CN112287863A - Computer portrait recognition system - Google Patents
Computer portrait recognition system Download PDFInfo
- Publication number
- CN112287863A CN112287863A CN202011241359.XA CN202011241359A CN112287863A CN 112287863 A CN112287863 A CN 112287863A CN 202011241359 A CN202011241359 A CN 202011241359A CN 112287863 A CN112287863 A CN 112287863A
- Authority
- CN
- China
- Prior art keywords
- region
- face
- dssd
- mouth
- image
- 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
- 210000004709 eyebrow Anatomy 0.000 claims abstract description 27
- 230000000977 initiatory effect Effects 0.000 claims abstract description 24
- 210000000697 sensory organ Anatomy 0.000 claims abstract description 4
- 238000012706 support-vector machine Methods 0.000 claims abstract description 4
- 238000001514 detection method Methods 0.000 claims description 28
- 230000008859 change Effects 0.000 claims description 22
- 238000013528 artificial neural network Methods 0.000 claims description 18
- 230000008921 facial expression Effects 0.000 claims description 18
- 238000004422 calculation algorithm Methods 0.000 claims description 9
- 230000014509 gene expression Effects 0.000 claims description 6
- 230000001815 facial effect Effects 0.000 description 17
- 210000000887 face Anatomy 0.000 description 5
- 238000000034 method Methods 0.000 description 3
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000012790 confirmation Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
Images
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
- G06V40/161—Detection; Localisation; Normalisation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- 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
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/40—Spoof detection, e.g. liveness detection
- G06V40/45—Detection of the body part being alive
Abstract
The invention relates to a face recognition system, in particular to a computer face recognition system, which comprises a face image acquisition module, a face image processing module and a face image processing module, wherein the face image acquisition module is used for acquiring a real-time dynamic face image group to be recognized; the image processing module is used for dividing the face image into an eye region, an eyebrow region, a nose region, a mouth region and a face contour region according to the positions of the five sense organs; the face recognition module firstly realizes the recognition of the identity information of an eye region, an eyebrow region, a nose region, a mouth region and a face outline region respectively based on a Dssd _ initiation _ V4 coco model, and then obtains a final face recognition result according to the identity recognition result of each region based on a support vector machine. The invention identifies the face based on the eye area, the eyebrow area, the nose area, the mouth area, the face contour area and the Dssd _ acceptance _ V4 coco model, thereby reducing the blind area of face identification as much as possible.
Description
Technical Field
The invention relates to a face recognition system, in particular to a computer face recognition system.
Background
At present, an existing portrait recognition system is generally implemented based on the overall features of a face, and once the overall features of the face change, for example, a certain part of the face has a shielding object, a failure of portrait recognition can be caused, and a large recognition blind area exists.
Disclosure of Invention
The invention aims to provide a computer portrait recognition system, which is used for recognizing human faces based on an eye region, an eyebrow region, a nose region, a mouth region and a human face outline region, so that blind areas of portrait recognition are reduced as much as possible, and a hardware device for detecting whether the portrait is a living body is not required to be configured independently.
In order to achieve the purpose, the invention adopts the technical scheme that:
a computer portrait recognition system comprises
The face image acquisition module is used for acquiring a real-time dynamic face image group to be identified;
the image processing module is used for dividing the face image into an eye region, an eyebrow region, a nose region, a mouth region and a face contour region according to the positions of the five sense organs;
the face recognition module firstly realizes the recognition of the identity information of an eye region, an eyebrow region, a nose region, a mouth region and a face outline region respectively based on a Dssd _ initiation _ V4 coco model, and then obtains a final face recognition result according to the identity recognition result of each region based on a support vector machine.
Furthermore, the facial image acquisition module comprises two working modes, wherein one mode is to detect free facial images and directly enter the facial image video acquisition process by clicking 'identity recognition', the other mode is to identify identity authority and acquire a dynamic facial image group by randomly calling the facial expression required audio, a user needs to make corresponding facial expressions according to the facial expression required audio, and the dynamic facial image group at least comprises two facial images with different expressions.
And further, the face image acquisition module detects whether the face image group is a living body face image by recording a dynamic video generated between two adjacent face expression images, and during detection, firstly, a video frame taking script is called to obtain one image at a certain frame number, and then, the detection of whether the face image is a living body is carried out by identifying the dynamic change information of the face carried in the two adjacent images.
Further, the identification of the dynamic change information of the human faces carried in the two adjacent images is realized through a Dssd _ initiation _ V4 coco model, the dynamic change information of the human faces is the form change information of an eye area and a mouth area, the Dssd _ initiation _ V4 coco model adopts a Dssd target detection algorithm, an initiation _ V4 deep neural network is pre-trained by a coco data set, then the model is trained by an eye area change image set and a mouth area change image set corresponding to different prepared expression changes, various parameters in the deep neural network are finely adjusted, and finally, a proper target detection model for detecting the form change information of the eye area and the mouth area is obtained.
Further, the image processing module realizes the delineation of an eye region, an eyebrow region, a nose region, a mouth region and a face contour region in the face image based on the Dssd _ initiation _ V4 coco model.
Further, the Dssd _ initiation _ V4 coco model adopts Dssd target detection algorithm, pre-trains initiation _ V4 deep neural network with coco data set, then trains the model with the previously prepared eye region, eyebrow region, nose region, mouth region and face contour region data set, fine-tunes each parameter in the deep neural network, and finally obtains the suitable target detection model for detecting eye region, eyebrow region, nose region, mouth region and face contour region.
Furthermore, the Dssd _ initiation _ V4 coco model loaded in the face recognition module uses Dssd target detection algorithm to pre-train initiation _ V4 deep neural network with coco data set, then trains the model with the previously prepared data set of eye region, eyebrow region, nose region, mouth region and face contour region carrying identity information, fine-tunes various parameters in the deep neural network, and finally obtains a suitable target detection model for detecting the identity information corresponding to the eye region, eyebrow region, nose region, mouth region and face contour region.
The invention has the following beneficial effects:
the face recognition is carried out based on the eye region, the eyebrow region, the nose region, the mouth region, the face contour region and the Dssd _ initiation _ V4 coco model, so that the blind area of face recognition is reduced as much as possible;
when the method is applied to authority identity recognition, a hardware device for detecting whether the portrait is a living body or not is not required to be configured independently, and the detection of whether the face image is a living body or not can be realized while the image is acquired.
Drawings
Fig. 1 is a system block diagram of a computer portrait recognition system according to an embodiment of the present invention.
Detailed Description
In order that the objects and advantages of the invention will be more clearly understood, the invention is further described in detail below with reference to examples. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
As shown in fig. 1, an embodiment of the present invention provides a computer portrait recognition system, including:
the face image acquisition module is used for acquiring a real-time dynamic face image group to be identified;
the image processing module is used for dividing the face image into an eye region, an eyebrow region, a nose region, a mouth region and a face contour region according to the positions of the five sense organs;
the face recognition module firstly realizes the recognition of the identity information of an eye region, an eyebrow region, a nose region, a mouth region and a face outline region respectively based on a Dssd _ initiation _ V4 coco model, and then obtains a final face recognition result according to the identity recognition result of each region based on a support vector machine.
In this embodiment, the facial image acquisition module includes two working modes, one is for free portrait detection, the process of facial image video acquisition can be directly entered by clicking 'identity recognition', the other is for identity authority recognition, a dynamic facial image group is obtained by randomly calling the facial expression requirement audio (such as tooth exposing smile, sipping mouth, mouth beeping, frown and the like), a user needs to make a corresponding facial expression according to the facial expression requirement audio, and the dynamic facial image group at least includes two different facial expression images; when the system is used, a user clicks identity recognition, the facial image acquisition module starts to play the audio required by the first facial expression, the user makes a corresponding expression according to needs, and clicks a confirmation button, after the acquisition of the first facial expression is finished, the audio required by the second facial expression is played, and the image acquisition function of the facial image acquisition module is always in an operating state after the acquisition of the first facial expression is finished and before the acquisition of the second facial expression is finished.
In this embodiment, the facial image acquisition module detects whether a facial image group is a living body facial image by recording a dynamic video generated between two adjacent facial expression images, and during detection, firstly, a video frame fetching script is called to obtain an image at a certain frame number, and then, the detection of whether the facial image is a living body is performed by identifying the dynamic change information of the facial image carried in the two adjacent images. The detection of whether the face image is a living body can be realized at the same time of image acquisition.
In this embodiment, the Dssd _ acceptance _ V4 coco model is used to recognize dynamic change information of faces carried in two adjacent images, where the dynamic change information of faces is form change information of an eye region and a mouth region, the Dssd _ acceptance _ V4 coco model adopts a Dssd target detection algorithm, an acceptance _ V4 deep neural network is pre-trained by a coco data set, then the model is trained by an eye region change image set and a mouth region change image set corresponding to different previously prepared expression changes, various parameters in the deep neural network are finely adjusted, and finally a suitable target detection model for detecting form change information of the eye region and the mouth region is obtained.
In this embodiment, the image processing module first defines an eye region, an eyebrow region, a nose region, a mouth region, and a face contour region in the face image based on the Dssd _ acceptance _ V4 coco model, and then calls an image mining program to mine the defined region, so as to obtain a corresponding eye region image, eyebrow region image, nose region image, mouth region image, and face contour region image.
In this embodiment, the Dssd _ initiation _ V4 coco model uses a Dssd target detection algorithm to pre-train an initiation _ V4 deep neural network with a coco data set, then trains the model with the previously prepared eye region, eyebrow region, nose region, mouth region and face contour region data sets, fine-tunes various parameters in the deep neural network, and finally obtains a suitable target detection model for detecting the eye region, eyebrow region, nose region, mouth region and face contour region.
In this embodiment, the Dssd _ initiation _ V4 coco model loaded in the face recognition module uses a Dssd target detection algorithm to pre-train an initiation _ V4 deep neural network with a coco data set, then trains the model with the previously prepared data sets of an eye region, an eyebrow region, a nose region, a mouth region, and a face contour region, which carry identity information, and fine-tunes various parameters in the deep neural network, so as to finally obtain a suitable target detection model for detecting the identity information corresponding to the eye region, the eyebrow region, the nose region, the mouth region, and the face contour region.
The foregoing is only a preferred embodiment of the present invention, and it should be noted that those skilled in the art can make various improvements and modifications without departing from the principle of the present invention, and these improvements and modifications should also be construed as the protection scope of the present invention.
Claims (7)
1. A computer portrait recognition system is characterized by comprising
The face image acquisition module is used for acquiring a real-time dynamic face image group to be identified;
the image processing module is used for dividing the face image into an eye region, an eyebrow region, a nose region, a mouth region and a face contour region according to the positions of the five sense organs;
the face recognition module firstly realizes the recognition of the identity information of an eye region, an eyebrow region, a nose region, a mouth region and a face outline region respectively based on a Dssd _ initiation _ V4 coco model, and then obtains a final face recognition result according to the identity recognition result of each region based on a support vector machine.
2. The computer face recognition system of claim 1, wherein the face image acquisition module comprises two modes, one mode is for free face detection and for "identity recognition" by clicking, and the other mode is for identity authority recognition, and the dynamic face image group is obtained by randomly calling the face expression requirement audio, and the user needs to make a corresponding face expression according to the face expression requirement audio, and the dynamic face image group comprises at least two different expression face images.
3. The computer face recognition system of claim 2, wherein the face image acquisition module detects whether the face image group is a living face image by recording a dynamic video generated between two adjacent face expression images, and during detection, firstly, a video frame fetching script is called to obtain an image at a certain frame number, and then, the detection of whether the face image is a living body is performed by recognizing the dynamic change information of the face carried in the two adjacent images.
4. The computer human image recognition system of claim 3, wherein the recognition of the dynamic human face change information carried in two adjacent images is realized through a Dssd _ initiation _ V4 coco model, the dynamic human face change information is the form change information of an eye region and a mouth region, the Dssd _ initiation _ V4 coco model adopts a Dssd target detection algorithm, an initiation _ V4 deep neural network is pre-trained by a coco data set, then the model is trained by an eye region change image set and a mouth region change image set corresponding to different prepared expression changes, parameters in the deep neural network are finely adjusted, and finally, a suitable target detection model for detecting the form change information of the eye region and the mouth region is obtained.
5. The computer human image recognition system of claim 1, wherein the image processing module implements delineation of eye regions, eyebrow regions, nose regions, mouth regions and face contour regions in the human face image based on Dssd _ acceptance _ V4 coco model.
6. The computer human recognition system of claim 5, wherein the Dssd _ initiation _ V4 coco model employs Dssd object detection algorithm to pre-train an initiation _ V4 deep neural network with a coco data set, then train the model with a previously prepared data set of eye region, eyebrow region, nose region, mouth region, and face contour region, fine tune parameters in the deep neural network, and finally obtain an appropriate object detection model for detecting the eye region, eyebrow region, nose region, mouth region, and face contour region.
7. The computer human image recognition system of claim 1, wherein the Dssd _ initiation _ V4 coco model loaded in the face recognition module uses Dssd target detection algorithm to pre-train the initiation _ V4 deep neural network with the coco data set, then train the model with the previously prepared eye region, eyebrow region, nose region, mouth region and face contour region data set carrying identification information, fine tune various parameters in the deep neural network, and finally obtain the appropriate target detection model for detecting the identification information corresponding to the eye region, eyebrow region, nose region, mouth region and face contour region.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011241359.XA CN112287863A (en) | 2020-11-09 | 2020-11-09 | Computer portrait recognition system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011241359.XA CN112287863A (en) | 2020-11-09 | 2020-11-09 | Computer portrait recognition system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112287863A true CN112287863A (en) | 2021-01-29 |
Family
ID=74351723
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011241359.XA Pending CN112287863A (en) | 2020-11-09 | 2020-11-09 | Computer portrait recognition system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112287863A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112990113A (en) * | 2021-04-20 | 2021-06-18 | 北京远鉴信息技术有限公司 | Living body detection method and device based on facial expression of human face and electronic equipment |
| CN113572955A (en) * | 2021-06-25 | 2021-10-29 | 维沃移动通信（杭州）有限公司 | Image processing method and device and electronic equipment |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2004272326A (en) * | 2003-03-05 | 2004-09-30 | Matsushita Electric Ind Co Ltd | Probabilistic facial component fusion method for face description and recognition using subspace component feature |
| CN108062546A (en) * | 2018-02-11 | 2018-05-22 | 厦门华厦学院 | A kind of computer face Emotion identification system |
| CN111368803A (en) * | 2020-03-28 | 2020-07-03 | 河南工业职业技术学院 | Face recognition method and system |
| CN111414896A (en) * | 2020-04-10 | 2020-07-14 | 吉林农业大学 | Crop disease image identification method based on convolutional neural network |
| CN111507978A (en) * | 2020-05-08 | 2020-08-07 | 延安大学 | Intelligent digital image processing system for urology surgery |
| CN211603979U (en) * | 2020-03-27 | 2020-09-29 | 九江产教融合发展中心 | Wisdom agricultural remote control system |
-
2020
- 2020-11-09 CN CN202011241359.XA patent/CN112287863A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2004272326A (en) * | 2003-03-05 | 2004-09-30 | Matsushita Electric Ind Co Ltd | Probabilistic facial component fusion method for face description and recognition using subspace component feature |
| CN108062546A (en) * | 2018-02-11 | 2018-05-22 | 厦门华厦学院 | A kind of computer face Emotion identification system |
| CN211603979U (en) * | 2020-03-27 | 2020-09-29 | 九江产教融合发展中心 | Wisdom agricultural remote control system |
| CN111368803A (en) * | 2020-03-28 | 2020-07-03 | 河南工业职业技术学院 | Face recognition method and system |
| CN111414896A (en) * | 2020-04-10 | 2020-07-14 | 吉林农业大学 | Crop disease image identification method based on convolutional neural network |
| CN111507978A (en) * | 2020-05-08 | 2020-08-07 | 延安大学 | Intelligent digital image processing system for urology surgery |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112990113A (en) * | 2021-04-20 | 2021-06-18 | 北京远鉴信息技术有限公司 | Living body detection method and device based on facial expression of human face and electronic equipment |
| CN113572955A (en) * | 2021-06-25 | 2021-10-29 | 维沃移动通信（杭州）有限公司 | Image processing method and device and electronic equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110287790B (en) | Learning state hybrid analysis method oriented to static multi-user scene | |
| CN108470169A (en) | Face identification system and method | |
| CN105809144A (en) | Gesture recognition system and method adopting action segmentation | |
| CN112287863A (en) | Computer portrait recognition system | |
| CN112818767B (en) | Data set generation and forgery detection methods and devices, electronic equipment and storage medium | |
| CN109766785A (en) | A kind of biopsy method and device of face | |
| CN106648078A (en) | Multimode interaction method and system applied to intelligent robot | |
| CN111611860B (en) | Micro-expression occurrence detection method and detection system | |
| CN104965589A (en) | Human living body detection method and device based on human brain intelligence and man-machine interaction | |
| CN112507772A (en) | Face recognition security system and suspicious person detection and early warning method | |
| CN108614987A (en) | The method, apparatus and robot of data processing | |
| CN111860394A (en) | Gesture estimation and gesture detection-based action living body recognition method | |
| CN107085709A (en) | A kind of Confidence determination methods based on video information, apparatus and system | |
| CN107862298B (en) | Winking living body detection method based on infrared camera device | |
| CN110633677B (en) | Face recognition method and device | |
| CN113627256B (en) | False video inspection method and system based on blink synchronization and binocular movement detection | |
| CN112509154B (en) | Training method of image generation model, image generation method and device | |
| CN114241584A (en) | Face optimization identification method and terminal | |
| Taskirar et al. | Face recognition using dynamic features extracted from smile videos | |
| CN104992085A (en) | Method and device for human body in-vivo detection based on touch trace tracking | |
| CN116883900A (en) | Video authenticity identification method and system based on multidimensional biological characteristics | |
| CN111414886A (en) | Intelligent recognition system for human body dynamic characteristics | |
| CN111241505A (en) | Terminal device, login verification method thereof and computer storage medium | |
| CN111898464A (en) | Face recognition method based on 5G network | |
| JP2020149361A (en) | Expression estimating apparatus, feeling determining apparatus, expression estimating method, and program |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20210129 |
|
| RJ01 | Rejection of invention patent application after publication |