule and zonule total quantity.Here,
Judge the texture in two rectangle cell domains it is whether identical can rule of thumb or statistical result determine, and similarity be same cells
The ratio in domain, preferred given threshold are 80%.
In another technical solution, the face identification method, databases contain the facial image for the people that puts on record
And textural characteristics set corresponding with the rotation process of the facial image for the people that puts on record, textural characteristics set in database obtain
Take mode identical as step 3.It here is the method and the acquisition side in specific identification process that database obtains face characteristic information
Method is identical, facilitates comparison.
Although the embodiments of the present invention have been disclosed as above, but its is not only in the description and the implementation listed
With it can be fully applied to various fields suitable for the present invention, for those skilled in the art, can be easily
Realize other modification, therefore without departing from the general concept defined in the claims and the equivalent scope, the present invention is simultaneously unlimited
In specific details and embodiment shown and described herein.
Claims (6)
1. face identification method characterized by comprising
Step 1: obtaining images to be recognized；
Step 2: obtaining the first human face region from images to be recognized according to edge detection algorithm, being chosen from images to be recognized
First human face region is overlapped by region of the color within the scope of benchmark rgb value as the second human face region with the second human face region
Region as third human face region；
Step 3: obtaining characteristic area from third human face region, a rectangular area is determined with the center of characteristic area, by rectangle
Region successively rotates 0~180 ° around center, after each rotation process, rectangular area is divided into multiple zonules, is extracted respectively
The textural characteristics of each zonule obtain textural characteristics set corresponding with each rotation process；
Step 4: textural characteristics set corresponding with each rotation process is compared with database, each rotation behaviour is calculated
Make corresponding similarity, if the corresponding similarity of each rotation process is above given threshold, by people corresponding in database
Face is as the face recognized；
Step 5: if the corresponding similarity of wherein one or more rotation process is lower than given threshold, by third human face region
Then 0.1~10 times of scaling repeats the operation of step 3 and step 4 using the third human face region after scaling as object.
2. face identification method as described in claim 1, which is characterized in that benchmark rgb value range includes the RGB of skin of face
It is worth the rgb value range of range, the rgb value range of lip and eyes.
3. face identification method as described in claim 1, which is characterized in that characteristic area is eyes or mouth.
4. face identification method as described in claim 1, which is characterized in that rectangular area is uniformly divided into multiple shape phases
Together, and be rectangle zonule.
5. face identification method as claimed in claim 4, which is characterized in that distinguish the textural characteristics in corresponding zonule
It compares, the corresponding similarity of each rotation process is the ratio of textural characteristics identical zonule and zonule total quantity.
6. face identification method as described in claim 1, which is characterized in that databases contain the facial image for the people that puts on record
And textural characteristics set corresponding with the rotation process of the facial image for the people that puts on record, textural characteristics set in database obtain
Take mode identical as step 3.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810796508.5A CN109002799B (en) | 2018-07-19 | 2018-07-19 | Face recognition method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810796508.5A CN109002799B (en) | 2018-07-19 | 2018-07-19 | Face recognition method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109002799A true CN109002799A (en) | 2018-12-14 |
| CN109002799B CN109002799B (en) | 2021-08-24 |
Family
ID=64596747
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810796508.5A Active CN109002799B (en) | 2018-07-19 | 2018-07-19 | Face recognition method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109002799B (en) |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109886213A (en) * | 2019-02-25 | 2019-06-14 | 湖北亿咖通科技有限公司 | Fatigue state judgment method, electronic equipment and computer readable storage medium |
| CN110956768A (en) * | 2019-12-05 | 2020-04-03 | 重庆电子工程职业学院 | Automatic anti-theft device of intelligence house |
| CN111582983A (en) * | 2020-05-07 | 2020-08-25 | 悠尼客(上海)企业管理有限公司 | Personalized control method based on face recognition and customer behaviors |
| CN111768545A (en) * | 2020-06-28 | 2020-10-13 | 北京伟杰东博信息科技有限公司 | Traffic safety monitoring method and system |
| CN111814603A (en) * | 2020-06-23 | 2020-10-23 | 汇纳科技股份有限公司 | Face recognition method, medium and electronic device |
| CN112131915A (en) * | 2019-06-25 | 2020-12-25 | 杭州海康威视数字技术股份有限公司 | Face attendance system, camera and code stream equipment |
| CN113011277A (en) * | 2021-02-25 | 2021-06-22 | 日立楼宇技术（广州）有限公司 | Data processing method, device, equipment and medium based on face recognition |
| CN113420663A (en) * | 2021-06-23 | 2021-09-21 | 深圳市海清视讯科技有限公司 | Child face recognition method and system |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1811793A (en) * | 2006-03-02 | 2006-08-02 | 复旦大学 | Automatic positioning method for characteristic point of human faces |
| US20090310828A1 (en) * | 2007-10-12 | 2009-12-17 | The University Of Houston System | An automated method for human face modeling and relighting with application to face recognition |
| CN106250843A (en) * | 2016-07-28 | 2016-12-21 | 北京师范大学 | A kind of method for detecting human face based on forehead region and system |
| US20170343481A1 (en) * | 2016-05-27 | 2017-11-30 | Purdue Research Foundation | Methods and systems for crack detection |
-
2018
- 2018-07-19 CN CN201810796508.5A patent/CN109002799B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1811793A (en) * | 2006-03-02 | 2006-08-02 | 复旦大学 | Automatic positioning method for characteristic point of human faces |
| US20090310828A1 (en) * | 2007-10-12 | 2009-12-17 | The University Of Houston System | An automated method for human face modeling and relighting with application to face recognition |
| US20170343481A1 (en) * | 2016-05-27 | 2017-11-30 | Purdue Research Foundation | Methods and systems for crack detection |
| CN106250843A (en) * | 2016-07-28 | 2016-12-21 | 北京师范大学 | A kind of method for detecting human face based on forehead region and system |
Cited By (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109886213A (en) * | 2019-02-25 | 2019-06-14 | 湖北亿咖通科技有限公司 | Fatigue state judgment method, electronic equipment and computer readable storage medium |
| CN109886213B (en) * | 2019-02-25 | 2021-01-08 | 湖北亿咖通科技有限公司 | Fatigue state determination method, electronic device, and computer-readable storage medium |
| CN112131915B (en) * | 2019-06-25 | 2023-03-24 | 杭州海康威视数字技术股份有限公司 | Face attendance system, camera and code stream equipment |
| CN112131915A (en) * | 2019-06-25 | 2020-12-25 | 杭州海康威视数字技术股份有限公司 | Face attendance system, camera and code stream equipment |
| CN110956768A (en) * | 2019-12-05 | 2020-04-03 | 重庆电子工程职业学院 | Automatic anti-theft device of intelligence house |
| CN111582983A (en) * | 2020-05-07 | 2020-08-25 | 悠尼客(上海)企业管理有限公司 | Personalized control method based on face recognition and customer behaviors |
| CN111814603A (en) * | 2020-06-23 | 2020-10-23 | 汇纳科技股份有限公司 | Face recognition method, medium and electronic device |
| CN111814603B (en) * | 2020-06-23 | 2023-09-05 | 汇纳科技股份有限公司 | Face recognition method, medium and electronic equipment |
| CN111768545A (en) * | 2020-06-28 | 2020-10-13 | 北京伟杰东博信息科技有限公司 | Traffic safety monitoring method and system |
| CN111768545B (en) * | 2020-06-28 | 2021-07-23 | 广东邦盛北斗科技股份公司 | Traffic safety monitoring method and system |
| CN113011277A (en) * | 2021-02-25 | 2021-06-22 | 日立楼宇技术（广州）有限公司 | Data processing method, device, equipment and medium based on face recognition |
| CN113011277B (en) * | 2021-02-25 | 2023-11-21 | 日立楼宇技术（广州）有限公司 | Face recognition-based data processing method, device, equipment and medium |
| CN113420663A (en) * | 2021-06-23 | 2021-09-21 | 深圳市海清视讯科技有限公司 | Child face recognition method and system |
| CN113420663B (en) * | 2021-06-23 | 2022-02-22 | 深圳市海清视讯科技有限公司 | Child face recognition method and system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109002799B (en) | 2021-08-24 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109002799A (en) | Face identification method | |
| Yan et al. | Biometric recognition using 3D ear shape | |
| Mozaffari et al. | Gender classification using single frontal image per person: combination of appearance and geometric based features | |
| CN109190535B (en) | Face complexion analysis method and system based on deep learning | |
| CN112232332B (en) | Non-contact palm detection method based on video sequence | |
| Atharifard et al. | Robust component-based face detection using color feature | |
| CN104765739B (en) | Extensive face database search method based on shape space | |
| Long et al. | Near infrared face image quality assessment system of video sequences | |
| Pervaiz | Real time face recognition system based on EBGM framework | |
| Dahal et al. | Incorporating skin color for improved face detection and tracking system | |
| Charoenpong et al. | A new method for occluded face detection from single viewpoint of head | |
| Batz et al. | A computer vision system for monitoring medication intake | |
| CN109753912A (en) | A kind of multi-light spectrum palm print matching process based on tensor | |
| Yi et al. | Face detection method based on skin color segmentation and facial component localization | |
| Singh et al. | Efficient tool for face detection and face recognition in color group photos | |
| Wu et al. | Real-time 2D hands detection and tracking for sign language recognition | |
| Le et al. | Fast and robust self-training beard/moustache detection and segmentation | |
| Guo et al. | Face detection with abstract template | |
| Mandhala et al. | Face detection using image morphology–a review | |
| Guerfi et al. | Implementation of the watershed method in the hsi color space for the face extraction | |
| Arca et al. | Face recognition based on 2d and 3d features | |
| Gupta et al. | Automatic and Robust Detection of Facial Features in Posed and Tilted Face Images | |
| Suksil et al. | Hand detection and feature extraction for static Thai Sign Language recognition | |
| Angulu et al. | Frontal face landmark displacement across age | |
| Adachi et al. | Extraction of face parts by using flesh color and boundary images |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |