CN109544859A - A kind of active safety household camera intelligent realization method - Google Patents
A kind of active safety household camera intelligent realization method Download PDFInfo
- Publication number
- CN109544859A CN109544859A CN201811515999.8A CN201811515999A CN109544859A CN 109544859 A CN109544859 A CN 109544859A CN 201811515999 A CN201811515999 A CN 201811515999A CN 109544859 A CN109544859 A CN 109544859A
- Authority
- CN
- China
- Prior art keywords
- image
- processing unit
- camera
- image processing
- realization method
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G08—SIGNALLING
- G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
- G08B19/00—Alarms responsive to two or more different undesired or abnormal conditions, e.g. burglary and fire, abnormal temperature and abnormal rate of flow
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/40—Scenes; Scene-specific elements in video content
- G06V20/46—Extracting features or characteristics from the video content, e.g. video fingerprints, representative shots or key frames
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/50—Context or environment of the image
- G06V20/52—Surveillance or monitoring of activities, e.g. for recognising suspicious objects
-
- 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
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N23/00—Cameras or camera modules comprising electronic image sensors; Control thereof
- H04N23/60—Control of cameras or camera modules
- H04N23/66—Remote control of cameras or camera parts, e.g. by remote control devices
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N23/00—Cameras or camera modules comprising electronic image sensors; Control thereof
- H04N23/80—Camera processing pipelines; Components thereof
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N7/00—Television systems
- H04N7/18—Closed-circuit television [CCTV] systems, i.e. systems in which the video signal is not broadcast
Landscapes
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Signal Processing (AREA)
- Theoretical Computer Science (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Oral & Maxillofacial Surgery (AREA)
- Human Computer Interaction (AREA)
- Studio Devices (AREA)
- Alarm Systems (AREA)
Abstract
A kind of active safety household camera intelligent realization method is based on following device: camera and control unit, image receiving unit, image processing unit, security logic, main control unit, motor circuit unit, machine assembly;Camera control unit control intake image is sent into image processing unit after image receiving unit receives image, and image processing unit uses deep learning algorithm, being capable of each type objects and human limbs movement in automatic identification image;Image processing unit exports the monitoring APP that user mobile phone is connectable to by network;User setting installs matched APP, and security logic is equipped in APP, emergency contact and dangerous scene etc. is arranged, and can check home background with real time remote;The judgment threshold of the dangerous scene of security logic setting makes dangerous judgement more than threshold value standard if meeting, alarms;If not meeting the standard of threshold value, dangerous judgement is not made.
Description
Technical field
The present invention relates to safe household camera field, especially a kind of active household camera and its safe and intelligent are real
Existing method.
Background technique
With the development of the times, smart home is more and more accepted, one as smart home of camera
Important component, also by users using installation.
Currently, the problem of an aging population is increasingly severe, the opening of two tire policies will be so that children's number will also increase, but blueness is strong
It is year powerless to look after the old and the children simultaneously, thus occur often old solitary people fall down unmanned discovery, be in injury etc. of playing of children it is new
It hears.Although there is part family to be mounted with household camera, traditional camera can only provide real-time video and functions, the user such as check
It can not monitor constantly.It when occurring unintended emergency situation in manager, can not often find in time, to delay rescue.
To reach the real-time monitoring to kinsfolk, when preventing fortuitous event from occurring or shortening the rescue after surprisingly occurring
Between, it is proposed that a kind of active safety household camera and its implementation realize the security monitoring of real-time active to family.
Summary of the invention
Object of the present invention is to against the background of the prior art, propose a kind of active safety household camera, and its realization side
Method, realize to dangerous scene judge automatically and initiative alarming.The method of household camera intelligent decision safety.
The technical scheme is that a kind of active safety household camera intelligent realization method, comprising:
E. it is based on following device: camera and control unit, image receiving unit, image processing unit, security logic list
Member, main control unit, motor circuit unit, machine assembly;Main control unit, horizontal stage electric machine circuit unit, the machine assembly point for including
Not Kong Zhi video information transmission, the rotation of camera holder, camera power supply;
F. camera control unit control intake image is sent into image processing unit after image receiving unit receives image,
Image processing unit uses deep learning algorithm, being capable of each type objects and human limbs movement in automatic identification image;Camera
Monitoring needs safe family to picture;Image processing unit exports the monitoring APP that user mobile phone is connectable to by network;
G. user setting installs matched APP, and security logic is equipped in APP, and emergency contact and danger is arranged
Scene etc., and home background can be checked with real time remote;The judgment threshold of the dangerous scene of security logic setting, if meeting
More than threshold value standard, dangerous judgement is made, is alarmed;If not meeting the standard of threshold value, dangerous judgement is not made;
H. after being provided with, APP control main control unit and horizontal stage electric machine, security logic judge that image processing unit is defeated
Whether video meets dangerous scene definition out, alarms immediately if meeting.
Further, the image receiving unit that device includes in step a is used to receive the image information of camera acquisition, and
Control the resolution ratio and frame number of video.
Further, the image processing unit that device includes in step a has decrease of noise functions, can carry out to video image
Pretreatment.
Further, the security logic for including in step a can judge automatically whether video scene is user's setting
Dangerous scene.
Further, in step b image processing unit using target detection deep learning algorithm such as faster-rcnn or
Ssd etc. marks the domestic articles such as people, sofa by hand, and data format is PascalVoc data set format, and data volume is 100,000
Picture trains respective objects detection model.
Further, image processing unit uses mtcnn et al. face detection algorithm and multilayer convolutional neural networks in step b
Algorithm, data set is the 500000+human face data collected, and marks position of good person's face in picture, and Face datection algorithm is used for
Detect video in face, multilayer convolutional neural networks algorithm for distinguish whether be stranger, the standard judged be detect people
Face generates corresponding face vector, and the vector generated with the face that user saves carries out similarity-rough set, and similarity value calculation is big
In 0.6 and it is highest be corresponding non-stranger, be otherwise stranger.
Further, image processing unit uses deep learning algorithm in step b, constructs a multilayer ResNet structure,
Purpose is classification, inputs the image of the people to detect, exports as the corresponding state of people in image.Trained data set is by hand
20,000+state diagram of screening.
Further, the matched APP of user installation in step c, the APP can obtain the control of corresponding camera by encryption
System, the settable emergency contact of user and dangerous scene, such as fall down, climb the alert fire alarm of window, cigarette, stranger's invasion etc..
Further, the matched APP of user installation in step c, which can remotely check family's real-time video, and control
Cloud platform rotation, watches the home background of different angle, while having the function of that playback is recorded.
Further, in step d in security logic combination step b image processing unit identify object space, object
Body classification, human action etc. determine whether to meet the dangerous scene set in step c, if meeting, set immediately into step c
Emergency contact push warning message.
The utility model has the advantages that by receiving the image information of camera acquisition, and control the resolution ratio of video and frame number can
To realize intelligent monitoring under conditions of flow and information processing capacity economy, and may it is critical under conditions of automatically setting it is fine
The information processing of image;Image processing unit uses deep learning algorithm, detects to object in video, mark object position
It sets and object category, human action identification, judgement is what state the people is currently in, carries out to the people detected in video
Recognition of face determines whether stranger and other possible predictor of risk situations --- it can by security logic
Judge automatically whether video scene is dangerous scene set by user, is alerted in time, intervene user in time.
Detailed description of the invention
Fig. 1 is a kind of active safety household camera of the present invention and its implementation flow diagram.
Fig. 2 is the structural schematic diagram the present invention is based on the picture retrieval method of convolution feature and system.
Specific embodiment
Below in conjunction with specific embodiment shown in the drawings, the present invention will be described in detail.
Fig. 1 is that the active safety of a kind of active safety household camera of the present invention and its implementation specific embodiment pushes away
Send flow chart.In the present embodiment, it mainly comprises the steps that
Step 11: user sets dangerous scene and emergency contact in APP, the data parameters of control camera acquisition and
Camera angle.
User can choose a variety of dangerous scenes of default in APP, such as old man falls down, child climbs window, catches fire in family
Deng if user also needs to define other dangerous scenes not provided, user can pass through shooting picture and carry out customized dangerous field
Scape, when there are can also provide active alarming function when the type scene in camera.
Multiple emergency contacts and type of alarm is arranged in user in APP, such as short message is reminded or made a phone call.
User can be according to the remotely located resolution ratio of camera head of self-demand and control camera shooting side in APP
To.
Step 12: camera acquires video data, and carries out noise reduction process, while being uploaded to cloud by network.
Camera acquires video data using image receiving unit, by video data transmission to image processing unit, image
Processing unit carries out noise reduction process to video, and by Video compression, is uploaded to cloud by network and saves, facilitate user logical
It crosses APP real time inspection home videos information or the later period is downloaded and reviews.
Step 13: local or cloud handle video data, using deep neural network, use to image in video
Object detection orients object space and object category, carries out action recognition and recognition of face to people in video.
Deep neural network module can be divided into processing locality version and cloud processing version in image processing unit, and function is
Object detection, action recognition and recognition of face are carried out to image in video.Object in tag image may be implemented in object detection
Position and object generic, action recognition may be implemented to differentiate personage's posture, the classification of motion, and recognition of face may be implemented to judge
Whether personage is that user sets non-stranger in image.If deep neural network module is beyond the clouds in image processing unit, cloud
Recognition result is returned to camera by network by end.
Step 14: by the comprehensive analysis of the results such as object detection, action recognition, recognition of face, judging whether to belong to step
Danger scene defined in rapid 11 pushes hydropac to emergency contact if belonging to.
Security logic analyzes image processing unit recognition result, scene is constructed according to result reasoning, if pushing away
The scene managed out belongs to the dangerous scene of definition, then security logic is issued by network to emergency contact set by user
Set the hydropac of type.
Fig. 2 is the structural schematic diagram of a kind of active safety household camera of the present invention and its implementation.
Module 21 is image receiving unit, and effect is to acquire video data using camera hardware, being capable of auto-focusing.
Module 22 is image processing unit, and effect is to carry out object to video image to the processing of received video image denoising
Physical examination survey, action recognition and recognition of face.
Module 23 is security logic, and effect is to judge whether image scene is dangerous field set by user in video
Scape pushes alarm to user if belonging to.
Module 24 is main control unit, and effect is data network transmission management, telecommunication management etc. between each unit module.
Module 25 is motor circuit unit and machine assembly, and effect is the power supply of control camera and holder rotation etc..
Module 26 is APP unit, and effect is to provide user setting danger scene, long-range to monitor, remote operation etc..
Claims (10)
1. a kind of active safety household camera intelligent realization method, characterized in that include the following steps:
A. it is based on following device: camera and control unit, image receiving unit, image processing unit, security logic, master
Control unit, horizontal stage electric machine circuit unit, machine assembly;Main control unit, horizontal stage electric machine circuit unit and the machine assembly for including point
Not Kong Zhi video information transmission, the rotation of camera holder;
B. camera control unit control intake image is sent into image processing unit, image after image receiving unit receives image
Processing unit uses deep learning algorithm, being capable of each type objects and human limbs movement in automatic identification image;Camera head monitor
Need safe family to picture;Image processing unit exports the monitoring APP that user mobile phone is connectable to by network;
C. user setting installs matched APP, and security logic is equipped in APP, and emergency contact and dangerous scene is arranged
Deng, and home background can be checked with real time remote;Security logic setting dangerous scene judgment threshold, be more than if meeting
Threshold value standard makes dangerous judgement, alarms;If not meeting the standard of threshold value, dangerous judgement is not made;
D. after being provided with, APP controls main control unit and horizontal stage electric machine, and security logic judges image processing unit output view
Whether frequency meets dangerous scene definition, alarms immediately if meeting.
2. household camera intelligent realization method according to claim 1, characterized in that the figure that device includes in step a
As the image information that receiving unit is acquired for receiving camera, and control the resolution ratio and frame number of video.
3. active safety household camera intelligent realization method according to claim 1, characterized in that device in step a
The image processing unit for including has decrease of noise functions, can enhance video image, increases contrast and clarity, removal
Image noise.
4. active safety household camera intelligent realization method according to claim 1, characterized in that include in step a
Security logic can judge automatically whether video scene is dangerous scene set by user.
5. active safety household camera intelligent realization method according to claim 1, characterized in that image in step b
Processing unit examines object each in video and people using target detection deep learning algorithm such as faster-rcnn or ssd etc.
It surveys, marks object and people position and classification.
6. active safety household camera intelligent realization method according to claim 1, characterized in that image in step b
Processing unit using mtcnn et al. face detection algorithm and multilayer convolutional neural networks algorithm, to the face detected in video into
Row recognition of face generates corresponding face vector, while the vector that the face saved with user generates carries out similarity-rough set, benefit
It is with the measuring similarity formula for calculating similarityWherein A, B are respectively the face spy saved
Sign vector and the face feature vector newly detected;If the similarity that some vector and B are calculated in A is greater than 0.6 and similarity highest,
The corresponding people of the vector in the artificial A then newly detected, is otherwise stranger.
7. active safety household camera intelligent realization method according to claim 1, characterized in that image in step b
Processing unit constructs a sorter network, the people detected in utilization 5 by multilayer ResNet structure using deep learning algorithm
People from position frame is selected, input of the figure as network exports as the status gestures of people, such as prone position, stance, falls down.
8. active safety household camera intelligent realization method according to claim 1, characterized in that user in step c
Matched APP is installed, which can obtain the control of corresponding camera, the settable emergency contact of user and danger by encryption
Dangerous scene, such as fall down, climb the alert fire alarm of window, cigarette, stranger's invasion etc..
9. active safety household camera intelligent realization method according to claim 1, characterized in that user in step c
Matched APP is installed, which can remotely check family's real-time video, and control cloud platform rotation, watch family's feelings of different angle
Condition.
10. active safety household camera intelligent realization method according to claim 1, characterized in that safety in step d
Object space, object category, the human action etc. that image processing unit identifies in logic unit combination step b determine whether to meet
The dangerous scene set in step c, if meeting, the emergency contact set immediately into step c pushes warning message.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811515999.8A CN109544859A (en) | 2018-12-12 | 2018-12-12 | A kind of active safety household camera intelligent realization method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811515999.8A CN109544859A (en) | 2018-12-12 | 2018-12-12 | A kind of active safety household camera intelligent realization method |
Publications (1)
Publication Number | Publication Date |
---|---|
CN109544859A true CN109544859A (en) | 2019-03-29 |
Family
ID=65854405
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201811515999.8A Pending CN109544859A (en) | 2018-12-12 | 2018-12-12 | A kind of active safety household camera intelligent realization method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109544859A (en) |
Cited By (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110458113A (en) * | 2019-08-14 | 2019-11-15 | 旭辉卓越健康信息科技有限公司 | A kind of non-small face identification method cooperated under scene of face |
CN111047828A (en) * | 2019-12-12 | 2020-04-21 | 天地伟业技术有限公司 | Household intelligent security alarm system |
CN111125768A (en) * | 2019-12-26 | 2020-05-08 | 联想(北京)有限公司 | Information processing method, information processing apparatus, electronic device, and medium |
CN111860034A (en) * | 2019-04-24 | 2020-10-30 | 广州煜煊信息科技有限公司 | Household accident handling method |
CN113158731A (en) * | 2020-12-31 | 2021-07-23 | 杭州拓深科技有限公司 | Push-up counting method and device, electronic device and storage medium |
CN113762184A (en) * | 2021-09-13 | 2021-12-07 | 北京市商汤科技开发有限公司 | Image processing method, image processing device, electronic equipment and computer storage medium |
CN113824917A (en) * | 2021-09-18 | 2021-12-21 | 厦门网拓科技有限公司 | Convenient video chat method based on third-party video software and terminal thereof |
CN113867163A (en) * | 2021-10-09 | 2021-12-31 | 深圳康佳电子科技有限公司 | Intelligent household scene switching method and device, intelligent terminal and storage medium |
CN117135324A (en) * | 2023-09-08 | 2023-11-28 | 南京启征信息技术有限公司 | Video integration intelligent identification system |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20090225750A1 (en) * | 2008-03-07 | 2009-09-10 | Embarq Holdings Company, Llc | System and Method for Remote Home Monitoring Utilizing a VoIP Phone |
CN104881966A (en) * | 2015-05-15 | 2015-09-02 | 焦一平 | Wireless remote monitoring system for gas leakage |
CN107045623A (en) * | 2016-12-30 | 2017-08-15 | 厦门瑞为信息技术有限公司 | A kind of method of the indoor dangerous situation alarm based on human body attitude trace analysis |
CN107249037A (en) * | 2017-07-02 | 2017-10-13 | 东华大学 | A kind of mobile communication Monitor Equipment of controllable head |
CN206611528U (en) * | 2016-12-28 | 2017-11-03 | 珠海迈科智能科技股份有限公司 | It is a kind of that there is the Intelligent set top box with IP Camera |
CN207070220U (en) * | 2017-03-01 | 2018-03-02 | 武汉摩菲智能科技有限公司 | Artificial intelligence security protection camera system |
CN207869208U (en) * | 2017-11-24 | 2018-09-14 | 核工业西南物理研究院 | A kind of wireless family monitoring camera |
-
2018
- 2018-12-12 CN CN201811515999.8A patent/CN109544859A/en active Pending
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20090225750A1 (en) * | 2008-03-07 | 2009-09-10 | Embarq Holdings Company, Llc | System and Method for Remote Home Monitoring Utilizing a VoIP Phone |
CN104881966A (en) * | 2015-05-15 | 2015-09-02 | 焦一平 | Wireless remote monitoring system for gas leakage |
CN206611528U (en) * | 2016-12-28 | 2017-11-03 | 珠海迈科智能科技股份有限公司 | It is a kind of that there is the Intelligent set top box with IP Camera |
CN107045623A (en) * | 2016-12-30 | 2017-08-15 | 厦门瑞为信息技术有限公司 | A kind of method of the indoor dangerous situation alarm based on human body attitude trace analysis |
CN207070220U (en) * | 2017-03-01 | 2018-03-02 | 武汉摩菲智能科技有限公司 | Artificial intelligence security protection camera system |
CN107249037A (en) * | 2017-07-02 | 2017-10-13 | 东华大学 | A kind of mobile communication Monitor Equipment of controllable head |
CN207869208U (en) * | 2017-11-24 | 2018-09-14 | 核工业西南物理研究院 | A kind of wireless family monitoring camera |
Cited By (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111860034A (en) * | 2019-04-24 | 2020-10-30 | 广州煜煊信息科技有限公司 | Household accident handling method |
CN110458113A (en) * | 2019-08-14 | 2019-11-15 | 旭辉卓越健康信息科技有限公司 | A kind of non-small face identification method cooperated under scene of face |
CN111047828A (en) * | 2019-12-12 | 2020-04-21 | 天地伟业技术有限公司 | Household intelligent security alarm system |
CN111125768A (en) * | 2019-12-26 | 2020-05-08 | 联想(北京)有限公司 | Information processing method, information processing apparatus, electronic device, and medium |
CN113158731A (en) * | 2020-12-31 | 2021-07-23 | 杭州拓深科技有限公司 | Push-up counting method and device, electronic device and storage medium |
CN113762184A (en) * | 2021-09-13 | 2021-12-07 | 北京市商汤科技开发有限公司 | Image processing method, image processing device, electronic equipment and computer storage medium |
CN113824917A (en) * | 2021-09-18 | 2021-12-21 | 厦门网拓科技有限公司 | Convenient video chat method based on third-party video software and terminal thereof |
CN113867163A (en) * | 2021-10-09 | 2021-12-31 | 深圳康佳电子科技有限公司 | Intelligent household scene switching method and device, intelligent terminal and storage medium |
CN117135324A (en) * | 2023-09-08 | 2023-11-28 | 南京启征信息技术有限公司 | Video integration intelligent identification system |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109544859A (en) | A kind of active safety household camera intelligent realization method | |
US10593174B1 (en) | Automatic setup mode after disconnect from a network | |
CN103839373B (en) | A kind of unexpected abnormality event Intelligent Recognition alarm device and warning system | |
US11252378B1 (en) | Batteryless doorbell with rectified power delivery | |
US11257226B1 (en) | Low-overhead motion classification | |
CN108492404B (en) | Face lock unlocking method and device with additional expression recognition function | |
US11341825B1 (en) | Implementing deterrent protocols in response to detected security events | |
US11217076B1 (en) | Camera tampering detection based on audio and video | |
US11240474B1 (en) | Reporting connectivity problems for electronic devices | |
US10803719B1 (en) | Batteryless doorbell with energy harvesters | |
CN104504793A (en) | Intelligent door safety control system and method based on video service | |
US10713928B1 (en) | Arming security systems based on communications among a network of security systems | |
US12073698B1 (en) | Security device with user-configurable motion detection settings | |
US11164435B1 (en) | Audio/video recording and communication doorbell devices with supercapacitors | |
Toreyin et al. | Wildfire detection using LMS based active learning | |
US10943442B1 (en) | Customized notifications based on device characteristics | |
US11546951B1 (en) | Touchless setup mode initiation for networked devices | |
US10733857B1 (en) | Automatic alteration of the storage duration of a video | |
US11423762B1 (en) | Providing device power-level notifications | |
CN110956768A (en) | Automatic anti-theft device of intelligence house | |
US12014271B2 (en) | Training image classifiers | |
CN115690653A (en) | Monitoring and early warning for realizing abnormal nursing behaviors of nursing staff based on AI behavior recognition | |
CN112418045A (en) | Safety monitoring method and system and intelligent household appliance | |
CN111908288A (en) | TensorFlow-based elevator safety system and method | |
US11163097B1 (en) | Detection and correction of optical filter position in a camera device |
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: 20190329 |