CN101201895B - Built-in human face recognizing monitor and detecting method thereof - Google Patents
Built-in human face recognizing monitor and detecting method thereof Download PDFInfo
- Publication number
- CN101201895B CN101201895B CN2007101519097A CN200710151909A CN101201895B CN 101201895 B CN101201895 B CN 101201895B CN 2007101519097 A CN2007101519097 A CN 2007101519097A CN 200710151909 A CN200710151909 A CN 200710151909A CN 101201895 B CN101201895 B CN 101201895B
- Authority
- CN
- China
- Prior art keywords
- face
- people
- self
- dsp
- human face
- 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.)
- Expired - Fee Related
Links
- 238000000034 method Methods 0.000 title claims description 8
- 238000001514 detection method Methods 0.000 claims abstract description 20
- 238000012360 testing method Methods 0.000 claims description 9
- 238000012545 processing Methods 0.000 claims description 6
- 239000000284 extract Substances 0.000 claims description 4
- 238000004364 calculation method Methods 0.000 claims description 2
- 238000004891 communication Methods 0.000 claims description 2
- 239000008358 core component Substances 0.000 claims description 2
- 238000000605 extraction Methods 0.000 claims description 2
- 230000006870 function Effects 0.000 description 6
- 238000005286 illumination Methods 0.000 description 5
- 238000010586 diagram Methods 0.000 description 4
- 230000008921 facial expression Effects 0.000 description 4
- 230000008569 process Effects 0.000 description 3
- 230000003993 interaction Effects 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 238000012549 training Methods 0.000 description 2
- 230000006399 behavior Effects 0.000 description 1
- 230000003542 behavioural effect Effects 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 230000008451 emotion Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000001815 facial effect Effects 0.000 description 1
- 238000003384 imaging method Methods 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 230000002093 peripheral effect Effects 0.000 description 1
- 230000009897 systematic effect Effects 0.000 description 1
Images
Abstract
The invention relates to an embedded face recognition monitor and a detection method thereof. The face recognition monitor provides detection of face video signal in the two input video signals and stacks the detected face video signal on the original video signal and outputs. The system detects face based on a digital signal processor (DSP). The system can real-time detect face image input by standard video and stacks the extracted signal on the given region of the original image to achieve the purpose of identifying face.
Description
Technical field
The present invention relates to a kind of human face recognizing monitor and detection method thereof, a kind of providing is provided the people's face vision signal in the two-path video signal of being imported is detected, and with detected people's face vision signal the be added to built-in human face recognizing monitor and the detection method thereof of the laggard line output of original video signal.
Background technology
Human face expression contains abundant human body behavioural information, play an important role in daily life, so FacialExpression Recognition and research is calculated and the intelligent interaction field has great significance in emotion.Because human face expression is very complicated, and be subjected to the influence of differences such as race, culture, present most of human face expression identification and the research that detects only limit to minority and typically express one's feelings, yet intelligent interaction requires to have the recognition capability near human expression, in order to realize this target, will realize the detection of people's face vision signal, and it will be handled.
Summary of the invention
The present invention relates to a kind of built-in human face recognizing monitor and detection method thereof, human face recognizing monitor provides the people's face vision signal in the two-path video signal of being imported is detected, and with detected people's face vision signal laggard line output of original video signal that is added to.Native system carries out people's face based on digital signal processor (DSP) and detects.People's face picture in the native system examination criteria video input in real time, and the appointed area of the original picture that is added to after it is extracted reach the purpose of clear and definite face.
Video standard signal is input to human face recognizing monitor, after in human face recognizing monitor, passing through the analog video signal digitizing, detect the Video Detection algorithm by special somebody's face, detect and contain somebody's face picture, then this person's face picture is extracted, be superimposed to the assigned address of original picture, the picture after will superposeing is then exported with the mock standard video mode.The benefit of this design is, this product directly can be inserted in the current video detection system, and need not existing equipment is upgraded, and energy and existing digital video monitor system carry out seamless integrated, the investment of farthest saving the user.
Description of drawings
Fig. 1 is the overall calcspar of built-in human face recognizing monitor according to an embodiment of the invention;
Fig. 2 is the hardware structure diagram of built-in human face recognizing monitor according to an embodiment of the invention;
Fig. 3 is the concrete structure figure of front panel, rear panel and the upper cover plate of built-in human face recognizing monitor according to an embodiment of the invention;
Fig. 4 is the particular flow sheet of the detection method of built-in human face recognizing monitor according to an embodiment of the invention.
Fig. 5 is the vision signal flow process synoptic diagram of built-in human face recognizing monitor according to an embodiment of the invention.
Embodiment
As shown in Figure 1, show the overall calcspar of built-in human face recognizing monitor according to an embodiment of the invention;
Described built-in human face recognizing monitor comprises front panel, rear panel, printed board and shell four parts, and its position relation as shown in the figure.
In Fig. 2, describe the hardware structure diagram of human face recognizing monitor of the present invention in detail, wherein DM642 is as the core component of whole human face recognizing monitor, it mainly finishes the control and the computing of total system, finishes the work such as digitizing of control, System self-test processing to each parts of periphery, people's face detection calculations, vision signal.
Video encoder is digitized into the data image signal that DSP can discern and handle with the normal video input signal under the control of DSP; Owing to can carry out the two-path video signal Processing, so in structural drawing, contain two video encoders.
Video Decoder is under the control of DSP, with the picture digital simulationization after carrying out the detection of people's face, extract and superposeing.Because computing is digitized picture signal entirely in DSP, after stack, need to convert it into the normal video simulating signal, so that next stage equipment directly uses.
Whether the System self-test circuit is finished the self check work of total system, and the recognition of face detecting device will move self-check program when powering on, working properly to judge total system.Self-checking circuit then cooperates self-check program to finish above-mentioned work.Simultaneously, self-checking circuit gives the user with prompting in the mode of light emitting diode when detecting fault.
SDRAM and FLASH belong to the storage device of whole human face recognizing monitor.SDRAM uses as the inside memory of dsp operation program, and what FLASH stored is application code, and the somebody of institute face programmed algorithm and self-check program etc. all leave in inside the FLASH.
Jtag interface can be used to the down load application program to FLASH as DSP and the outside interface that is connected, and also can hold under the software control at PC and use as program debug.
Working power circuit provides the various working powers of multichannel human face recognizing monitor.
Referring to Fig. 3, legend shows the concrete structure figure of front panel, rear panel and the upper cover plate of built-in human face recognizing monitor according to an embodiment of the invention;
Described front panel comprises power switch, pilot lamp and toggle switch, and rear panel comprises network interface input/output end port and power supply input port, in addition, also shows the structure of upper cover plate.
At last,, describe the software flow of described human face recognizing monitor in detail, specifically describe as follows referring to Fig. 4:
S1: after powering on, start the power-on self-test measuring program, self-check program detects all functions module of human face recognizing monitor, if the fault of discovery is then carried out the S2 operation; Otherwise provide indicating fault, the S1 that circulates simultaneously is till self check is normal.
S2: after self check is normal, carries out the user trace routine is set.The user is provided with the outer setting situation that trace routine detects the user, and content comprises: big picture whether illumination compensation, little picture whether illumination compensation, whether show little picture, little picture people's face detect, people's face extracts the picture size setting, people's face extracts the picture superposed positions and is provided with etc.Detect the back and be provided with to preset various work registers, with sign as the down-stream operation according to the user.Move S3 then
S3: will enter the Video Detection program after the user is provided with trace routine and has moved, whether program will constantly detect the vision signal input, if there is not the vision signal input, then constantly moves the Video Detection program; If the vision signal input is arranged, then with behind the video signal digitization, according to user's setting, illumination compensation then carries out illumination compensation and handles if desired, proceeds to S4 then.
S4: operation people face detection algorithm, this step is the core of whole procedure.People's face trace routine will be carried out people's face according to the data that training draws and be detected, and have people's face if detect, and then move S5, otherwise directly move S6.
S5: detected people's face is extracted, and overlay drafting size and the position setting that is provided with according to the user simultaneously is added to people's face picture on the former big picture.
S6: this moment, picture signal was a digital image signal still, so need DSP control peripheral chip to change digital image signal the output of into standard analog vision signal.
S7: after the one-period operation is finished, carry out the Video Detection of next cycle, carry out S3.
After having described hardware structure diagram and software flow, describe specific implementation process of the present invention in detail, detailed process is as follows:
1, power on and System self-test: after human face recognizing monitor powered on, system software carried out self check to itself, and whether detect each parts work normal.If the discovery fault, then working routine provides fault alarm (realizing by the LED indication), and constantly detection disappears up to fault.If do not find fault, then carry out next step operation.
2, system software is provided with according to the user each parts of system and carries out initialization, and exposure compensating is handled, people's face detects work such as back superposed positions setting, video coding chip initialization, video decoding chip initialization, network communication interface initialization as people's face.
3, obtain view data:, the video image of input is extracted with the form of frame by operation to video coding chip.
4, light is handled: system adopts the average and the variance of pixel value to correct the influence of light to each pixel of image, reaches with this and eliminates the purpose of light to the influence of people's face pixel value.
5, people's face detects: carry out people's face by people's face detection algorithm and detect.When native system has the image of a plurality of people's faces in processing, only detect behind the facial image and just no longer detect.
6, signal extraction of people's face and stack: detected people's face picture is extracted, according to the user the upper left corner, the lower left corner, the upper right corner and lower right corner assigned address to its original picture that is added to are set again, the size of people's face picture also can be provided with by the user.
Describe software algorithm of the present invention below in detail, it has drawn state-of-the-art in the world at present theoretical result, has added the learning algorithm of the uniqueness of oneself simultaneously again.
Its training study algorithm is as follows:
1, provides a systematic learning sample (x
1, y
1), (x
2, y
2) ..., (x
n, y
n), y wherein
iIt is non-face sample for=0 expression, y
i=1 its behaviour face sample of expression.
2、w
1，i＝D(i)；
3, to t=1 ..., T
(1) advanced line parameter normalization,
(2) to each face feature f, train a classification function h (x, f, p, θ); Calculate the weighting (q of the classification function of corresponding all face features
t) error rate is σ
f:
(3) choose best classification function h
t(x) (have minimal error rate σ
t)
(4), adjust weight according to this optimal classification function:
E wherein
i=0 expression x
iCorrectly classified e
i=1 expression x
iClassified mistakenly.
3, the classification function that forms at last is:
C (x)=1, when
The time
C (x)=0, when
The time
After adopting the method study, all kinds of people's faces under the detection of complex condition preferably, as: people's face of different macroscopic featuress, people's face with abundant expression, by people's face that exterior object blocks, people's face of imaging angle height rotation and the people's face under various illumination conditions.
Innovation in this detection software is, by the various face features of dynamic adjustment shared weight in whole people's face detects, dynamically remove the influence of irrelevant background image, can reach the purpose of the people's face that detects preferably under the various conditions, the time complexity and the space complexity of computing are all reduced greatly, both improved detection efficiency, reduced requirement again hardware.
Exemplary embodiments of the present invention has been described.Yet, being understandable that not breaking away from the spirit and scope of the present invention can make various changes, it all falls in the follow-up claim institute restricted portion.
Claims (1)
1. embedded human face detecting method that utilizes built-in human face recognizing monitor is characterized in that:
Described built-in human face recognizing monitor comprises:
The DSP core processing unit, it mainly finishes the control and the computing of total system as the core component of human face recognizing monitor, finishes the digitizing work of control, System self-test processing to each parts of periphery, people's face detection calculations, vision signal;
Video encoder under the control of DSP, is digitized into the data image signal that DSP can discern and handle with the normal video input signal, and the video encoder number is 2, can carry out the two-path video signal Processing;
Video Decoder, under the control of DSP, with carry out people's face detect, extract and stack after the picture digital simulationization, will be in DSP after the stack computing be that digitized picture signal changes into the normal video simulating signal entirely so that next stage equipment directly uses;
Memory device is used to store dsp operation program and application code, the somebody of institute face programmed algorithm and self-check program; Memory device comprises SDRAM and FLASH, and SDRAM uses as the inside memory of dsp operation program, and what FLASH stored is application code, and the somebody of institute face programmed algorithm and self-check program leave the FLASH the inside in;
Interface unit is used to realize being connected of DSP and codec, memory device and external unit, and wherein jtag interface is as DSP and the outside interface that is connected, is used for the down load application program to hold under the software control to FLASH or at PC and use as program debug;
The System self-test circuit, finish the self check work of total system, built-in human face recognizing monitor will move self-check program when powering on, to judge whether total system is working properly, the System self-test circuit then cooperates self-check program to finish above-mentioned self check work, the System self-test circuit gives the user with prompting in the mode of light emitting diode when detecting fault;
Working power circuit, it provides the various working powers of multichannel human face recognizing monitor;
Described embedded human face detecting method may further comprise the steps:
1. power on and System self-test: after built-in human face recognizing monitor powered on, system software carried out self check to system, and whether detect each parts work normal, if the discovery fault, fault alarm then, and constantly detect and disappear up to fault, if do not find fault, then execution in step 2.;
2. system software is provided with according to the user each parts of system and carries out initialization: handle, people's face detects back superposed positions setting, video encoder initialization, Video Decoder initialization, network communication interface initialization by exposure compensating for people's face;
3. obtain view data:, the video image of input is extracted with the form of frame by operation to video encoder;
4. light is handled: system adopts the average and the variance of pixel value to correct the influence of light to each pixel of image;
5. people's face detects: carry out people's face by people's face detection algorithm and detect;
6. signal extraction of people's face and stack: detected people's face picture is extracted, according to the user the upper left corner, the lower left corner, the upper right corner and lower right corner assigned address with its original picture that is added to are set again;
7. digital image signal is changed into the output of standard analog vision signal.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2007101519097A CN101201895B (en) | 2007-09-20 | 2007-09-20 | Built-in human face recognizing monitor and detecting method thereof |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2007101519097A CN101201895B (en) | 2007-09-20 | 2007-09-20 | Built-in human face recognizing monitor and detecting method thereof |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN101201895A CN101201895A (en) | 2008-06-18 |
| CN101201895B true CN101201895B (en) | 2010-06-02 |
Family
ID=39517054
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2007101519097A Expired - Fee Related CN101201895B (en) | 2007-09-20 | 2007-09-20 | Built-in human face recognizing monitor and detecting method thereof |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101201895B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101957909B (en) * | 2009-07-15 | 2012-09-05 | 青岛科技大学 | Digital signal processor (DSP)-based face detection method |
| CN103220471A (en) * | 2013-03-26 | 2013-07-24 | 苏州福丰科技有限公司 | Face video detection overlaying device |
| CN103607554B (en) * | 2013-10-21 | 2017-10-20 | 易视腾科技股份有限公司 | It is a kind of based on full-automatic face without the image synthesizing method being stitched into |
| CN104751197A (en) * | 2015-04-22 | 2015-07-01 | 安徽金赛弗信息技术有限公司 | Device and method for recognizing faces of drivers during vehicle running on basis of video analysis |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1411277A (en) * | 2001-09-26 | 2003-04-16 | Lg电子株式会社 | Video-frequency communication system |
| CN1428694A (en) * | 2001-12-29 | 2003-07-09 | 成都银晨网讯科技有限公司 | Embedded human face automatic detection equipment based on DSP and its method |
| CN2718676Y (en) * | 2004-06-09 | 2005-08-17 | 上海银晨智能识别科技有限公司 | Embedded identifier for human face |
| CN101025790A (en) * | 2007-01-15 | 2007-08-29 | 中北大学 | Image-based bus passenger number automatic statistics meter |
-
2007
- 2007-09-20 CN CN2007101519097A patent/CN101201895B/en not_active Expired - Fee Related
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1411277A (en) * | 2001-09-26 | 2003-04-16 | Lg电子株式会社 | Video-frequency communication system |
| CN1428694A (en) * | 2001-12-29 | 2003-07-09 | 成都银晨网讯科技有限公司 | Embedded human face automatic detection equipment based on DSP and its method |
| CN2718676Y (en) * | 2004-06-09 | 2005-08-17 | 上海银晨智能识别科技有限公司 | Embedded identifier for human face |
| CN101025790A (en) * | 2007-01-15 | 2007-08-29 | 中北大学 | Image-based bus passenger number automatic statistics meter |
Also Published As
| Publication number | Publication date |
|---|---|
| CN101201895A (en) | 2008-06-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105137887B (en) | The material objectization programmed method of plate and its application in robot field based on programming | |
| Matusiak et al. | Object recognition in a mobile phone application for visually impaired users | |
| CN101201895B (en) | Built-in human face recognizing monitor and detecting method thereof | |
| CN104850542B (en) | Non-audible voice input correction | |
| CN107635147A (en) | Health information management TV based on multi-modal man-machine interaction | |
| Conti et al. | Accelerated visual context classification on a low-power smartwatch | |
| CN201397512Y (en) | Embedded-type infrared human face image recognition device | |
| CN106887059A (en) | A kind of intelligent electronic lock system based on face recognition | |
| CN101923433A (en) | Man-computer interaction mode based on hand shadow identification | |
| CN202584048U (en) | Smart mouse based on DSP image location and voice recognition | |
| US20090179887A1 (en) | Digital photo frame with a power saving function and a power saving method thereof | |
| CN103019381B (en) | Control the method for display screen automatic backlight | |
| US20210393141A1 (en) | Method and device for measuring human body temperature | |
| CN207096984U (en) | A kind of chemical illumination immunity analysis instrument inspection data inquiry unit | |
| Sareeka et al. | pseudoEye—Mobility assistance for visually impaired using image recognition | |
| CN208298198U (en) | A kind of optical finger print identification mould group | |
| CN105468095B (en) | A kind of angle detecting method and electronic equipment | |
| JP2019028968A (en) | Biological information analyzer capable of marking cheek rouge region | |
| CN203102318U (en) | Iris information acquisition and identification device | |
| CN209821912U (en) | No screen face identification intelligence lock | |
| KR20220027241A (en) | Motion recognition method, device and storage medium | |
| CN211691896U (en) | Intelligent lock | |
| CN219147548U (en) | Household hand-held tongue health information acquisition device | |
| CN111210438A (en) | Mirror | |
| CN116092129B (en) | Intelligent bookshelf and control method thereof |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| ASS | Succession or assignment of patent right |
Owner name: GAO YUXIN Free format text: FORMER OWNER: GAO YUXIN; APPLICANT Effective date: 20080926 |
|
| C41 | Transfer of patent application or patent right or utility model | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20080926 Address after: Beijing City, Chaoyang District Road No. 16 Baiziwan 100 yuan (urban) No. 4 Building A Room 601 - 602 post encoding: 100022 Applicant after: Gao Yuxin Address before: A5-3005 room, Fuli City, Beijing, Chaoyang District: 100022 Applicant before: High education Co-applicant before: Hao Xiaogu |
|
| C57 | Notification of unclear or unknown address | ||
| DD01 | Delivery of document by public notice |
Addressee: Gao Yuxin Document name: Notification of Passing Examination on Formalities |
|
| C41 | Transfer of patent application or patent right or utility model | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20090327 Address after: Room 1101, Dadi science building, No. 67 West Fourth Ring Road, Beijing, Haidian District, China: 100080 Applicant after: Beijing Qingda Weisen Technology Co., Ltd. Address before: Beijing City, Chaoyang District Road No. 16 Baiziwan 100 yuan (urban) No. 4 Building A Room 601 - 602 post encoding: 100022 Applicant before: Gao Yuxin |
|
| ASS | Succession or assignment of patent right |
Owner name: BEIJING QINGDA WEISEN SCIENCE AND TECHNOLOGY CO., Free format text: FORMER OWNER: GAO YUXIN Effective date: 20090327 |
|
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| EE01 | Entry into force of recordation of patent licensing contract |
Assignee: Jiangsu Tsingda Vision Technology Co., Ltd. Assignor: Beijing Qingda Weisen Technology Co., Ltd. Contract record no.: 2012320000306 Denomination of invention: Built-in human face recognizing monitor and detecting method thereof Granted publication date: 20100602 License type: Exclusive License Open date: 20080618 Record date: 20120326 |
|
| DD01 | Delivery of document by public notice |
Addressee: Gao Yuxin Document name: Notification of Passing Examination on Formalities |
|
| ASS | Succession or assignment of patent right |
Owner name: SHANXI ZHONGZI BAIJIA IOT TECHNOLOGY CO., LTD. Free format text: FORMER OWNER: BEIJING QINGDA VISION TECHNOLOGY CORP. Effective date: 20150701 |
|
| C41 | Transfer of patent application or patent right or utility model | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20150701 Address after: High tech Zone of Shanxi province Taiyuan City Xuefu street 030006 No. 122 Kay building room 9921A Patentee after: Shanxi ever since the Internet of things Technology Co., Ltd. Address before: 100080, room 1101, Dadi science and technology building, No. 67 West Fourth Ring Road, Haidian District, Beijing Patentee before: Beijing Qingda Weisen Technology Co., Ltd. |
|
| 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: 20100602 Termination date: 20170920 |