CN108154116A - A kind of image-recognizing method and system - Google Patents
A kind of image-recognizing method and system Download PDFInfo
- Publication number
- CN108154116A CN108154116A CN201711412133.XA CN201711412133A CN108154116A CN 108154116 A CN108154116 A CN 108154116A CN 201711412133 A CN201711412133 A CN 201711412133A CN 108154116 A CN108154116 A CN 108154116A
- Authority
- CN
- China
- Prior art keywords
- image
- face
- region
- pretreated
- feature information
- 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
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
- G06V40/169—Holistic features and representations, i.e. based on the facial image taken as a whole
-
- 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
- G06V10/20—Image preprocessing
- G06V10/22—Image preprocessing by selection of a specific region containing or referencing a pattern; Locating or processing of specific regions to guide the detection or recognition
-
- 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
- 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
Abstract
The present invention provides a kind of image-recognizing method and system, method therein includes:Real-time image acquisition, obtain original image, image preprocessing is carried out to original image, obtain pretreated image, pretreated image is detected, sub-argument goes out the human face region in image, and face feature information is extracted from human face region, face feature information and face characteristic library are compared, identify the highest face of similarity.The image-recognizing method and system of the present invention, by Face datection, pretreatment, feature extraction, matches the face of most similarity from database, ensure that the accuracy of recognition of face, be with a wide range of applications.
Description
Technical field
The present invention relates to image identification art field, more particularly to a kind of image-recognizing method and system.
Background technology
Increasingly extensive with the application of human-computer interaction technology, in field of human-computer interaction, face recognition technology has very
Important meaning.Recognition of face refers to the facial face of people and the distribution of profile, these distribution characteristics vary with each individual, all with life
Come.
However, face can change with expression, the variation at age, can because illumination, angle, distance etc. to image into
As being affected, these can all influence the accuracy of recognition of face.
Invention content
In view of the above problems, it is an object of the present invention to provide one kind to overcome the above problem or solve at least partly
The certainly image-recognizing method and system of the above problem.
The present invention one is further objective is that improving the precision of recognition of face and realizing quick identification face.
According to an aspect of the present invention, the present invention provides a kind of image-recognizing method, including:
Real-time image acquisition obtains original image;
Image preprocessing is carried out to original image, obtains pretreated image;
Pretreated image is detected, sub-argument goes out the human face region in image;
Face feature information is extracted from human face region;
Face feature information and face characteristic library are compared, identify the highest face of similarity.
Optionally, the step of carrying out image preprocessing to original image specifically includes:
Original image is carried out successively at image gray processing processing, image scaling processing and histogram light equalization
Reason.
Optionally, pretreated image is detected, the step of human face region that sub-argument goes out in image specifically includes:
Rough detection is carried out to pretreated image, obtains face candidate region in image;
Face candidate region is detected using AdaBoost algorithms, isolates human face region.
Optionally, the step of carrying out rough detection to pretreated image, obtaining face candidate region in image is specifically wrapped
It includes:
It is constantly slided in the image of one sliding window of setting after the pre-treatment, the every position of sliding window is just counted
The harr-like features at the position are calculated, if harr-like characteristic values are more than preset candidate threshold, it is determined that the region is
Face candidate region.
Optionally, the step of face feature information is extracted from human face region specifically includes:
Face feature information in extraction human face region is converted using Gabor wavelet.
According to a further aspect of the present invention, a kind of image identification system is additionally provided, including:
Image capture module is configured to real-time image acquisition, obtains original image;
Image pre-processing module is configured to carry out image preprocessing to original image, obtains pretreated image;
Face detection module is configured to be detected pretreated image, and sub-argument goes out the human face region in image;
Characteristic extracting module is configured to extract face feature information from human face region;
Identification module is configured to face feature information and face characteristic library being compared, identifies that similarity is highest
Face.
Optionally, image pre-processing module is additionally configured to:
Original image is carried out successively at image gray processing processing, image scaling processing and histogram light equalization
Reason.
Optionally, face detection module is additionally configured to:
Rough detection is carried out to pretreated image, obtains face candidate region in image;
Face candidate region is detected using AdaBoost algorithms, isolates human face region.
Optionally, face detection module is additionally configured to:
It is constantly slided in the image of one sliding window of setting after the pre-treatment, the every position of sliding window is just counted
The harr-like features at the position are calculated, if harr-like characteristic values are more than preset candidate threshold, it is determined that the region is
Face candidate region.
Optionally, characteristic extracting module is additionally configured to:
Face feature information in extraction human face region is converted using Gabor wavelet.
The image-recognizing method and system of the present invention, by Face datection, pretreatment, feature extraction, from database
The face of most similarity is allotted, the accuracy of recognition of face is ensure that, is with a wide range of applications.
According to the accompanying drawings to the detailed description of the specific embodiment of the invention, those skilled in the art will be brighter
The above and other objects, advantages and features of the present invention.
Description of the drawings
Some specific embodiments of detailed description of the present invention by way of example rather than limitation with reference to the accompanying drawings hereinafter.
Identical reference numeral denotes same or similar component or part in attached drawing.It should be appreciated by those skilled in the art that these
What attached drawing was not necessarily drawn to scale.In attached drawing:
Fig. 1 is the flow chart of image-recognizing method according to an embodiment of the invention;
Fig. 2 is the schematic diagram of image identification system according to an embodiment of the invention.
Specific embodiment
The present embodiment provides firstly a kind of image-recognizing method, and Fig. 1 is that image according to an embodiment of the invention is known
The flow chart of other method.
As shown in Figure 1, the image-recognizing method of the present embodiment includes:
S102, real-time image acquisition obtain original image;
S104 carries out image preprocessing to original image, obtains pretreated image;
S106 is detected pretreated image, and sub-argument goes out the human face region in image;
S108 extracts face feature information from human face region;
Face feature information and face characteristic library are compared by S110, identify the highest face of similarity.
In step S104, the step of original image progress image preprocessing, is included:Figure is carried out successively to original image
As gray processing processing, image scaling processing and histogram light equalization processing.It is handled by image gray processing by colour
Original image is converted to gray level image, and image down is suitably sized at one, promotion image inspection is handled by image scaling
The speed of survey improves contrast and the brightness of image by histogram light equalization processing, to avoid insufficient light or light
Cross the bright influence to subsequent detection.
In step s 106, pretreated image is detected, the step of human face region that sub-argument goes out in image has
Body includes:Rough detection is carried out to pretreated image, face candidate region in image is obtained, using AdaBoost algorithms to people
Face candidate region is detected, and isolates human face region.Wherein, rough detection is carried out to pretreated image, obtained in image
The step of face candidate region, specifically includes:It constantly slides, slides in the image of one sliding window of setting after the pre-treatment
The every position of window just calculates the harr-like features at the position, if harr-like characteristic values are more than preset candidate
Threshold, it is determined that the region is face candidate region.
Harr-1ike features are exactly that the picture frame of black and white point half is put into sliding window, and the pixel of white area is subtracted
The pixel of black region and, then set a wealthy value of candidate, the face candidate area that the region more than candidate wealthy value is exactly
Domain.
AdaBoost algorithms are that Adaboost is a kind of iterative algorithm, and core concept is trained for same training set
Different graders (Weak Classifier), then gets up these weak classifier sets, and it is (strong to form a stronger final classification device
Grader).Its algorithm realizes that it divides according to sample each among each training set by change data distribution in itself
Whether class correct and the accuracy rate of general classification of last time, to determine the weights of each sample.The new number of weights will be changed
Sub-classification device is given according to collection to be trained, the grader for finally obtaining each training finally merges, as last
Decision Classfication device.Some unnecessary training data features can be excluded using adaboost graders, and are placed on crucial instruction
Practice above data.
In step S108, specifically included the step of extraction face feature information from human face region:Using Gabor wavelet
Face feature information in transformation extraction human face region.
Multiple dimensioned, the multi-direction spatial frequency features in image specific region can be extracted due to Gabor wavelet transformation, as
Microscope equally amplifies the variation of gray scale, and eyes, nose and the mouth and other local features in such facial image are amplified,
Some key features in facial image can be enhanced, distinguish different facial images.
As shown in Fig. 2, based on above-mentioned image-recognizing method, the present embodiment additionally provides a kind of image identification system 10, wraps
Include image capture module 11, image pre-processing module 12, face detection module 13, characteristic extracting module 14 and identification module 15.
Image capture module 11 is configured to real-time image acquisition, obtains original image.Image pre-processing module 12 is configured to
Image preprocessing is carried out to original image, obtains pretreated image, image pre-processing module 12 is additionally configured to original graph
As carrying out image gray processing processing, image scaling processing and histogram light equalization processing successively.Pass through image gray processing
Colored original image is converted to gray level image by processing, is handled by image scaling by image down in a suitable ruler
It is very little, the speed of image detection is promoted, improves contrast and the brightness of image by histogram light equalization processing, to avoid light
Line is insufficient or the excessively bright influence to subsequent detection of light.
Face detection module 13 is configured to be detected pretreated image, and sub-argument goes out the human face region in image,
Face datection mould is additionally configured to carry out rough detection to pretreated image, obtains face candidate region in image, uses
AdaBoost algorithms are detected face candidate region, isolate human face region.Face detection module 13 is additionally configured to set
It is constantly slided in the image of one sliding window after the pre-treatment, the every position of sliding window is just calculated at the position
Harr-like features, if harr-like characteristic values are more than preset candidate threshold, it is determined that the region is face candidate region.
Harr-1ike features are exactly that the picture frame of black and white point half is put into sliding window, and the pixel of white area is subtracted
The pixel of black region and, then set a wealthy value of candidate, the face candidate area that the region more than candidate wealthy value is exactly
Domain.
Characteristic extracting module 14 is configured to extract face feature information from human face region, and characteristic extracting module 14 also configures
For using the face feature information in Gabor wavelet transformation extraction human face region.Since Gabor wavelet transformation can extract image
Multiple dimensioned, multi-direction spatial frequency features in specific region amplify the variation of gray scale, such facial image as microscope
In eyes, nose and mouth and other local features be amplified, some key features in facial image can be enhanced, distinguish
Different facial images.
Identification module 15 is configured to face feature information and face characteristic library being compared, and identifies that similarity is highest
Face.
So far, although those skilled in the art will appreciate that detailed herein have shown and described multiple showing for the present invention
Example property embodiment, still, without departing from the spirit and scope of the present invention, still can according to the present disclosure directly
Determine or derive many other variations or modifications consistent with the principles of the invention.Therefore, the scope of the present invention is understood that and recognizes
It is set to and covers other all these variations or modifications.
Claims (10)
1. a kind of image-recognizing method, which is characterized in that including:
Real-time image acquisition obtains original image;
Image preprocessing is carried out to the original image, obtains pretreated image;
The pretreated image is detected, sub-argument goes out the human face region in described image;
Face feature information is extracted from the human face region;
The face feature information and face characteristic library are compared, identify the highest face of similarity.
2. image-recognizing method according to claim 1, which is characterized in that described pre- to original image progress image
The step of processing, specifically includes:
The original image is carried out successively at image gray processing processing, image scaling processing and histogram light equalization
Reason.
3. image-recognizing method according to claim 1, which is characterized in that described to be carried out to the pretreated image
The step of detection, the human face region that sub-argument goes out in described image, specifically includes:
Rough detection is carried out to the pretreated image, obtains face candidate region in described image;
The face candidate region is detected using AdaBoost algorithms, isolates the human face region.
4. image-recognizing method according to claim 3, which is characterized in that described to be carried out to the pretreated image
Rough detection, obtain described image in face candidate region the step of specifically include:
One sliding window of setting constantly slides in the pretreated image, and the every position of sliding window is just counted
The harr-like features at the position are calculated, if harr-like characteristic values are more than preset candidate threshold, it is determined that the region is
Face candidate region.
5. image-recognizing method according to claim 1, which is characterized in that described that face is extracted from the human face region
The step of characteristic information, specifically includes:
The face feature information in the human face region is extracted using Gabor wavelet transformation.
6. a kind of image identification system, which is characterized in that including:
Image capture module is configured to real-time image acquisition, obtains original image;
Image pre-processing module is configured to carry out image preprocessing to the original image, obtains pretreated image;
Face detection module is configured to be detected the pretreated image, and sub-argument goes out the face area in described image
Domain;
Characteristic extracting module is configured to extract face feature information from the human face region;
Identification module is configured to the face feature information and face characteristic library being compared, identifies that similarity is highest
Face.
7. image identification system according to claim 6, which is characterized in that described image preprocessing module is additionally configured to:
The original image is carried out successively at image gray processing processing, image scaling processing and histogram light equalization
Reason.
8. image identification system according to claim 6, which is characterized in that the face detection module is additionally configured to:
Rough detection is carried out to the pretreated image, obtains face candidate region in described image;
The face candidate region is detected using AdaBoost algorithms, isolates the human face region.
9. image identification system according to claim 8, which is characterized in that the face detection module is additionally configured to:
One sliding window of setting constantly slides in the pretreated image, and the every position of sliding window is just counted
The harr-like features at the position are calculated, if harr-like characteristic values are more than preset candidate threshold, it is determined that the region is
Face candidate region.
10. image identification system according to claim 6, which is characterized in that the characteristic extracting module is additionally configured to:
The face feature information in the human face region is extracted using Gabor wavelet transformation.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201711412133.XA CN108154116A (en) | 2017-12-23 | 2017-12-23 | A kind of image-recognizing method and system |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201711412133.XA CN108154116A (en) | 2017-12-23 | 2017-12-23 | A kind of image-recognizing method and system |
Publications (1)
Publication Number | Publication Date |
---|---|
CN108154116A true CN108154116A (en) | 2018-06-12 |
Family
ID=62465506
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201711412133.XA Pending CN108154116A (en) | 2017-12-23 | 2017-12-23 | A kind of image-recognizing method and system |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108154116A (en) |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108985224A (en) * | 2018-07-13 | 2018-12-11 | 江苏慧学堂系统工程有限公司 | A kind of method and face identification system promoting face recognition accuracy rate |
CN110755847A (en) * | 2019-10-30 | 2020-02-07 | 腾讯科技(深圳)有限公司 | Virtual operation object generation method and device, storage medium and electronic device |
US11380037B2 (en) | 2019-10-30 | 2022-07-05 | Tencent Technology (Shenzhen) Company Limited | Method and apparatus for generating virtual operating object, storage medium, and electronic device |
-
2017
- 2017-12-23 CN CN201711412133.XA patent/CN108154116A/en active Pending
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108985224A (en) * | 2018-07-13 | 2018-12-11 | 江苏慧学堂系统工程有限公司 | A kind of method and face identification system promoting face recognition accuracy rate |
CN110755847A (en) * | 2019-10-30 | 2020-02-07 | 腾讯科技(深圳)有限公司 | Virtual operation object generation method and device, storage medium and electronic device |
CN110755847B (en) * | 2019-10-30 | 2021-03-16 | 腾讯科技(深圳)有限公司 | Virtual operation object generation method and device, storage medium and electronic device |
WO2021082787A1 (en) * | 2019-10-30 | 2021-05-06 | 腾讯科技(深圳)有限公司 | Virtual operation object generation method and device, storage medium and electronic apparatus |
US11380037B2 (en) | 2019-10-30 | 2022-07-05 | Tencent Technology (Shenzhen) Company Limited | Method and apparatus for generating virtual operating object, storage medium, and electronic device |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109154978B (en) | System and method for detecting plant diseases | |
Savkare et al. | Automatic system for classification of erythrocytes infected with malaria and identification of parasite's life stage | |
Shen et al. | Improving OCR performance with background image elimination | |
Konwar et al. | An American sign language detection system using HSV color model and edge detection | |
CN103824091B (en) | A kind of licence plate recognition method for intelligent transportation system | |
CN106610969A (en) | Multimodal information-based video content auditing system and method | |
KR20170006355A (en) | Method of motion vector and feature vector based fake face detection and apparatus for the same | |
CN106022231A (en) | Multi-feature-fusion-based technical method for rapid detection of pedestrian | |
CN108564092A (en) | Sunflower disease recognition method based on SIFT feature extraction algorithm | |
KR20160119932A (en) | Method and apparatus for face recognition based quality assessment | |
CN110659649A (en) | Image processing and character recognition algorithm based on near infrared light imaging | |
CN104484652A (en) | Method for fingerprint recognition | |
CN108154116A (en) | A kind of image-recognizing method and system | |
Monwar et al. | Pain recognition using artificial neural network | |
CN112464885A (en) | Image processing system for future change of facial color spots based on machine learning | |
Agarwal et al. | An efficient algorithm for automatic car plate detection & recognition | |
KR20080079798A (en) | Method of face detection and recognition | |
Hu et al. | Fast face detection based on skin color segmentation using single chrominance Cr | |
Das et al. | Human face detection in color images using HSV color histogram and WLD | |
Fathy et al. | Benchmarking of pre-processing methods employed in facial image analysis | |
Jindal et al. | Sign Language Detection using Convolutional Neural Network (CNN) | |
De Siqueira et al. | Adaptive detection of human skin in color images | |
CN109960976A (en) | A kind of image identification system | |
Mehta et al. | An Efficient Way to Detect and Recognize the Overlapped Coins using Otsu's Algorithm based on Hough Transform Technique | |
Gandhi et al. | Image based sign language recognition on android |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20180612 |
|
WD01 | Invention patent application deemed withdrawn after publication |