CN109064696A - The optical fiber perimeter security system realized based on deep learning - Google Patents
The optical fiber perimeter security system realized based on deep learning Download PDFInfo
- Publication number
- CN109064696A CN109064696A CN201810940120.8A CN201810940120A CN109064696A CN 109064696 A CN109064696 A CN 109064696A CN 201810940120 A CN201810940120 A CN 201810940120A CN 109064696 A CN109064696 A CN 109064696A
- Authority
- CN
- China
- Prior art keywords
- optical fiber
- digital signal
- unit
- identification
- deep learning
- 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
- G08B13/00—Burglar, theft or intruder alarms
- G08B13/02—Mechanical actuation
- G08B13/12—Mechanical actuation by the breaking or disturbance of stretched cords or wires
- G08B13/122—Mechanical actuation by the breaking or disturbance of stretched cords or wires for a perimeter fence
- G08B13/124—Mechanical actuation by the breaking or disturbance of stretched cords or wires for a perimeter fence with the breaking or disturbance being optically detected, e.g. optical fibers in the perimeter fence
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
Abstract
The invention discloses a kind of optical fiber perimeter security systems realized based on deep learning, including light pulse output unit, fiber optic sensing devices, D/A conversion unit, digital signal processing unit, storage unit, identification categorization module and application end processing module；Fiber optic sensing devices are connected with light pulse output unit and D/A conversion unit respectively, the output end of D/A conversion unit is connect with digital signal processing unit, digital signal processing unit is connected with identification categorization module and storage unit respectively, and identification categorization module connects application end processing module.Optical fiber perimeter security system of the invention effectively increases the recognition capability of optical fiber security protection perimeter system, optical fiber perimeter system is enhanced to adverse circumstances anti-interference ability, system can provide guarantee to self iteration optimization of identification model with the accumulation of time for the status tracking in later period.
Description
Technical field
The invention belongs to intelligent identification technology field, in particular to a kind of optical fiber perimeter security protection realized based on deep learning
System.
Background technique
With the continuous development of China's economy and science and technology, the high-technology in security protection safeguard work changes biography
The security protection system of system.Modern security system requires to must be equipped with boundary defence means, can be real to threatening the event of safety to carry out
When monitor and be accurately positioned, so as to put into strength terminate crime, timely control threat event generation.Application environment not only relates to
And live to school periphery, residential quarters, industrial park etc., the complex environments such as oil field oil depot, power plant, airport, especially
It is the important area that place of military importance, judicial prison, government bodies etc. are related to national safety.Therefore, one safely and effectively
Circumference security protection monitoring system becomes particularly important.
Traditional perimeter security system mainly has infrared monitoring formula perimeter security system, electronic tensile formula circumference security protection system
System, video monitoring formula perimeter security system, vibration wireline formula perimeter security system, that there are performances is poor for traditional security and guard technology, wrong report
Rate is high, be subject to be struck by lightning, is not easy the problems such as installation maintenance, poor anti jamming capability, monitoring are apart from limited, higher cost, optical fiber peace
Anti- monitoring system uses optical cable as carrier as security protection monitoring system of new generation, by directly contacting optical cable or pass through carrying
(such as earthing, wire netting, fence-pass to the various disturbances of optical cable to object, are continued and are monitored in real time, and by electronically
Figure shows the place of interference source, realizes system Realtime Alerts with this.System can be efficiently against existing perimeter security system
The shortcomings that, but also have monitoring distance, electromagnetic-radiation-free, strong antijamming capability, high reliablity, engineering construction relatively easy
The features such as, it is current more advanced technology, and the main flow direction of current security protection market development, there is wide prospect and city
?.
But the feature recognition algorithms of optical fiber perimeter security system, the artificially consciousness that is all based on are sentenced otherwise in the market,
Sensing data is parsed, subjective consciousness is relied on to carry out the differentiation of vibration-sensing signal under specific behavior, feature identification is being carried out and is dividing
When class process, the feature identification library (being not easy to update optimization and upgrading, imperfect not comprehensive) artificially constructed early period is relied on.With strong
Strong subjective consciousness, and the limitation artificially recognized, lead to current existing optical fiber perimeter security system there are rate of false alarms it is higher,
Feature recognition capability is poor, recognition rate is relatively slow, the limitation of feature database, monistic problem.
Summary of the invention
It is an object of the invention to overcome the deficiencies of the prior art and provide one kind can effectively improve optical fiber security protection circumference system
The recognition capability of system, the optical fiber perimeter based on deep learning realized of the enhancing optical fiber perimeter system to adverse circumstances anti-interference ability
Security system.
The purpose of the present invention is achieved through the following technical solutions: the optical fiber perimeter security protection realized based on deep learning
System, including light pulse output unit, fiber optic sensing devices, D/A conversion unit, digital signal processing unit, storage unit,
Identify categorization module and application end processing module；
Light pulse output unit, for passing through hardware optoelectronic switch, power amplifier, by the transmission of laser pulse period
To fiber optic sensing devices；
D/A conversion unit, the light intensity signal for reflecting fiber optic sensing devices are converted to digital signal；
Digital signal processing unit is filtered digital signal by bandpass filter；
It identifies categorization module, establishes neural network for the method by deep learning, and have using labeled
Effect sensing data obtains identification sorter network by neural network repetition training, is carried out using identification sorter network to digital signal
Analysis processing identifies illegal invasion state and matches phagocytic process；
Storage unit, light intensity signal, modulus for output signal, sensing equipment reflection to light pulse output unit
Digital signal, identification classification results after conversion are stored；
Application end processing module, for handling illegal invasion behavior, sound an alarm, mark invasion point coordinate and
Record the state of illegal invasion personnel.
Further, the neural network includes input layer, output layer and five hidden layers, there was only adjacent layer in neural network
There is connection between node, it is mutually connectionless between same layer and cross-layer node.
The beneficial effects of the present invention are: optical fiber perimeter security system of the invention effectively increases optical fiber security protection perimeter system
Recognition capability, enhance optical fiber perimeter system to adverse circumstances anti-interference ability, system, can be to knowledge with the accumulation of time
Other self iteration optimization of model can provide guarantee for the status tracking in later period.
Detailed description of the invention
Fig. 1 is the structure chart of optical fiber perimeter security system of the invention.
Specific embodiment
Technical solution of the present invention is further illustrated with reference to the accompanying drawing.
The advantages of main realization principle of the present invention is using optical fiber sensing technology high sensitivity, loss low, electromagnetism interference,
Based on fiber interference principle, realizes and monitor in real time on a large scale over long distances.The light in boundary is enclosed in pressure, vibration by the physics that is laid in
Fine (cable) front end sensing equipment, generates detectable signal, by the data processing and intelligent recognition of rear end, carries out to different movements
Classification such as climbs enclosure wall, excavates isolation strip, sets up a separatist regime by force of arms protective net, walking in entry region, judges whether it is intrusion behavior, in fact
The timely early warning of existing fiber-optic probe perimeter security system or Realtime Alerts.Fiber-optic probe perimeter security system high sensitivity, anti-electricity
Magnetic disturbance can be used for combustible and explosive area, is not afraid of lightning stroke lighting, field is facilitated to be laid with, convenient for installation and maintenance, meet people to peace
The requirement that all risk insurance is defended.
Optical fiber main flow, which is system, is periodically output to laser sensor fibre front end by light pulse output unit,
According to BOTDR (Brillouin light Time Domain Reflectometry) principle, light has faint optical back scattering and returns during optical fiber transmits
Come, the transducing signal that fiber reflection is returned is converted by accessible number by photoelectric conversion and digital-to-analog circuit transfer principle and is believed
Number, on the one hand used predefined identification/sorter network model treatment utilizes application end pair after the matching of feature recognition mode
On the other hand event handling is stored by storage unit, after particular value must be demarcated, in the way of deep learning, to net
The iterative training of network, and then realize the building of feature identification/disaggregated model, to reach the problem of optimizing to model continuous updating.
As shown in Figure 1, based on the optical fiber perimeter security system that deep learning is realized, including light pulse output unit, optical fiber
Sensing equipment, D/A conversion unit, digital signal processing unit, storage unit, identification categorization module and application end processing module；
Fiber optic sensing devices are connected with light pulse output unit and D/A conversion unit respectively, the output end and number of D/A conversion unit
Signal processing unit connection, digital signal processing unit are connected with identification categorization module and storage unit respectively, identification classification mould
Block connects application end processing module；
Light pulse output unit, for passing through hardware optoelectronic switch, power amplifier, by the transmission of laser pulse period
To fiber optic sensing devices；
D/A conversion unit, the light intensity signal for reflecting fiber optic sensing devices are converted to digital signal；
Digital signal processing unit is filtered digital signal by bandpass filter；
It identifies categorization module, establishes neural network for the method by deep learning, and have using labeled
Effect sensing data obtains identification sorter network by neural network repetition training, and identification classification network is in practical application model
On the basis of iterative update, with the accumulation of sensing data, and become more reliable and stablize, so improve feature identification and
Pattern match efficiency has given full play to deep learning bring identification advantage, has avoided the difference of artificial subjective consciousness and cause
The unstable problem of discrimination；Then digital signal is analyzed and processed using identification sorter network, identifies illegal invasion shape
State simultaneously matches phagocytic process；
Storage unit, light intensity signal, modulus for output signal, sensing equipment reflection to light pulse output unit
Digital signal, identification classification results after conversion are stored；
Application end processing module, for handling illegal invasion behavior, sound an alarm, mark invasion point coordinate and
Record the state of illegal invasion personnel.
Further, the neural network includes input layer, output layer and five hidden layers, there was only adjacent layer in neural network
There is connection between node, it is mutually connectionless between same layer and cross-layer node.Training process includes preceding to trained and backward instruction
Practice.Forward direction training process is unsupervised learning from bottom to top, i.e., since bottom, gradually trains to top layer, in training process,
After training study obtains the (n-1)th layer parameter, by n-1 layers of the input exported as n-th layer, thus training n-th layer is respectively obtained
The parameter of each layer；Backward training process is top-down supervised learning, i.e. the top-down transmission of training error carries out parameter
Fine tuning.Its method trained is the customary means of this field, and details are not described herein again.
Those of ordinary skill in the art will understand that the embodiments described herein, which is to help reader, understands this hair
Bright principle, it should be understood that protection scope of the present invention is not limited to such specific embodiments and embodiments.This field
Those of ordinary skill disclosed the technical disclosures can make according to the present invention and various not depart from the other each of essence of the invention
The specific variations and combinations of kind, these variations and combinations are still within the scope of the present invention.
Claims (2)
1. the optical fiber perimeter security system realized based on deep learning, which is characterized in that passed including light pulse output unit, optical fiber
Feel equipment, D/A conversion unit, digital signal processing unit, storage unit, identification categorization module and application end processing module；
Light pulse output unit, for by hardware optoelectronic switch, power amplifier, laser pulse period to be sent to light
Fine sensing equipment；
D/A conversion unit, the light intensity signal for reflecting fiber optic sensing devices are converted to digital signal；
Digital signal processing unit is filtered digital signal by bandpass filter；
It identifies categorization module, establishes neural network for the method by deep learning, and utilize labeled effective biography
Sense data obtain identification sorter network by neural network repetition training, are analyzed using identification sorter network digital signal
Processing identifies illegal invasion state and matches phagocytic process；
Storage unit, for the output signal to light pulse output unit, the light intensity signal of sensing equipment reflection, analog-to-digital conversion
Digital signal afterwards, identification classification results are stored；
Application end processing module sounds an alarm, marks the coordinate and record of invasion point for handling illegal invasion behavior
The state of illegal invasion personnel.
2. the optical fiber perimeter security system according to claim 1 realized based on deep learning, which is characterized in that the mind
Include input layer, output layer and five hidden layers through network, only there is connection in neural network between adjacent node layer, same layer with
And it is mutually connectionless between cross-layer node.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810940120.8A CN109064696A (en) | 2018-08-17 | 2018-08-17 | The optical fiber perimeter security system realized based on deep learning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810940120.8A CN109064696A (en) | 2018-08-17 | 2018-08-17 | The optical fiber perimeter security system realized based on deep learning |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109064696A true CN109064696A (en) | 2018-12-21 |
Family
ID=64687214
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810940120.8A Pending CN109064696A (en) | 2018-08-17 | 2018-08-17 | The optical fiber perimeter security system realized based on deep learning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109064696A (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110133382A (en) * | 2019-04-10 | 2019-08-16 | 上海交通大学 | Wide-band microwave Instantaneous Frequency Measurement optimization method and device neural network based |
| CN110492926A (en) * | 2019-07-26 | 2019-11-22 | 北京光锁科技有限公司 | A kind of optical fiber eavesdropping localization method, system and electronic equipment based on machine learning |
| CN110794462A (en) * | 2019-11-06 | 2020-02-14 | 广东博智林机器人有限公司 | Building site safety monitoring system and monitoring method and device thereof |
| CN110995339A (en) * | 2019-11-26 | 2020-04-10 | 电子科技大学 | Method for extracting and identifying time-space information of distributed optical fiber sensing signal |
| CN111458059A (en) * | 2019-01-21 | 2020-07-28 | 中国科学院上海光学精密机械研究所 | Continuous vehicle collision detection system and method thereof |
| CN111554079A (en) * | 2020-04-26 | 2020-08-18 | 成都铭钛科技有限公司 | Trigger judgment method and device based on multi-channel data |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20140050473A1 (en) * | 2012-08-17 | 2014-02-20 | Cleveland Electric Laboratories Co. | Sagnac interferometer event sensing and locating device |
| CN104751580A (en) * | 2015-04-08 | 2015-07-01 | 武汉理工光科股份有限公司 | Distributed optical fiber sensing signal mode identifying method and system |
| CN105869327A (en) * | 2016-04-20 | 2016-08-17 | 宁波绿阳智汇电气科技有限公司 | Optical fiber vibration detection system and invasion detection method thereof |
| CN106683305A (en) * | 2017-01-22 | 2017-05-17 | 唐山市镜圆科技有限公司 | Anti-eavesdrop optical-fiber alarm system |
| CN107067605A (en) * | 2016-12-31 | 2017-08-18 | 上海华魏光纤传感技术有限公司 | A kind of distributed optical fiber fence vibrating intruding recognition methods |
| CN107180521A (en) * | 2017-04-19 | 2017-09-19 | 天津大学 | Optical fiber perimeter security protection intrusion event recognition methods and device based on comprehensive characteristics |
| DE102016114722A1 (en) * | 2016-08-09 | 2018-02-15 | Fibre Optics Ct Gmbh | Intruder alarm notification device |
| CN108399696A (en) * | 2018-03-22 | 2018-08-14 | 中科润程(北京)物联科技有限责任公司 | Intrusion behavior recognition methods and device |
-
2018
- 2018-08-17 CN CN201810940120.8A patent/CN109064696A/en active Pending
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20140050473A1 (en) * | 2012-08-17 | 2014-02-20 | Cleveland Electric Laboratories Co. | Sagnac interferometer event sensing and locating device |
| CN104751580A (en) * | 2015-04-08 | 2015-07-01 | 武汉理工光科股份有限公司 | Distributed optical fiber sensing signal mode identifying method and system |
| CN105869327A (en) * | 2016-04-20 | 2016-08-17 | 宁波绿阳智汇电气科技有限公司 | Optical fiber vibration detection system and invasion detection method thereof |
| DE102016114722A1 (en) * | 2016-08-09 | 2018-02-15 | Fibre Optics Ct Gmbh | Intruder alarm notification device |
| CN107067605A (en) * | 2016-12-31 | 2017-08-18 | 上海华魏光纤传感技术有限公司 | A kind of distributed optical fiber fence vibrating intruding recognition methods |
| CN106683305A (en) * | 2017-01-22 | 2017-05-17 | 唐山市镜圆科技有限公司 | Anti-eavesdrop optical-fiber alarm system |
| CN107180521A (en) * | 2017-04-19 | 2017-09-19 | 天津大学 | Optical fiber perimeter security protection intrusion event recognition methods and device based on comprehensive characteristics |
| CN108399696A (en) * | 2018-03-22 | 2018-08-14 | 中科润程(北京)物联科技有限责任公司 | Intrusion behavior recognition methods and device |
Non-Patent Citations (1)
| Title |
|---|
| 董海鹰等: "《智能控制理论及应用》", 30 September 2016, 中国铁道出版社 * |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111458059A (en) * | 2019-01-21 | 2020-07-28 | 中国科学院上海光学精密机械研究所 | Continuous vehicle collision detection system and method thereof |
| CN111458059B (en) * | 2019-01-21 | 2022-01-28 | 中国科学院上海光学精密机械研究所 | Continuous vehicle collision detection system and method thereof |
| CN110133382A (en) * | 2019-04-10 | 2019-08-16 | 上海交通大学 | Wide-band microwave Instantaneous Frequency Measurement optimization method and device neural network based |
| CN110133382B (en) * | 2019-04-10 | 2021-07-27 | 上海交通大学 | Broadband microwave instantaneous frequency measurement optimization method and device based on neural network |
| CN110492926A (en) * | 2019-07-26 | 2019-11-22 | 北京光锁科技有限公司 | A kind of optical fiber eavesdropping localization method, system and electronic equipment based on machine learning |
| CN110794462A (en) * | 2019-11-06 | 2020-02-14 | 广东博智林机器人有限公司 | Building site safety monitoring system and monitoring method and device thereof |
| CN110794462B (en) * | 2019-11-06 | 2021-08-03 | 广东博智林机器人有限公司 | Building site safety monitoring system and monitoring method and device thereof |
| CN110995339A (en) * | 2019-11-26 | 2020-04-10 | 电子科技大学 | Method for extracting and identifying time-space information of distributed optical fiber sensing signal |
| CN110995339B (en) * | 2019-11-26 | 2022-08-19 | 电子科技大学 | Method for extracting and identifying time-space information of distributed optical fiber sensing signal |
| CN111554079A (en) * | 2020-04-26 | 2020-08-18 | 成都铭钛科技有限公司 | Trigger judgment method and device based on multi-channel data |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109064696A (en) | The optical fiber perimeter security system realized based on deep learning | |
| Li et al. | Pattern recognition for distributed optical fiber vibration sensing: A review | |
| US20180357542A1 (en) | 1D-CNN-Based Distributed Optical Fiber Sensing Signal Feature Learning and Classification Method | |
| CN110995339B (en) | Method for extracting and identifying time-space information of distributed optical fiber sensing signal | |
| US20180195883A1 (en) | Optical fibre sensor system | |
| CN104574742B (en) | A kind of optical fiber perimeter safety-protection system based on Φ OTDR techniques | |
| CN103824406A (en) | Optical fiber sensing-based tunnel cable invasive detection method and system | |
| CN103236127A (en) | Fiber fence intrusion monitoring system and pattern recognition method thereof | |
| CN101556724A (en) | Safety management system of optical fiber perimeter and pattern recognition method thereof | |
| CN205428085U (en) | Indoor security protection system based on fiber grating | |
| CN104008621A (en) | Defense area type optical fiber disturbance periphery security and protection system and quick invasion disturbance judgment method | |
| WO2016198683A1 (en) | Method and apparatus for monitoring pipeline using an optical fiber sensor system | |
| Wu et al. | Intelligent detection and identification in fiber-optical perimeter intrusion monitoring system based on the FBG sensor network | |
| CN112907869B (en) | Intrusion detection system based on multiple sensing technologies | |
| CN206610396U (en) | A kind of fiber grating fence intrusion alarm detecting system | |
| CN202205305U (en) | Defence area type security intrusion system based on vibration and video monitoring | |
| Yang et al. | Early safety warnings for long-distance pipelines: A distributed optical fiber sensor machine learning approach | |
| CN103794010A (en) | Perimeter safety protection system | |
| CN203799460U (en) | Tunnel cable intrusion detection system based on optical fiber sensing | |
| CN109374116A (en) | The excavation Activity recognition method of buried Fibre Optical Sensor vibration-detection system | |
| Tabi Fouda et al. | Research and software design of an Φ-OTDR-based optical fiber vibration recognition algorithm | |
| CN113936402A (en) | Perimeter security system and security scheme based on distributed optical fiber early warning system and video target identification technology | |
| CN106781152A (en) | A kind of fiber grating fence intrusion alarm detecting system and method | |
| CN114093105A (en) | Active and passive combined security and protection enclosure alarm system | |
| Mahmoud | Practical Aspects of Perimeter Intrusion Detection and Nuisance Suppression for Distributed Fibre-optic Sensors |
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: 20181221 |