CN109410497B - Bridge opening space safety monitoring and alarming system based on deep learning - Google Patents
Bridge opening space safety monitoring and alarming system based on deep learning Download PDFInfo
- Publication number
- CN109410497B CN109410497B CN201811383003.2A CN201811383003A CN109410497B CN 109410497 B CN109410497 B CN 109410497B CN 201811383003 A CN201811383003 A CN 201811383003A CN 109410497 B CN109410497 B CN 109410497B
- Authority
- CN
- China
- Prior art keywords
- smoke
- bridge opening
- module
- fire
- pedestrian
- 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.)
- Active
Links
- 238000013135 deep learning Methods 0.000 title claims abstract description 39
- 238000012544 monitoring process Methods 0.000 title claims abstract description 34
- 238000001514 detection method Methods 0.000 claims abstract description 70
- 239000000779 smoke Substances 0.000 claims abstract description 66
- 230000009545 invasion Effects 0.000 claims abstract description 34
- 238000007689 inspection Methods 0.000 claims abstract description 26
- 238000005516 engineering process Methods 0.000 claims abstract description 10
- 238000012545 processing Methods 0.000 claims abstract description 5
- 238000007726 management method Methods 0.000 claims description 30
- 238000012549 training Methods 0.000 claims description 20
- 238000013523 data management Methods 0.000 claims description 8
- 238000004458 analytical method Methods 0.000 claims description 6
- 238000013528 artificial neural network Methods 0.000 claims description 6
- 230000006399 behavior Effects 0.000 claims description 6
- 238000003062 neural network model Methods 0.000 claims description 6
- 230000002159 abnormal effect Effects 0.000 claims description 5
- 238000007405 data analysis Methods 0.000 claims description 4
- 238000000034 method Methods 0.000 abstract description 14
- 230000008569 process Effects 0.000 description 5
- 230000006870 function Effects 0.000 description 4
- 238000010586 diagram Methods 0.000 description 3
- 238000003384 imaging method Methods 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- 230000002349 favourable effect Effects 0.000 description 2
- 238000005070 sampling Methods 0.000 description 2
- 238000012935 Averaging Methods 0.000 description 1
- 230000004913 activation Effects 0.000 description 1
- 238000013527 convolutional neural network Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000011835 investigation Methods 0.000 description 1
- 230000004807 localization Effects 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000011176 pooling Methods 0.000 description 1
- 230000003449 preventive effect Effects 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 238000005728 strengthening Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G08—SIGNALLING
- G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
- G08B13/00—Burglar, theft or intruder alarms
- G08B13/18—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength
- G08B13/189—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength using passive radiation detection systems
- G08B13/194—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength using passive radiation detection systems using image scanning and comparing systems
- G08B13/196—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength using passive radiation detection systems using image scanning and comparing systems using television cameras
- G08B13/19602—Image analysis to detect motion of the intruder, e.g. by frame subtraction
- G08B13/19613—Recognition of a predetermined image pattern or behaviour pattern indicating theft or intrusion
- G08B13/19615—Recognition of a predetermined image pattern or behaviour pattern indicating theft or intrusion wherein said pattern is defined by the user
-
- 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
-
- 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
-
- 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
- G06V40/172—Classification, e.g. identification
-
- 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
- G06V40/20—Movements or behaviour, e.g. gesture recognition
- G06V40/23—Recognition of whole body movements, e.g. for sport training
- G06V40/25—Recognition of walking or running movements, e.g. gait recognition
-
- G—PHYSICS
- G08—SIGNALLING
- G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
- G08B17/00—Fire alarms; Alarms responsive to explosion
- G08B17/12—Actuation by presence of radiation or particles, e.g. of infrared radiation or of ions
- G08B17/125—Actuation by presence of radiation or particles, e.g. of infrared radiation or of ions by using a video camera to detect fire or smoke
Abstract
The invention provides a bridge opening space safety monitoring and alarming system based on deep learning, which relates to the technical field of intelligent security and protection and comprises a video acquisition module, a pedestrian intrusion detection module, a smoke and fire detection module, an information management module and an alarming module; the video acquisition module is respectively connected with the input ends of the pedestrian invasion detection module and the smoke and fire detection module, the output ends of the pedestrian invasion detection module and the smoke and fire detection module are respectively connected with the input end of the information management module, and the output end of the information management module is connected with the alarm module. The method is used for carrying out intrusion detection on the specified area of the bridge opening space based on deep learning combined with face recognition and matching, carrying out smoke and fire recognition in the monitoring coverage range under the bridge based on the deep learning combined with an image processing technology, acquiring video of the bridge opening space through a monitoring camera, detecting pedestrian intrusion and smoke and fire, sending alarm information to inspection personnel in time, and adopting related measures to ensure the safety of the bridge opening.
Description
Technical Field
The invention belongs to the technical field of intelligent security and particularly relates to a bridge opening space safety monitoring and alarming system based on deep learning.
Background
Along with the rapid development of economy, the viaduct bridge of the highway trunk has a lot of high level, and the transportation is convenient, and meanwhile, some potential safety hazards exist, for example, wave personnel occupy the vacant bridge opening space as a temporary residence, and nearby residents even burn household garbage directly in the bridge opening, so that fire disasters are easily caused, and the texture structure of bridge pipeline facilities and the bridge is damaged.
The existing monitoring video is utilized to analyze the bridge opening space scene, and because the ordinary face detection recognition rate is not high, the face is required to face the camera and be static for a moment to be recognized in most cases, so that the face recognition accuracy is low, and whether the pedestrian invades in the bridge opening space is difficult to accurately detect.
The color characteristics of flame and smoke are utilized for detection, the detection is easily interfered by external environment, for example, red imaging at night is realized, flame loses an original color space, car light interference, haze interference and the like, the calculated amount is large in practical application, the scene adaptability is poor, and the phenomenon of false alarm missing is easily caused.
Technical problem to be solved
The invention provides a bridge opening space safety monitoring and alarming system based on deep learning, aiming at the defect that the phenomena of pedestrian invasion, garbage incineration and the like in the bridge opening space are difficult to accurately detect in the prior art.
(II) technical scheme
In order to achieve the purpose, the invention is realized by the following technical scheme:
a bridge opening space safety monitoring and alarming system based on deep learning comprises a video acquisition module, a pedestrian invasion detection module, a smoke and fire detection module, an information management module and an alarming module; the video acquisition module is respectively connected with the input ends of the pedestrian invasion detection module and the smoke and fire detection module, the output ends of the pedestrian invasion detection module and the smoke and fire detection module are respectively connected with the input end of the information management module, and the output end of the information management module is connected with the alarm module;
the video acquisition module acquires a bridge opening video by using a monitoring camera;
the pedestrian intrusion detection module captures pedestrians intruding into the bridge opening space through the video acquisition module, and detects pedestrian intrusion behaviors by adopting videos acquired by the deep learning face detection and matching analysis bridge opening space monitoring camera;
the smoke and fire detection module trains smoke and fire data by adopting a deep learning neural network model and detects smoke and fire by deep learning residual error network training image characteristics;
the information management module establishes a pedestrian invasion data management and firework data management base and manages data information involved in the bridge opening space safety monitoring process;
the alarm module sends out an alarm signal to the detection of pedestrian invasion or smoke and fire events.
According to one embodiment of the invention, the information management module sends the information of abnormal time, place and people to inspection personnel for alarming, marks warning icons of pedestrian invasion and smoke on a map interface, and realizes a report form containing the time, place and event type of the alarm event for subsequent viewing of historical events and data analysis.
According to an embodiment of the present invention, the pedestrian intrusion detection module for detecting pedestrian intrusion includes the following steps:
s1, establishing a pedestrian intrusion model based on deep learning neural network training bridge opening space;
s2, capturing pedestrians invading the bridge opening space through the network camera, and processing the acquired data through deep learning face recognition.
According to an embodiment of the present invention, in step S1, the open-source image set is trained first, and iterates 200000 times to obtain a general pedestrian feature representation model, and on the basis of the general model, a bridge opening space and a pedestrian sample are added to perform a detailed training to obtain a bridge opening pedestrian intrusion detection model.
According to an embodiment of the present invention, in step S2, once it is recognized that someone invades the bridge opening space, alarm information including information about the place of the incident and the person of the incident is sent to the inspection staff, and meanwhile, whether the intruder invades the bridge opening for many times is judged by face matching, if the face matching is the intruder appearing for many times, the inspection staff is informed to check the information management record and judge whether the intruder stays in the bridge opening by combining with the video picture, if so, the inspection staff is informed to move away as soon as possible, and if suspected thieves and the like, the alarm information is timely reflected to the relevant departments.
According to an embodiment of the present invention, the smoke and fire detection process performed by the smoke and fire detection module is as follows:
collecting roadside burning images through a web crawler technology, adding normal scene images of a bridge opening space to form a positive sample set and a negative sample set, training a classifier in a Caffe environment, and identifying three scenes of normal, fire and smoke;
training smoke and fire data by adopting a deep learning neural network model, outputting 3 types by a full connection layer, and respectively representing fire, smoke and normal scenes;
when the system detects smoke and fire, the system sends the information of time, place and people to the inspection personnel for timely and effective management, thereby ensuring the safety of the bridge opening space.
(III) advantageous effects
The invention has the beneficial effects that: a bridge opening space safety monitoring and alarming system based on deep learning has the following beneficial effects:
(1) the method has the advantages that the pedestrian invasion behavior is detected by adopting the deep learning face detection and matching analysis of the video collected by the bridge opening space monitoring camera, and compared with the traditional shallow learning technology for detecting and matching the face through simple binocular features, the method not only solves the condition limitation of face detection, but also improves the detection accuracy.
(2) The method for detecting smoke and fire based on the deep learning neural network training image features is adopted, so that the problems that the flame and smoke are easily interfered by external environments such as red imaging at night, the flame losing original color, car light interference and the like when being detected by using the color features are solved, the problem of large calculation amount is solved, and the scene adaptability and the detection hit rate are improved.
(3) The system comprises an information management module, wherein warning icons of pedestrian invasion and smoke are marked on a map interface, historical data and a current data report including time, place and event type of an alarm event are realized, and the historical event can be checked later; the information management module can not only save current data, but also can strengthen the inspection force of the bridge opening with frequent and multiple smoke and fire phenomena caused by pedestrian invasion by analyzing historical data, call out the picture of the monitoring camera, analyze the reason of the picture, take corresponding measures, correspondingly punish the smoke and fire caused by malicious persons, and ensure the safety of the bridge opening space.
(4) The bridge opening abnormity alarm system comprises a bridge opening abnormity alarm module, videos collected by a monitoring camera are analyzed based on the deep learning face detection and matching technology and the image classification technology, pedestrian invasion and smoke are detected, an alarm signal is sent to an inspection worker in time, the alarm module sends a place of affair to the inspection worker, and the inspection worker can conveniently arrive at the site in time to process abnormal events.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts.
FIG. 1 is a schematic block diagram of the present invention;
FIG. 2 is a flow chart of the present invention;
FIG. 3 is a flow chart of smoke and fire detection;
FIG. 4 is an information management flow diagram;
fig. 5 is an alarm flow chart.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, but not all, embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
With reference to fig. 1, a bridge opening space safety monitoring and alarming system based on deep learning comprises a video acquisition module, a pedestrian intrusion detection module, a smoke and fire detection module, an information management module and an alarming module; the video acquisition module is respectively connected with the input ends of the pedestrian invasion detection module and the smoke and fire detection module, the output ends of the pedestrian invasion detection module and the smoke and fire detection module are respectively connected with the input end of the information management module, and the output end of the information management module is connected with the alarm module.
The video acquisition module acquires a bridge opening video by using the monitoring camera.
The pedestrian intrusion detection module captures pedestrians intruding into the bridge opening space through the video acquisition module, and detects pedestrian intrusion behaviors by adopting deep learning face detection and matching analysis of videos acquired by the bridge opening space monitoring camera.
The smoke and fire detection module trains smoke and fire data by adopting a deep learning neural network model, and smoke and fire are detected by training image characteristics through a deep learning residual error network.
The information management module establishes a pedestrian invasion data management and firework data management base and manages data information involved in the bridge opening space safety monitoring process.
The alarm module sends out an alarm signal to the detection of the invasion of pedestrians or the occurrence of smoke and fire events.
With reference to fig. 2, a specific flow chart is shown, in which a video acquisition module acquires a bridge opening video through a monitoring camera, trains a correlation model of a bridge opening space through a deep learning residual error network, and analyzes the video acquired by the bridge opening monitoring camera; the method comprises the steps of loading the pedestrian invasion detection module and the smoke and fire detection module, and detecting whether pedestrian invasion and smoke and fire exist in a bridge opening space by using a bridge opening pedestrian invasion detection model and a smoke and fire classifier; the information management module sends the abnormal time, place and person information to inspection personnel for alarming, meanwhile, warning icons of pedestrian invasion and smoke and fire are marked on a map interface, and reports containing the time, place and event types of the alarm events are realized for follow-up viewing of historical events and data analysis.
Aiming at the bridge opening frequently invaded by pedestrians, a guardrail is additionally arranged, the inspection force is enhanced, and corresponding measures are taken for invading personnel still appearing after multiple warnings so as to ensure the safety of the bridge opening space; and calling a monitoring camera picture aiming at the bridge opening with the smoke and fire phenomenon for many times, analyzing the generated reason, and taking corresponding measures to correspondingly penalize the smoke and fire caused by malicious persons so as to ensure the safety of the bridge opening space.
The pedestrian intrusion detection module for detecting the intrusion of the pedestrian comprises the following steps:
s1, establishing a pedestrian intrusion model based on deep learning neural network training bridge opening space;
firstly, training an open source image set, iterating 200000 times to obtain a general pedestrian characteristic representation model, adding a bridge opening space and pedestrian samples on the basis of the general model, and performing detailed training to obtain a bridge opening pedestrian intrusion detection model;
s2, capturing pedestrians invading the bridge opening space through the network camera, and processing the acquired data through deep learning face recognition.
Using a cross entropy loss function when performing face detection:
bounding box regression uses the sum of squares loss function:
face feature point localization also uses the sum of squares loss function:
there are many different tasks across the convolutional neural network framework, so the following functions are used in the multitask training:
Once the fact that people invade the bridge opening space is recognized, alarm information including the information of the accident site and the accident personnel is sent to the inspection personnel, whether the intruder invades the bridge opening for multiple times or not is judged through face matching, if the intruder appears for multiple times, the inspection personnel is informed to check information management records and judge whether the intruder stays at the place by combining video pictures, if yes, the inspection personnel is informed to move away as soon as possible, and if suspected thieves and the like, the alarm information is timely reflected to relevant departments.
As the appearance characteristic of smoke and fire is changed greatly, false alarm is easy to miss under the interference of the external environment, and the adaptability of the scene is poor. In conjunction with the smoke and fire detection flow diagram of fig. 3, the smoke and fire detection module performs smoke and fire detection. Collecting roadside burning images through a web crawler technology, adding normal scene images of a bridge opening space to form a positive sample set and a negative sample set, training a classifier in a Caffe environment, and identifying three scenes of normal, fire and smoke; training smoke and fire data by adopting a deep learning neural network model, outputting 3 types by a full connection layer, and respectively representing fire, smoke and normal scenes; when the system detects smoke and fire, the system sends the information of time, place and people to the inspection personnel for timely and effective management, thereby ensuring the safety of the bridge opening space.
In the process of processing the image by applying the deep learning convolution neural network, calculating the characteristics of the image by adopting matrix convolution, wherein the definition of the effective value convolution is as follows:
when calculating the activation value by forward propagation, the output of the convolutional layer with the previous layer as the input layer is:
The output of the sub-sampling layer is:
using an averaging pooling method, using convolutionThe weight of each unit of the kernel is beta(l+1)After each convolution operation, a bias unit b is still added(l+1)。
The output of the convolutional layer with the next layer being a sub-sampling layer is:
management of data information is involved in the process of bridge opening space safety monitoring, and as shown in fig. 4, an information management module respectively establishes a pedestrian invasion data management base and a firework data management base. Pedestrian intrusion mainly records the time, the place (namely, a bridge opening address), people, the intrusion frequency and the number of people intruding at the same time. The time record is favorable for analyzing the time tendency of pedestrian invasion, the site record is favorable for analyzing information such as safety factor of a bridge opening, geographic advantages of the bridge opening and the like, the person record aims to prevent bad persons such as thieves from gathering at squatting points, people who frequently invade need to find the person through face identification information positioning, ask the person for questions and the like, and the danger of the space of the bridge opening and the nearby space caused by bad behaviors is prevented. The information management aiming at fireworks mainly comprises the steps of recording time and places, strengthening patrol of bridge holes with frequent fireworks, calling out monitoring camera pictures, analyzing reasons of the pictures, taking corresponding measures, carrying out corresponding punishment on the fireworks caused by malicious persons, avoiding occurrence of heavy loss and ensuring safety of bridge hole spaces. The management and the recording of the information are important bases for later investigation and data analysis, and have good preventive effect on the safety of the bridge opening space.
Combine 5 warning flow charts of figure, carry out the analysis to the video that surveillance camera machine gathered based on degree of depth learning face detection and matching and image classification, detect pedestrian's invasion and firework and in time send alarm signal to patrolling and examining personnel, alarm module can send the personnel of patrolling and examining with the place of affairs.
In summary, the bridge opening space safety monitoring and alarming system based on deep learning in the embodiment of the invention has the following advantages:
(1) the method has the advantages that the pedestrian invasion behavior is detected by adopting the deep learning face detection and matching analysis of the video collected by the bridge opening space monitoring camera, and compared with the traditional shallow learning technology for detecting and matching the face through simple binocular features, the method not only solves the condition limitation of face detection, but also improves the detection accuracy.
(2) The method for detecting smoke and fire based on the deep learning neural network training image features is adopted, so that the problems that the flame and smoke are easily interfered by external environments such as red imaging at night, the flame losing original color, car light interference and the like when being detected by using the color features are solved, the problem of large calculation amount is solved, and the scene adaptability and the detection hit rate are improved.
(3) The system comprises an information management module, wherein warning icons of pedestrian invasion and smoke are marked on a map interface, historical data and a current data report including time, place and event type of an alarm event are realized, and the historical event can be checked later; the information management module can not only save current data, but also can strengthen the inspection force of the bridge opening with frequent and multiple smoke and fire phenomena caused by pedestrian invasion by analyzing historical data, call out the picture of the monitoring camera, analyze the reason of the picture, take corresponding measures, correspondingly punish the smoke and fire caused by malicious persons, and ensure the safety of the bridge opening space.
(4) The bridge opening abnormity alarm system comprises a bridge opening abnormity alarm module, videos collected by a monitoring camera are analyzed based on the deep learning face detection and matching technology and the image classification technology, pedestrian invasion and smoke are detected, an alarm signal is sent to an inspection worker in time, the alarm module sends a place of affair to the inspection worker, and the inspection worker can conveniently arrive at the site in time to process abnormal events.
The above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.
Claims (2)
1. A bridge opening space safety monitoring and alarming system based on deep learning is characterized by comprising a video acquisition module, a pedestrian invasion detection module, a smoke and fire detection module, an information management module and an alarming module; the video acquisition module is respectively connected with the input ends of the pedestrian invasion detection module and the smoke and fire detection module, the output ends of the pedestrian invasion detection module and the smoke and fire detection module are respectively connected with the input end of the information management module, and the output end of the information management module is connected with the alarm module;
the video acquisition module acquires a bridge opening video by using a monitoring camera;
the pedestrian intrusion detection module captures pedestrians intruding into the bridge opening space through the video acquisition module, and detects pedestrian intrusion behaviors by adopting videos acquired by the deep learning face detection and matching analysis bridge opening space monitoring camera;
the smoke and fire detection module trains smoke and fire data by adopting a deep learning neural network model and detects smoke and fire by deep learning residual error network training image characteristics;
the information management module establishes a pedestrian invasion data management and firework data management base and manages data information involved in the bridge opening space safety monitoring process;
the alarm module sends an alarm signal to the detection of pedestrian invasion or smoke and fire events;
the pedestrian intrusion detection module for carrying out pedestrian intrusion detection comprises the following steps:
s1, establishing a pedestrian intrusion model based on deep learning neural network training bridge opening space:
firstly, training an open source image set, iterating 200000 times to obtain a general pedestrian characteristic representation model, adding a bridge opening space and pedestrian samples on the basis of the general model, and performing detailed training to obtain a bridge opening pedestrian intrusion detection model;
s2, capturing pedestrians invading the bridge opening space through a network camera, and processing the acquired data through deep learning face recognition:
once the step S2 identifies that someone invades the bridge opening space, alarm information is sent to the inspection personnel, wherein the alarm information comprises information of an incident place and the incident personnel, meanwhile, whether the intruder invades the bridge opening for multiple times is judged through face matching, if the intruder appears for multiple times, the inspection personnel is informed to check information management records and judge whether the intruder stays at the place by combining video pictures, if yes, the inspection personnel is informed to move away as soon as possible, and if suspected, the inspection personnel timely reflects the information to relevant departments;
the firework detection flow of the firework detection module is as follows:
collecting roadside burning images through a web crawler technology, adding normal scene images of a bridge opening space to form a positive sample set and a negative sample set, training a classifier in a Caffe environment, and identifying three scenes of normal, fire and smoke;
training smoke and fire data by adopting a deep learning neural network model, outputting 3 types by a full connection layer, and respectively representing fire, smoke and normal scenes;
when the system detects smoke and fire, the system sends the information of time, place and people to the inspection personnel for timely and effective management, thereby ensuring the safety of the bridge opening space.
2. The bridge opening space safety monitoring and alarming system based on deep learning of claim 1, wherein the information management module sends information of abnormal time, place and people to inspection personnel for alarming, meanwhile, warning icons of pedestrian invasion and smoke and fire are marked on a map interface, and a report containing time, place and event types of alarm events is realized for follow-up viewing of historical events and data analysis.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811383003.2A CN109410497B (en) | 2018-11-20 | 2018-11-20 | Bridge opening space safety monitoring and alarming system based on deep learning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811383003.2A CN109410497B (en) | 2018-11-20 | 2018-11-20 | Bridge opening space safety monitoring and alarming system based on deep learning |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109410497A CN109410497A (en) | 2019-03-01 |
| CN109410497B true CN109410497B (en) | 2021-01-19 |
Family
ID=65474093
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811383003.2A Active CN109410497B (en) | 2018-11-20 | 2018-11-20 | Bridge opening space safety monitoring and alarming system based on deep learning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109410497B (en) |
Families Citing this family (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111553264B (en) * | 2020-04-27 | 2023-04-18 | 中科永安(安徽)科技有限公司 | Campus non-safety behavior detection and early warning method suitable for primary and secondary school students |
| CN111432182B (en) * | 2020-04-29 | 2021-10-08 | 上善智城(苏州)信息科技有限公司 | Safety supervision method and system for oil discharge place of gas station |
| CN113657298A (en) * | 2021-08-20 | 2021-11-16 | 软通动力信息技术(集团)股份有限公司 | Pedestrian intrusion identification method, device, equipment and medium based on large displacement tracking |
| CN114267082B (en) * | 2021-09-16 | 2023-08-11 | 南京邮电大学 | Bridge side falling behavior identification method based on depth understanding |
| CN115082834B (en) * | 2022-07-20 | 2023-03-17 | 成都考拉悠然科技有限公司 | Engineering vehicle black smoke emission monitoring method and system based on deep learning |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106372576A (en) * | 2016-08-23 | 2017-02-01 | 南京邮电大学 | Deep learning-based intelligent indoor intrusion detection method and system |
| CN106355162A (en) * | 2016-09-23 | 2017-01-25 | 江西洪都航空工业集团有限责任公司 | Method for detecting intrusion on basis of video monitoring |
| US10691950B2 (en) * | 2017-03-10 | 2020-06-23 | Turing Video, Inc. | Activity recognition method and system |
| CN108389359B (en) * | 2018-04-10 | 2020-03-24 | 中国矿业大学 | Deep learning-based urban fire alarm method |
| CN108777777A (en) * | 2018-05-04 | 2018-11-09 | 江苏理工学院 | A kind of monitor video crop straw burning method for inspecting based on deep neural network |
-
2018
- 2018-11-20 CN CN201811383003.2A patent/CN109410497B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN109410497A (en) | 2019-03-01 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109410497B (en) | Bridge opening space safety monitoring and alarming system based on deep learning | |
| CN1148710C (en) | Monitoring system | |
| CN101751744B (en) | Detection and early warning method of smoke | |
| CN106650584B (en) | Flame detecting method and system | |
| CN106454250A (en) | Intelligent recognition and early warning processing information platform | |
| CN210899299U (en) | Tunnel monitoring system | |
| CN201424042Y (en) | Rail transportation road condition automatic alarm system | |
| CN108009488A (en) | The street security joint defense system of Behavior-based control analysis | |
| CN113011833A (en) | Safety management method and device for construction site, computer equipment and storage medium | |
| CN110867046A (en) | Intelligent car washer video monitoring and early warning system based on cloud computing | |
| CN107729850A (en) | Broadcast system is supervised in Internet of Things outdoor advertising | |
| KR20190035187A (en) | Sound alarm broadcasting system in monitoring area | |
| CN114724330A (en) | Implementation method of self-adaptive mode switching multi-channel video fire real-time alarm system | |
| CN111476964A (en) | Remote forest fire prevention monitoring system and method | |
| CN112383615A (en) | Residential building entrance guard security system and method based on edge calculation | |
| KR20200052418A (en) | Automated Violence Detecting System based on Deep Learning | |
| CN112767196A (en) | Intelligent property management information sharing platform for smart community | |
| CN110780356A (en) | Subway platform clearance foreign matter detecting system | |
| CN113538825A (en) | Campus wall-turning event alarm method and system | |
| CN112132048A (en) | Community patrol analysis method and system based on computer vision | |
| CN204375138U (en) | Based on the intelligent early-warning system of people's current density recognition technology | |
| CN116434533A (en) | AI wisdom highway tunnel synthesizes monitoring platform based on 5G | |
| CN109274945B (en) | Method and system for self-adaptively performing true color restoration on image | |
| KR101046819B1 (en) | Method and system for watching an intrusion by software fence | |
| CN210222962U (en) | Intelligent electronic fence system |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |