CN109871797A - A face recognition and remote intelligent remote control method for Yuanwei shared kitchen - Google Patents
A face recognition and remote intelligent remote control method for Yuanwei shared kitchen Download PDFInfo
- Publication number
- CN109871797A CN109871797A CN201910103288.8A CN201910103288A CN109871797A CN 109871797 A CN109871797 A CN 109871797A CN 201910103288 A CN201910103288 A CN 201910103288A CN 109871797 A CN109871797 A CN 109871797A
- Authority
- CN
- China
- Prior art keywords
- face
- picture
- recognition
- raspberry pie
- raspberry
- 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)
Abstract
一种源味共享厨房的人脸识别和远程智能遥控方法，包括以下步骤：步骤1：开通树莓派SSH、VNC服务，购买能连接树莓派的usb摄像头，开通树莓派摄像头服务；使用树莓派GPIO高低电平控制继电器开关，达到操控设备的作用；步骤2：安装编译环境(python，c++)；步骤3：进行人脸识别；步骤4：配置lnmp环境，服务器端放置html+css+js+php写的Web端，手机上使用浏览器打开网页进行相关设备开关操作；步骤5：进行Web端控制系统功能的完善；步骤6：设置脚本开机自启动。本发明提供了一种智能化程度较高、可靠性良好的源味共享厨房的人脸识别和远程智能遥控方法，实现人脸识别和远程智能控制。
A face recognition and remote intelligent remote control method for Yuanwei shared kitchen, comprising the following steps: Step 1: activate Raspberry Pi SSH and VNC services, purchase a usb camera that can be connected to the Raspberry Pi, and activate the Raspberry Pi camera service; use Raspberry Pi GPIO high and low level controls the relay switch to control the device; Step 2: Install the compilation environment (python, c++); Step 3: Perform face recognition; Step 4: Configure the lnmp environment and place html+css on the server side +js+php written on the web side, use the browser on the mobile phone to open the web page to perform related device switch operations; Step 5: Improve the functions of the web-side control system; Step 6: Set the script to start automatically after booting. The invention provides a face recognition and remote intelligent remote control method for a shared kitchen with a high degree of intelligence and good reliability, so as to realize the face recognition and remote intelligent control.
Description
Technical field
The present invention relates to the communications between raspberry pie and Cloud Server, mobile phone terminal.The present invention is a kind of shared based on source taste
The recognition of face and long-distance intelligent remote control method added on cooking apparatus.
Background technique
Source taste kitchen Project-developing final goal is to realize that user places an order (deletion server give an order part)-by APP
Central kitchen allocation semi-finish cooked food and it is sent to-intelligent terminal refrigeration and switching culinary art in due course.
The terminal mechanical equipment of kitchen system is made of host and distribution box two parts, and wherein there are four independent controls for host
Working bin, four distribution boxes of arranging in pairs or groups, it is possible to provide 1 meal, 3 dish totally 16 sets of canteens.Distribution box is made of incubator and liner case, is protected
Incubator is used to guarantee preservation temperature of the vegetable in dispatching, and liner case is for placement tray and meal bowl.Liner case can be layered placement
Four pallets have multiple units on each pallet, can place the meal bowl of different model.Meal bowl is made of bowl and bowl cover, by right
The case where design of its structure can prevent taint of odour, drip.Since control robot has permission, and the people having permission is absent from the scene
When fail to open equipment take out food.
Summary of the invention
Intelligent program in order to overcome the shortcomings of existing kitchen system is lower, reliability is poor, and the present invention provides one
Kind intelligence degree is higher, the recognition of face and long-distance intelligent remote control method in the shared kitchen of the good source taste of reliability, realizes people
Face identification and long-distance intelligent control.
The technical solution adopted by the present invention to solve the technical problems is:
One provenance taste shares the recognition of face and long-distance intelligent remote control method in kitchen, comprising the following steps:
Step 1: the respbian jessie system of burning raspberry pie opens the service of raspberry pie SSH, VNC into SD card,
Purchase can connect the usb camera of raspberry pie, open raspberry pie camera service；Using raspberry pie GPIO low and high level control after
Electric switch has the function that controlling equipment；
Step 2: installation translation and compiling environment (python, c++)；
Step 3: carrying out recognition of face, to guarantee that camera photographed complete face, show that camera is clapped using led display screen
The picture arrived, is convenient for knowing others and carries out face's alignment；
Step 4: configuration lnmp environment, server end are placed the end Web that html+css+js+php writes, are used on mobile phone clear
Device of looking at opens webpage and carries out relevant device switch operation；
Step 5: carrying out the perfect of Web end control system function；
Step 6: setting script booting self-starting.
Further, in the step 3, recognition of face process is described as follows:
Step 3.1: the position of face in picture is determined using HOG --- 16x16 is divided the image into after picture gray processing
The square of pixel calculates how many gradient of principal direction in each small cube, is replaced using the strongest gradient of directive property
Original image is finally converted into a very simple HOG expression-form by small cube originally, captures facial basic structure；
Step 3.2: in order to eliminate face towards error caused by different directions, we use the calculation of facial facial feature estimation
Method finds 68 points (landmark) such as table 1 generally existing on face:
Table 1
The position at each position of face is known according to this 68 points, picture pivots about operation with nose, so that
Face location is substantially aligned, then carries out next step operation；
Step 3.3: face encodings are given, for distinguishing different faces, concrete operations are one depth convolutional network of training,
128 measured values are generated for face, network can obtain almost the same value when the photo of two same persons is as input.Tool
Body is video under typing state using frame as unit screenshot by configuring openface program, training dataset in raspberry pie,
It carrying out recognition of face and the picture of Video segmentation is then inputted into training pattern, output test value is similar with the picture recognition of script, depending on
Frequency identification is exactly a large amount of picture recognition in fact, and the frame screenshot of video is stored under same file folder by the present invention, when identification is tied
Picture is deleted after beam, also screenshot is left out after same recognition of face detection, prevents picture from excessively causing raspberry pie can not
128 measured values of different people's training results are stored in cloud database by storage to identify；
Step 3.4: distinguishing different people 128 by personage in 128 measured values of measurement of comparison result and database
The Euclidean distance of measured value is compared, it is contemplated that volume will not be arranged in equipment rights management personnel, so this process
Performance raspberry pie can bear completely, one threshold value of default then thinks the non-same person more than threshold value, and authentication authorization and accounting loses
It loses, Euclidean distance is defined as follows d (x, y)
Wherein, xiRepresent the ith measurement value that instrument detects face, yiRepresent comparison people's image data in database
Ith measurement value.
In the present invention, using the raspberry pie operation associated depth learning program training facial image of connection camera, pass through
The video taken, which is cut into frame and identify, reaches face identification functions, while being communicated by socket and Cloud Server,
Cloud Server is communicated by http agreement with mobile phone mobile terminal.
Beneficial effects of the present invention are mainly manifested in: carrying out the deep learning of intelligent terminal by raspberry pie and items are counted
It calculates, control, since required calculating power is little, it is possible to complete to operate by raspberry pie, realize recognition of face and long-range intelligence
It can control.
Detailed description of the invention
Fig. 1 is that source taste shares the recognition of face in kitchen and the raspberry pie data communication process figure of long-distance intelligent remote control method.
Specific embodiment
The invention will be further described below in conjunction with the accompanying drawings.
Referring to Fig.1, a provenance taste shares the recognition of face and long-distance intelligent remote control method in kitchen, comprising the following steps:
Step 1: into SD card, (some raspberry pies are bought to be passed through the respbian jessie system of burning raspberry pie
Businessman's burning, the step for can saving), the service of raspberry pie SSH, VNC is opened, purchase can connect the usb camera of raspberry pie,
Open raspberry pie camera service.Originally small kitchen terminal mechanical equipment is using relay control switch, and the present invention is intended that with tree
The certain kind of berries sends GPIO low and high level to control relay switch, has the function that controlling equipment；
Step 2: installation translation and compiling environment (python, c++)；
Step 3: installation opencv, the method for handling video is to be split video according to frame, logical when typing information
It crosses web terminal manipulation raspberry pie to start to record face information, starts to detect by starting to record by the switch on instrument, pass through threaded tree
The certain kind of berries group infrared module detecting instrument before whether someone, when detection complete or it is infrared can not detect people station 5 seconds before instrument
After close detection function.Opencv determines the position of face in picture using HOG --- it is divided the image into after picture gray processing
The square of 16x16 pixel calculates how many gradient of principal direction in each small cube, is come using the strongest gradient of directive property
Instead of original small cube, original image is finally converted into a very simple HOG expression-form, captures the basic knot of face
Structure；
Step 4: in order to eliminate face towards error caused by different directions, we use the algorithm of facial facial feature estimation
(Vahid Kazemi and Josephine Sullivan were invented in 2014), finds 68 points generally existing on face
(landmark) such as table 1:
| Characteristic point | Chin | Left eyebrow | Right eyebrow | The bridge of the nose | Nose | Left eye | Right eye | Outer lip | Interior lip |
| Number | 0-16 | 17-21 | 22-26 | 27-30 | 31-35 | 36-41 | 42-47 | 48-59 | 60-67 |
Table 1
The position at each position of face can be substantially known according to this 68 points, picture pivots about behaviour with nose
Make, so that face location is substantially aligned, by treated, image saves the original image of covering, then carries out next step operation；
Step 5: giving face encodings, for distinguishing different faces, concrete operations are one depth convolutional network of training, are
Face generates 128 measured values, and when the photo of two same persons is as input, network can obtain almost the same value.Specifically
By configuring openface program in raspberry pie, training dataset is video under typing state using frame as unit screenshot, into
The picture of Video segmentation is then inputted training pattern by row recognition of face, and output test value is similar with the picture recognition of script, video
Identification is exactly a large amount of picture recognition in fact, and the frame screenshot of video is stored under same file folder by the present invention, works as end of identification
Picture is deleted afterwards, is also left out screenshot after same recognition of face detection, preventing picture excessively causes raspberry pie that can not deposit
128 measured values of different people's training results are stored in cloud database by storage to identify；
Step 6: distinguishing different people and pass through 128 surveys of personage in 128 measured values of measurement of comparison result and database
The Euclidean distance of magnitude is compared, it is contemplated that and volume will not be arranged in equipment rights management personnel, so this process
Performance raspberry pie can be born completely, one threshold value of default, and the non-same person is then thought more than threshold value, and authentication authorization and accounting fails,
Euclidean distance is defined as follows d (x, y)
Wherein, xiRepresent the ith measurement value that instrument detects face, yiRepresent comparison people's image data in database
Ith measurement value；
Step 7: purchase Ali's Cloud Server configures lnmp environment, and server end places what html+css+js+php write
The end Web, using browser opening webpage on mobile phone and carrying out relevant device switch operation (has unlatching to eat in script terminal mechanical equipment
Object heating cabinet and the button for starting culinary art, stopping culinary art are all relay switch operations, and operation here also includes these), the phase
Between connection relationship such as Fig. 1；
Step 8: carrying out perfect (such as user's logins, rights management etc., the progress phase of Web end control system correlation function
Close security setting).
Step 9: setting script booting self-starting；
The present invention carries out the deep learning of intelligent terminal by raspberry pie and every calculating, control centre act on, due to this
Calculating power needed for invention is little, so can complete to operate by raspberry pie substantially.
In order to make the object, technical scheme and advantages of the embodiment of the invention clearer, above embodiments implement the present invention
Technical solution in example is clearly and completely described, it is clear that and described embodiments are some of the embodiments of the present invention, and
The embodiment being not all of, based on the embodiments of the present invention, those of ordinary skill in the art are not making creative work
Under the premise of every other embodiment obtained, shall fall within the protection scope of the present invention.
Claims (2)
1. a provenance taste share kitchen recognition of face and long-distance intelligent remote control method, which is characterized in that the method includes with
Lower step:
Step 1: the respbian jessie system of burning raspberry pie opens the service of raspberry pie SSH, VNC, purchase into SD card
The usb camera that raspberry pie can be connected opens raspberry pie camera service；Relay is controlled using raspberry pie GPIO low and high level
Switch, has the function that controlling equipment；
Step 2: installation translation and compiling environment (python, c++)；
Step 3: carrying out recognition of face, to guarantee that camera photographed complete face, show what camera photographed using led display screen
Picture, is convenient for knowing others and carries out face's alignment；
Step 4: configuration lnmp environment, server end place the end Web that html+css+js+php writes, use browser on mobile phone
It opens webpage and carries out relevant device switch operation；
Step 5: carrying out the perfect of Web end control system function；
Step 6: setting script booting self-starting.
2. provenance taste as described in claim 1 shares the recognition of face in kitchen and long-distance intelligent remote control method, feature exist
In being described as follows to recognition of face process in the step 3:
Step 3.1: the position of face in picture is determined using HOG --- 16x16 pixel is divided the image into after picture gray processing
Square, how many gradient of principal direction is calculated in each small cube, is replaced using the strongest gradient of directive property original
Small cube, original image is finally converted into a very simple HOG expression-form, captures facial basic structure；
Step 3.2: in order to eliminate face towards error caused by different directions, we use the algorithm of facial facial feature estimation,
Find 68 points (landmark) such as table 1 generally existing on face:
Table 1
The position at each position of face is known according to this 68 points, picture pivots about operation with nose, so that face
Position is substantially aligned, then carries out next step operation；
Step 3.3: giving face encodings, for distinguishing different faces, it is face that concrete operations, which are one depth convolutional network of training,
Portion generate 128 measured values, when the photo of two same persons as input when network can obtain almost the same value, by
Openface program is configured in raspberry pie, training dataset is video under typing state using frame as unit screenshot, carries out face
The picture of Video segmentation is then inputted training pattern by identification, and output test value is similar with the picture recognition of script, video identification its
Real is exactly a large amount of picture recognition, and the frame screenshot of video is stored under same file folder by the present invention, will figure after end of identification
Piece is deleted, and also leaves out screenshot after same recognition of face detection, prevent picture excessively and cause raspberry pie that can not store to
It can not identify, while 128 measured values of different people's training results being stored in cloud database；
Step 3.4: distinguishing different people and pass through 128 measurements of personage in 128 measured values of measurement of comparison result and database
The Euclidean distance of value is compared, it is contemplated that volume will not be arranged in equipment rights management personnel, so the property of this process
Energy raspberry pie can be born completely, one threshold value of default, and the non-same person, authentication authorization and accounting failure, Europe are then thought more than threshold value
The following d (x, y) of formula distance definition
Wherein, xiRepresent the ith measurement value that instrument detects face, yiRepresent i-th of comparison people's image data in database
Measured value.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910103288.8A CN109871797A (en) | 2019-02-01 | 2019-02-01 | A face recognition and remote intelligent remote control method for Yuanwei shared kitchen |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910103288.8A CN109871797A (en) | 2019-02-01 | 2019-02-01 | A face recognition and remote intelligent remote control method for Yuanwei shared kitchen |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109871797A true CN109871797A (en) | 2019-06-11 |
Family
ID=66918520
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910103288.8A Pending CN109871797A (en) | 2019-02-01 | 2019-02-01 | A face recognition and remote intelligent remote control method for Yuanwei shared kitchen |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109871797A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114842584A (en) * | 2022-04-19 | 2022-08-02 | 常州工学院 | Smart access control service method based on RASPBERRY Pi 4B |
| CN117993855A (en) * | 2024-02-02 | 2024-05-07 | 北京中友科技有限公司 | Power industry park personnel management software system |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN204090035U (en) * | 2014-09-23 | 2015-01-07 | 西安盈姆电子信息科技有限公司 | A kind of cloud perimeter system of sending framework based on raspberry |
| CN204090034U (en) * | 2014-09-23 | 2015-01-07 | 西安盈姆电子信息科技有限公司 | A kind of cloud door bell and button system of sending framework based on raspberry |
| CN205614688U (en) * | 2016-04-28 | 2016-10-05 | 漆伟 | Portable service robot of family based on raspberry group and arduino |
| CN107491770A (en) * | 2017-09-12 | 2017-12-19 | 深圳市炫彩酷游科技有限公司 | A kind of monitoring method and system based on recognition of face |
| CN207198851U (en) * | 2017-05-09 | 2018-04-06 | 南昌航空大学 | Intelligent identity identification system based on Raspberry Pi |
| CN108090433A (en) * | 2017-12-12 | 2018-05-29 | 厦门集微科技有限公司 | Face identification method and device, storage medium, processor |
| CN108304800A (en) * | 2018-01-30 | 2018-07-20 | 厦门启尚科技有限公司 | A kind of method of Face datection and face alignment |
| US20180308100A1 (en) * | 2017-04-19 | 2018-10-25 | Risto Haukioja | System and method of client recognition for service provider transactions |
| 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
- 2019-02-01 CN CN201910103288.8A patent/CN109871797A/en active Pending
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN204090035U (en) * | 2014-09-23 | 2015-01-07 | 西安盈姆电子信息科技有限公司 | A kind of cloud perimeter system of sending framework based on raspberry |
| CN204090034U (en) * | 2014-09-23 | 2015-01-07 | 西安盈姆电子信息科技有限公司 | A kind of cloud door bell and button system of sending framework based on raspberry |
| CN205614688U (en) * | 2016-04-28 | 2016-10-05 | 漆伟 | Portable service robot of family based on raspberry group and arduino |
| US20180308100A1 (en) * | 2017-04-19 | 2018-10-25 | Risto Haukioja | System and method of client recognition for service provider transactions |
| CN207198851U (en) * | 2017-05-09 | 2018-04-06 | 南昌航空大学 | Intelligent identity identification system based on Raspberry Pi |
| CN107491770A (en) * | 2017-09-12 | 2017-12-19 | 深圳市炫彩酷游科技有限公司 | A kind of monitoring method and system based on recognition of face |
| CN108090433A (en) * | 2017-12-12 | 2018-05-29 | 厦门集微科技有限公司 | Face identification method and device, storage medium, processor |
| CN108304800A (en) * | 2018-01-30 | 2018-07-20 | 厦门启尚科技有限公司 | A kind of method of Face datection and face alignment |
| 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 |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114842584A (en) * | 2022-04-19 | 2022-08-02 | 常州工学院 | Smart access control service method based on RASPBERRY Pi 4B |
| CN117993855A (en) * | 2024-02-02 | 2024-05-07 | 北京中友科技有限公司 | Power industry park personnel management software system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN205314704U (en) | Intelligence lock and system | |
| CN103375880B (en) | Remote control device and method for air conditioner | |
| CN104394044B (en) | The method and apparatus of self-defined smart machine scene mode | |
| CN105095882B (en) | Gesture recognition method and device | |
| CN202521793U (en) | Remote control device for air conditioner | |
| KR102242648B1 (en) | Artificial intelligence cooking system | |
| CN110098984A (en) | A kind of household appliance networking implementation method | |
| CN107728482A (en) | Control system, control process method and device | |
| WO2022233123A1 (en) | Method and apparatus for controlling air conditioner, and air conditioner | |
| CN104092775A (en) | Intelligent household electrical appliance self-learning method and system | |
| CN103345204B (en) | A kind of home control system | |
| CN107909605A (en) | Control method, device, storage medium and the server of cooking equipment | |
| CN109871797A (en) | A face recognition and remote intelligent remote control method for Yuanwei shared kitchen | |
| CN113486436A (en) | Soft installation design method, device, equipment and storage medium based on tree structure | |
| CN106054620A (en) | Intelligent control device and system | |
| CN110448250B (en) | Tableware storing and taking method and system | |
| CN107131606A (en) | Proximity induction line controller, control method thereof and air conditioner | |
| CN106789456A (en) | A kind of home equipment control method and device | |
| CN107682236A (en) | Smart home interactive system and method based on computer picture recognition | |
| US20250249581A1 (en) | Method, apparatus, device, and computer medium for task execution | |
| CN104601629A (en) | Processing method and processing apparatus, control apparatus and working method thereof as well as control method and control system | |
| CN105225303A (en) | Based on the management system of a kind of network control gate inhibition of face identification system exploitation | |
| CN205176788U (en) | Interactive installation of space atmospheric pressure response | |
| CN111007806B (en) | Smart home control method and device | |
| WO2016101272A1 (en) | Method and apparatus for adjusting network mode |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190611 |