CN109740577A - A kind of real-time face based on raspberry pie identifies camera system and its adjustment method again - Google Patents
A kind of real-time face based on raspberry pie identifies camera system and its adjustment method again Download PDFInfo
- Publication number
- CN109740577A CN109740577A CN201910149786.6A CN201910149786A CN109740577A CN 109740577 A CN109740577 A CN 109740577A CN 201910149786 A CN201910149786 A CN 201910149786A CN 109740577 A CN109740577 A CN 109740577A
- Authority
- CN
- China
- Prior art keywords
- camera
- raspberry pie
- face
- real
- video
- 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
Landscapes
- Image Analysis (AREA)
- Alarm Systems (AREA)
- Image Processing (AREA)
Abstract
The invention discloses a kind of real-time faces based on raspberry pie to identify camera system and its adjustment method again, including raspberry pie system and camera, camera is connected with raspberry pie development board, raspberry pie system one end connects camera, the other end passes through network interface connection internet, raspberry pie development board embeds video identification program, the pictorial information training program for not needing alarm of user setting is acquired simultaneously, when work, raspberry pie development board acquires video information by camera, the video information of identification camera input is gone by trained program, judge whether to identify, appearance cannot identify to obtain personage or when image information, warning message is sent to user by network interface.Effective warning message can be transmitted in the present invention, saves the time and efforts of user, solves the problems, such as that privacy of existing video monitoring apparatus during monitoring is easy leakage.
Description
Technical field
The invention belongs to technical field of video monitoring, and in particular to a kind of real-time face based on raspberry pie identifies camera shooting again
System and its adjustment method.
Background technique
With the development of technology, Levels of Social Economic Development improves, and rhythm of life is accelerated, demand of the people to safety precaution
Also increase year by year, ensure that family old man, infant and property safety etc. have become Modern Family and mainly state and ask, and people are safe
The increase of consciousness and intelligentized high speed development, household safety-protection market emerge therewith.
It is shown according to public data, existing about 200,000,000 family of urban household of China, the family of the five-year estimated at least 5%
It can consider the product for installing intelligent video camera head at home, will there is 2,000,000,000 yuan or so of the market demand every year on average.Moreover, current state
Interior domestic security market entirety accounting is only 6% or so, and far below the 10% of global average level, growth space is huge；In addition
Smart home is set with the influence in market, and it is latent that the home intelligent camera that future takes into account security protection performance undoubtedly has huge development
Power.
Compared with traditional camera, intelligent video camera head has all obtained very big upgrading in terms of software and hardware, in hardware aspect
Realize the functions such as high-resolution, low-power consumption, night market, dynamic monitoring and wireless, miniaturization；Software aspects gradually to low cost and
Intelligentized embedded plan development.Simultaneously intelligent monitor system also with the technologies such as long-range monitoring, HD video transmission, Internet of Things
It combines, already becomes the middle control of security system in family.When intelligent video camera head is recognizing stranger and swarms into house
Trigger automatic alarm；Some intelligent video camera heads even have the partial function of voice assistant, are meeting the same of people's demand for security
When, greatly facilitate daily life.
But current intelligent video camera head can only be realized and be detected by video analysis to the object moved in video, and from
It is dynamic to send warning message, need user can just judge whether it is because of stranger by the video of viewing one section of camera storage
Swarm into pet, the old man, child that the information triggered in the family of oneself is still in alone；Intelligent video camera head can not independently judge
Accurate warning message, this will lead to the transmission of invalid warning message, waste a large amount of time and efforts of user.Except this it
Outside, current intelligent video camera head safety problem occurs frequently, and the use of camera needs to pass through handset binding and downloads in intelligence system
The app that relevant manufacturers provide, is inputted, some criminals can easily be logged on to binding by hacking technique and be set by account number cipher
It is standby to steal personal video data, cause privacy leakage.
It is therefore desirable to carry out upgrading to existing video monitoring, privacy of user is protected while monitoring to realize, is obtained
Obtain better using effect and more preferably practicability.
Summary of the invention
The technical problem to be solved by the present invention is to solve the above shortcomings of the prior art and to provide a kind of based on raspberry pie
Real-time face identifies camera system and its adjustment method again.
To realize the above-mentioned technical purpose, the technical scheme adopted by the invention is as follows:
A kind of real-time face based on raspberry pie identifies camera system, including raspberry pie system and camera again；Camera and tree
The certain kind of berries sends development board to be connected, and raspberry pie system one end connects camera, and the other end passes through network interface connection internet, raspberry pie development board
Embedded video identification program, while acquiring the pictorial information training program for not needing alarm of user setting, when work, raspberry pie
Development board acquires video information by camera, and the video information of identification camera input, judgement are gone by trained program
Whether can identify, appearance cannot identify to obtain personage or when image information, send warning message to user by network interface.
To optimize above-mentioned technical proposal, the concrete measure taken further include:
Above-mentioned raspberry pie is 3 generation B+ models, and the camera is the mini-camcorder of 500w pixel 1080p sensor.
A kind of real-time face based on raspberry pie identifies the adjustment method of camera system again, includes the following steps:
Step1 downloads operating system in raspberry pie, configures raspberry pie；
Step2, debugs the camera module of raspberry pie, and test camera is in available mode；
Step3 creates a human-face detector in raspberry pie, and real-time face detection can be realized in the video that camera is shot,
Return to the rectangle frame that face is marked；
Step4 creates a subdirectory and is used to store facial sample, by the people of user's understanding or frequently appears in user
The people of family carries out face sample collection；
Step5, training face identifier, obtains trained recognition of face device；
Step6 calls camera to detect again, encounters the face transmission warning message for failing detection.
In above-mentioned Step3, Face datection specifically: create a human-face detector with OpenCV, use haar feature
Cascade classifier test object, by calling OpenCV, trained classifier carries out Face datection in advance.
The inside circulation that camera is arranged in above-mentioned calling process loads input video with grayscale mode, and to classifier
Function passes related coefficient, the proportionality coefficient of setting twice sweep search window are 1.2；Detection target adjacent rectangle number be
5；The minimum dimension of the face detected is 20*20.
In above-mentioned Step5, recognition of face device training method are as follows: the partial binary in OpenCV is called to encode histogram
(LBPH) algorithm carries out recognition of face as origin identifier, and the facial sample of collection is input in origin identifier and is instructed
Get the recognition of face device of id in recognizable database.
Above-mentioned Step6 specifically: detect the video of camera shooting, Haar again with trained identifier
Cascade detection of classifier is to facial information and returns to an id parameter, and the id parameter is input in trained identifier
It is matched, finally shows id and confidence level in video, a Unknown Label is obtained if it can not match, it will screenshot saves
And it generates warning message and is sent to user.
The invention has the following advantages:
It (1) is alone using after the solution of the present invention, enable that camera removes judgement triggering warning message by self training
The pet, child, old man or the stranger that are in and issue authentic and valid warning message, do not need user by checking view
Frequency judges, the time and efforts of user is greatly saved, keeps camera more intelligent.
(2) by using raspberry pie, the app or other bound devices that downloading producer provides is not needed, is considerably increased
The safety that camera uses, avoids criminal from stealing video data, causes privacy leakage.
(3) raspberry pie can realize that real-time face is identified and alarmed again, and user can take emergency measures reduction at the first time
The loss of property.
Detailed description of the invention
Fig. 1 is the composition block diagram of present system.
Fig. 2 is the flow chart of the method for the present invention.
Specific embodiment
The embodiment of the present invention is described in further detail below in conjunction with attached drawing.
Referring to Fig. 1, a kind of real-time face based on raspberry pie of the invention identifies camera system, including raspberry pie system again
And camera；Camera is connected with raspberry pie development board, and raspberry pie system one end connects camera, and the other end passes through network interface connection
Internet, raspberry pie development board embed video identification program, while acquiring the pictorial information instruction for not needing alarm of user setting
Practice program, when work, raspberry pie development board acquires video information by camera, goes to identify camera by trained program
The video information of input judges whether to identify, appearance cannot identify to obtain personage or when image information, passes through network interface and sends report
Alert information is to user.
In embodiment, raspberry pie is 3 generation B+ models, and camera is the mini-camcorder of 500w pixel 1080p sensor.
Referring to fig. 2, a kind of real-time face based on raspberry pie identifies the adjustment method of camera system again, including walks as follows
It is rapid:
Step1 downloads operating system in raspberry pie, then configures raspberry pie；The present invention recommends using raspberry pie official website
Mirror image raspbi-an.Raspberry pie is configured after downloading mirror image, starts camera, the functions such as ssh, VNC.In windows system
The input address raspberry pie ip downloading putty remotely connects raspberry pie by ssh and is operated on system.Pass through command download in terminal
Python3.5 and OpenCV environment.
Step2, debugs the camera module of raspberry pie, and test camera is in available mode；Whether test camera is in can
With state, it must assure that camera keeps available mode in following all operating process.
Step3 creates a human-face detector in raspberry pie, can realize real-time face in the video that camera is shot
Detection returns to the rectangle frame that face is marked；
In embodiment, a human-face detector is created with OpenCV, uses the cascade classifier test object of haar feature.Pass through
Calling OpenCV, trained classifier carries out Face datection in advance.Need to be arranged the inside circulation of camera in calling process,
Input video is loaded with grayscale mode, and related coefficient must be transmitted to classifier functions, twice sweep search window is set
Proportionality coefficient be 1.2；The number for detecting target adjacent rectangle is 5；The minimum dimension of the face detected is 20*20.Operation
Real-time face detection can be realized after program in the video that camera is shot, returns to the rectangle frame that face is marked.
Step4 creates a subdirectory and is used to store facial sample, the people or frequently appear in of user's understanding is made
The people of user's family carries out face sample collection；
Step5, training face identifier, obtains trained recognition of face device；
In embodiment, using user understanding people or frequently appear in user's family people carry out face sample collection as
We train the database of identifier, everyone captures 30 samples as an id, each id；
The partial binary coding histogramming algorithm in OpenCV is called to carry out recognition of face as origin identifier, by collection
Facial sample is input to the recognition of face device for being trained to obtain in origin identifier and can recognize id in database.
Step6 calls camera to detect again, encounters the face transmission warning message for failing detection；
In embodiment, the video of camera shooting is detected again with trained identifier, Haar Cascade classifier is still
It can detect face and return, but not return to a rectangle frame directly on video, but return to an id parameter and input
It is matched into our trained identifiers, finally shows id and confidence level in video, obtain one if it can not match
A Unknown Label, it will screenshot, which saves and generates warning message, is sent to user.
The above is only the preferred embodiment of the present invention, protection scope of the present invention is not limited merely to above-described embodiment,
All technical solutions belonged under thinking of the present invention all belong to the scope of protection of the present invention.It should be pointed out that for the art
For those of ordinary skill, several improvements and modifications without departing from the principles of the present invention should be regarded as protection of the invention
Range.
Claims (7)
1. a kind of real-time face based on raspberry pie identifies camera system again, it is characterised in that: including raspberry pie system and camera shooting
Head；The camera is connected with raspberry pie development board, and raspberry pie system one end connects camera, and the other end is mutual by network interface connection
Networking, raspberry pie development board embed video identification program, while acquiring the pictorial information training for not needing alarm of user setting
Program, when work, raspberry pie development board acquires video information by camera, goes identification camera defeated by trained program
The video information entered judges whether to identify, appearance cannot identify to obtain personage or when image information, passes through network interface and sends alarm
Information is to user.
2. a kind of real-time face based on raspberry pie according to claim 1 identifies camera system again, it is characterised in that: institute
Stating raspberry pie is 3 generation B+ models, and the camera is the mini-camcorder of 500w pixel 1080p sensor.
3. a kind of real-time face based on raspberry pie according to claim 1 identifies the adjustment method of camera system again,
It is characterized in that:
Include the following steps:
Step1 downloads operating system in raspberry pie, configures raspberry pie；
Step2, debugs the camera module of raspberry pie, and test camera is in available mode；
Step3 creates a human-face detector in raspberry pie, and real-time face detection can be realized in the video that camera is shot,
Return to the rectangle frame that face is marked；
Step4 creates a subdirectory and is used to store facial sample, by the people of user's understanding or frequently appears in user
The people of family carries out face sample collection；
Step5, training face identifier, obtains trained recognition of face device；
Step6 calls camera to detect again, encounters the face transmission warning message for failing detection.
4. a kind of real-time face based on raspberry pie according to claim 3 identifies the adjustment method of camera system again,
It is characterized in that: Face datection described in Step3 specifically: create a human-face detector with OpenCV, use the grade of haar feature
Join detection of classifier object, trained classifier carries out Face datection in advance by calling OpenCV.
5. a kind of real-time face based on raspberry pie according to claim 4 identifies the adjustment method of camera system again,
Be characterized in that: the inside circulation that camera is arranged in the calling process loads input video with grayscale mode, and to classifier
Function passes related coefficient, the proportionality coefficient of setting twice sweep search window are 1.2；Detection target adjacent rectangle number be
5；The minimum dimension of the face detected is 20*20.
6. a kind of real-time face based on raspberry pie according to claim 5 identifies the adjustment method of camera system again,
It is characterized in that: recognition of face device training method described in Step5 are as follows: the partial binary in OpenCV is called to encode histogramming algorithm
Recognition of face is carried out as origin identifier, the facial sample of collection is input in origin identifier and is trained to obtain and can know
The recognition of face device of id in other database.
7. a kind of real-time face based on raspberry pie according to claim 6 identifies the adjustment method of camera system again,
It is characterized in that: Step6 specifically: detect the video of camera shooting again with trained identifier, Haar Cascade divides
Class device detects facial information and returns to an id parameter, and the id parameter, which is input in trained identifier, to be matched,
Id and confidence level are finally shown in video, a Unknown Label is obtained if it can not match, it will screenshot saves and generates report
Alert information is sent to user.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910149786.6A CN109740577A (en) | 2019-02-28 | 2019-02-28 | A kind of real-time face based on raspberry pie identifies camera system and its adjustment method again |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910149786.6A CN109740577A (en) | 2019-02-28 | 2019-02-28 | A kind of real-time face based on raspberry pie identifies camera system and its adjustment method again |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109740577A true CN109740577A (en) | 2019-05-10 |
Family
ID=66368780
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910149786.6A Pending CN109740577A (en) | 2019-02-28 | 2019-02-28 | A kind of real-time face based on raspberry pie identifies camera system and its adjustment method again |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109740577A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110232347A (en) * | 2019-06-06 | 2019-09-13 | 上海海事大学 | A kind of Intelligent roll calling face identification device |
| CN110457987A (en) * | 2019-06-10 | 2019-11-15 | 中国刑事警察学院 | Face identification method based on unmanned plane |
| CN110569827A (en) * | 2019-09-28 | 2019-12-13 | 华南理工大学 | Face recognition reminding system based on convolutional neural network |
| CN113221612A (en) * | 2020-11-30 | 2021-08-06 | 南京工程学院 | Visual intelligent pedestrian monitoring system and method based on Internet of things |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103761514A (en) * | 2014-01-26 | 2014-04-30 | 公安部第三研究所 | System and method for achieving face recognition based on wide-angle gun camera and multiple dome cameras |
| CN104780308A (en) * | 2014-01-09 | 2015-07-15 | 联想(北京)有限公司 | Information processing method and electronic device |
| CN107613272A (en) * | 2017-11-13 | 2018-01-19 | 惠州学院 | A kind of safety defense monitoring system based on Raspberry Pi |
| CN207319328U (en) * | 2017-10-18 | 2018-05-04 | 浙江泰科力合科技有限公司 | A kind of intelligent bullet management control system for realizing Face datection |
| CN108958053A (en) * | 2018-07-20 | 2018-12-07 | 杭州电子科技大学 | Wechat long-distance monitoring method based on raspberry pie |
-
2019
- 2019-02-28 CN CN201910149786.6A patent/CN109740577A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104780308A (en) * | 2014-01-09 | 2015-07-15 | 联想(北京)有限公司 | Information processing method and electronic device |
| CN103761514A (en) * | 2014-01-26 | 2014-04-30 | 公安部第三研究所 | System and method for achieving face recognition based on wide-angle gun camera and multiple dome cameras |
| CN207319328U (en) * | 2017-10-18 | 2018-05-04 | 浙江泰科力合科技有限公司 | A kind of intelligent bullet management control system for realizing Face datection |
| CN107613272A (en) * | 2017-11-13 | 2018-01-19 | 惠州学院 | A kind of safety defense monitoring system based on Raspberry Pi |
| CN108958053A (en) * | 2018-07-20 | 2018-12-07 | 杭州电子科技大学 | Wechat long-distance monitoring method based on raspberry pie |
Non-Patent Citations (1)
| Title |
|---|
| SPOONY: "用树莓派实现实时的人脸检测", pages 1 - 24, Retrieved from the Internet <URL:https://shumeipai.nxez.com/2018/03/09/real-time-face-recognition-an-end-to-end-project-with-raspberry-pi.html?variant=zh-cn> * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110232347A (en) * | 2019-06-06 | 2019-09-13 | 上海海事大学 | A kind of Intelligent roll calling face identification device |
| CN110457987A (en) * | 2019-06-10 | 2019-11-15 | 中国刑事警察学院 | Face identification method based on unmanned plane |
| CN110569827A (en) * | 2019-09-28 | 2019-12-13 | 华南理工大学 | Face recognition reminding system based on convolutional neural network |
| CN110569827B (en) * | 2019-09-28 | 2024-01-05 | 华南理工大学 | Face recognition reminding system based on convolutional neural network |
| CN113221612A (en) * | 2020-11-30 | 2021-08-06 | 南京工程学院 | Visual intelligent pedestrian monitoring system and method based on Internet of things |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109740577A (en) | A kind of real-time face based on raspberry pie identifies camera system and its adjustment method again | |
| US11354993B2 (en) | Image surveillance and reporting technology | |
| US9386281B2 (en) | Image surveillance and reporting technology | |
| CN206515931U (en) | A kind of face identification system | |
| AU2019295856B2 (en) | Object tracking using disparate monitoring systems | |
| US10867217B1 (en) | Fusion of visual and non-visual information for training deep learning models | |
| CN103731633A (en) | Television device capable of carrying out remote monitoring and control method of television device | |
| CN108364374A (en) | Face access control device based on deep learning and method | |
| CN102572388A (en) | Face-recognition-based network video monitoring device and monitoring recognition method | |
| CN106157505A (en) | One is illegally entered the room alarm method and device | |
| CN105139563A (en) | Intelligent door and control method thereof | |
| CN107426533A (en) | A kind of video monitoring image recognition system based on video-encryption compression and image identification | |
| Lee et al. | A multilevel home security system (mhss) | |
| CN106780875A (en) | A kind of recognition of face gate inhibition integrated apparatus and its recognition methods | |
| CN210110004U (en) | Oil field behavior monitoring system based on artificial intelligence | |
| CN105245836A (en) | Monitoring system for security door | |
| CN109841034A (en) | A kind of intelligent monitor system based on Internet of Things | |
| CN105450992A (en) | Smart home monitoring method and system based on motion detection | |
| CN111400415B (en) | Personnel management method and related device | |
| CN104463114A (en) | Method for catching images and quickly recognizing targets and embedded device | |
| CN115065812B (en) | Real-time monitoring method based on user behavior and related equipment | |
| CN105094068A (en) | Intelligent monitoring system for buildings | |
| Hang et al. | Design and implementation of intelligent fire notification service using IP camera in smart home | |
| CN110886571A (en) | Intelligent safety door | |
| CN205140103U (en) | Intelligence door |
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 |