WO2024124750A1 - 一种基于图像识别技术的工业场所不安全行为识别系统 - Google Patents

一种基于图像识别技术的工业场所不安全行为识别系统 Download PDF

Info

Publication number
WO2024124750A1
WO2024124750A1 PCT/CN2023/086167 CN2023086167W WO2024124750A1 WO 2024124750 A1 WO2024124750 A1 WO 2024124750A1 CN 2023086167 W CN2023086167 W CN 2023086167W WO 2024124750 A1 WO2024124750 A1 WO 2024124750A1
Authority
WO
WIPO (PCT)
Prior art keywords
module
regional
unit
layer
information
Prior art date
Application number
PCT/CN2023/086167
Other languages
English (en)
French (fr)
Inventor
赵文博
周刚
徐柯
邓岳辉
吴昌兵
张伟
向东
刘晶
苏林
张晨雨
张双
康倩
石灿
冯鹏宇
张大勇
陈柄宏
李涛
Original Assignee
华能重庆两江燃机发电有限责任公司
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by 华能重庆两江燃机发电有限责任公司 filed Critical 华能重庆两江燃机发电有限责任公司
Priority to ZA2023/09548A priority Critical patent/ZA202309548B/en
Publication of WO2024124750A1 publication Critical patent/WO2024124750A1/zh

Links

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W4/00—Services specially adapted for wireless communication networks; Facilities therefor
    • H04W4/02—Services making use of location information
    • H04W4/029—Location-based management or tracking services
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N20/00—Machine learning
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/10—Image acquisition
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
    • G06V10/764—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using classification, e.g. of video objects
    • 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

Definitions

  • the present invention relates to the technical field of unsafe behavior recognition, and in particular to an industrial site unsafe behavior recognition system based on image recognition technology.
  • the existing method is usually to increase the number of supervisors and use manual supervision to supervise the working status of workers.
  • the manual method still has major disadvantages, such as the inability to monitor in real time and the inability to communicate in time when encountering situations. It is even more difficult to supervise and manage them centrally, which can easily lead to weak safety supervision measures and large monitoring blind spots. Not only can the safety of workers not be guaranteed, but it can also easily bring unnecessary troubles. Therefore, the present invention proposes an industrial site unsafe behavior recognition system based on image recognition technology to solve the problems existing in the prior art.
  • the purpose of the present invention is to propose an industrial site unsafe behavior identification system based on image recognition technology.
  • This industrial site unsafe behavior identification system based on image recognition technology has the advantages of real-time monitoring and real-time fixed-point reporting, and solves the problems existing in the prior art.
  • an industrial site unsafe behavior identification system based on image recognition technology, including a device layer, a data layer, a management layer and a communication layer, the device layer, the data layer and the management layer are connected through the communication layer, and the device layer includes an image acquisition module for collecting personnel work, an information communication module for conveying warning information and a control terminal for supervising personnel operations, the data layer includes a regional database for regional computing and storage and a cloud database for cloud computing and storage, the management layer includes a personnel management module for personnel information management, a device management module for equipment information management and a communication management module for communication information management, the communication layer includes a 4G communication module, a 5G communication module and a wireless communication module, the information communication module is provided with a voice broadcast unit for broadcasting prompt voice, a personnel positioning unit for locating personnel positions, a personnel information binding unit for binding personnel information, a wireless connection unit for wireless connection and a rechargeable battery unit.
  • the regional database is provided with multiple groups, and the multiple groups of regional databases are connected to the cloud database, the multiple groups of regional databases correspond to one region respectively, and the regional database is provided with a regional data processing unit, a regional data calculation unit, a regional data storage unit, a regional personnel information storage unit and a regional system operating platform.
  • the cloud database includes a cloud data acquisition unit, a cloud data calculation unit, a cloud model training unit, a recognition model storage unit, a recognition model decentralization unit, a cloud data storage unit and a cloud operation platform.
  • a further improvement is that: the regional database is also provided with a regional 3D visualization module, the regional 3D visualization module is connected to the regional system operation platform, and the regional 3D visualization module is provided with a 3D model map of the building of the corresponding area, the regional system
  • the system operation platform is equipped with a voice information editing module and a voice notification module.
  • a further improvement is that it also includes a power supply layer, which includes a device power module for providing power to the device, a solar power generation module for clean energy power generation, and a charging module for charging.
  • a further improvement is that a timing module and a password unlocking module are also provided in the regional system operating platform, and the output end of the timing module is connected to the input end of the password unlocking module.
  • a further improvement is that a log recording module and a log uploading module are also provided in the regional database, the output end of the log recording module is connected to the input end of the log uploading module, and the output end of the log uploading module is connected to the input end of the cloud database.
  • a further improvement is that an alarm unit is also provided in the information transmission module, and an output end of the alarm unit is connected to a regional database via a wireless connection unit.
  • control terminal is a mobile phone or a tablet computer.
  • a further improvement is that a pressure monitoring unit is provided in the information transmission module, and the output end of the pressure monitoring unit is connected to the input end of the alarm unit.
  • the beneficial effects of the present invention are as follows: the industrial workplace unsafe behavior identification system based on image recognition technology is bound to each staff member through the information communication module set up, and then the image acquisition module collects the work image data of the staff member and identifies it by referring to the image recognition technology.
  • the corresponding staff member’s information communication module can be used to give a voice broadcast reminder, which is convenient for the staff member to correct it, and solves the problems of difficult supervision and inadequate communication in the prior art, thereby ensuring the life safety and property safety of the operating personnel, effectively implementing the safety management responsibilities of enterprises and personnel at all levels, and reducing the unsafe behavior of the staff in the industrial workplace, that is, effectively reducing The probability of occurrence of low unsafe events is reduced, and the regional database corresponds to one area, so that it does not need to calculate the data of other areas, thereby reducing the calculation amount of each regional database, thereby improving the efficiency and effect of unsafe behavior identification in each regional database.
  • FIG1 is a schematic diagram of the system architecture of the present invention.
  • FIG. 2 is a schematic diagram of the information communication module architecture of the present invention.
  • FIG. 3 is a schematic diagram of the connection status between the regional database and the cloud database of the present invention.
  • this embodiment proposes an unsafe behavior identification system for industrial sites based on image recognition technology, including a device layer, a data layer, a management layer and a communication layer.
  • the device layer, the data layer and the management layer are connected through the communication layer, and the device layer includes an image acquisition module for collecting personnel work, an information communication module for conveying warning information and a control terminal for supervising personnel operations.
  • the specific image acquisition module corresponds to a panoramic camera.
  • the image acquisition module is provided with multiple groups, that is, multiple groups of panoramic cameras are used to monitor the working status of personnel in industrial sites to obtain corresponding image data, which is convenient for subsequent identification.
  • the image information communication module corresponds to an information communicator.
  • the information communication module is provided with a voice broadcast unit for broadcasting prompt voice, a personnel positioning unit for locating personnel positions, a personnel information binding unit for binding personnel information, a wireless connection unit for wireless connection, and a rechargeable battery unit.
  • the information communicator has the functions of voice broadcast and personnel positioning, and the information communication module is provided with multiple groups, that is, each staff member entering the industrial site is equipped with an information communication module, and at the same time, each staff member is bound to the corresponding information communication module through the personnel information binding unit.
  • a voice broadcast will be played from the information communication module of the staff member, so as to remind the staff member to make corrections and ensure the safety of the staff member
  • the wireless connection unit is used for wireless connection of the information communication module to provide convenience
  • the rechargeable battery unit provides a rechargeable and recyclable function for the information communication module, without the need to replace the battery, reducing unnecessary equipment investment
  • the further personnel positioning unit is used to record the range of the corresponding staff member in the industrial site, so as to facilitate subsequent tracing.
  • the further data layer includes a regional database for regional computing and storage and a cloud database for cloud computing and storage.
  • a plurality of regional databases are provided, and the plurality of regional databases are connected to the cloud database.
  • the plurality of regional databases correspond to one region respectively, and the regional database is provided with a regional data processing unit for processing data within the region, a regional data computing unit for computing according to an unsafe behavior model, a regional data storage unit for storing regional data, a regional personnel information storage unit for recording personnel information corresponding to the region, and a regional system operation platform.
  • a login module is provided in the regional operation platform, and relevant management personnel are required to log in to perform corresponding operations.
  • the data collected by the image acquisition module enters the regional data processing unit, and the regional data processing unit performs image processing on the data, such as data noise reduction, data feature extraction, etc., and then the regional data computing unit calculates according to the unsafe behavior model to determine whether the current image data is within the regional data processing unit. Whether there are any unsafe behaviors of the staff, whether there are any or not, are all stored in the regional data storage unit. Specifically, since the scope of the industrial site itself is relatively wide, the number of image acquisition modules deployed is also very large, and the amount of data generated is also very large. If only one database is used to calculate the data, the recognition efficiency of the system will be greatly reduced, resulting in the failure to timely transmit the reminder correction information.
  • the industrial site is divided into multiple areas, each area corresponds to a regional database, that is, a distributed computing method is adopted, the regional database performs the calculation, and the cloud database performs the unsafe behavior model training and other operations, which greatly reduces the efficiency of each area.
  • the data that needs to be calculated by the database is collected, thereby improving the calculation efficiency of the regional database, that is, improving the efficiency and effect of identifying unsafe behaviors.
  • the facial data of the staff can also be identified.
  • the personnel information stored in the regional personnel information storage unit will be called and compared to confirm the personnel information corresponding to the unsafe behavior.
  • the system automatically sends instructions to the voice broadcast unit in the corresponding information transmission module for broadcast, and then reminds the staff to make corrections.
  • the broadcast time is 15s, so there is no need for the supervisor to communicate again, which is highly efficient.
  • unsafe behaviors in industrial sites include but are not limited to not using safety belts when working at heights; safety belts are not hung on solid components when working at heights; not wearing safety helmets when working, resting on high platforms, at the edge of holes, in safety nets, or leaning on railings, etc. Therefore, when unsafe behaviors occur, the information communication module corresponding to the staff will send out a broadcast message to remind the staff to make corrections.
  • a log recording module and a log uploading module are also provided in the regional database. The output end of the log recording module is connected to the input end of the log uploading module. The output end of the log upload module is connected to the input end of the cloud database. The log recording module records the personnel information of each time the broadcast command is triggered, and attaches the corresponding timeline to facilitate subsequent tracing, so as to strengthen the safety operation knowledge in a targeted manner and ensure that subsequent staff can operate safely.
  • the further management layer includes a personnel management module for personnel information management, a device management module for device information management, and a communication management module for communication information management.
  • the personnel management module is the overall management module, which records the personnel information corresponding to each area, while the device management module for device information management is used to manage the status of each device in the system.
  • the communication layer includes a 4G communication module, a 5G communication module and a wireless communication module to provide network continuity.
  • the specific cloud database includes a cloud data acquisition unit, a cloud data calculation unit, a cloud model training unit, a recognition model storage unit, a recognition model decentralization unit, a cloud data storage unit and a cloud operation platform, wherein the cloud data acquisition unit is connected to multiple groups of regional databases to obtain data in the regional databases. At the same time, it can also obtain data input from the outside. Then, for the unsafe behavior model, it is necessary for personnel to first input on-site materials of the industrial site.
  • the on-site materials correspond to different safety behavior materials, thereby generating corresponding training data sets, and then the cloud model training unit uses the training data sets for classification training to obtain an unsafe behavior recognition model for the industrial site, and then the recognition model decentralization unit decentralizes it to the regional database, and the regional database recognizes the collected image data according to the trained unsafe behavior model.
  • the regional database also has a regional 3D visualization module, which is connected to the regional system operation platform. After the system operation platform is opened, the map information of the corresponding area can be observed through the regional 3D visualization module. At the same time, the regional 3D visualization module is provided with a 3D model diagram of the building in the corresponding area.
  • the regional 3D visualization module corresponds to the image acquisition module, that is, the position corresponding to each group of image acquisition modules is reflected in the regional 3D visualization module, which is convenient for management personnel to view.
  • a voice information editing module and a voice notification module are provided in the specific regional system operation platform. The voice information editing module is used to edit the voice information to be sent, and the voice notification module is connected to the information communication module through wireless to send out relevant voice broadcast information.
  • an alarm unit is also provided in the information communication module.
  • the output end of the alarm unit is connected to the regional database through a wireless connection unit.
  • the alarm unit corresponds to a button on the information communicator.
  • the staff presses the button, that is, the information communicator (information communication module) sends an alarm message to the regional database. At this time, it is directly fed back to the regional operation platform, and then the management personnel can directly see it and perform subsequent processing.
  • the system also includes a power supply layer, which includes a device power module for providing power to the device, a solar power generation module for clean energy generation, and a charging module for charging. That is, clean energy is used to generate electricity to achieve an energy-saving effect and reduce costs.
  • the corresponding device power module is mainly used to power the image acquisition module.
  • the information communication module can be charged through the set charging module, and the charging method is wireless charging.
  • a timing module and a password unlocking module are provided in the regional system operation platform.
  • the output end of the timing module is connected to the input end of the password unlocking module.
  • the timing module is set to perform timing. The timing time 30 minutes later, the regional system operation platform will be locked and you need to enter the correct password through the password unlocking module before you can operate.
  • control terminal is a mobile phone or tablet computer, that is, the supervisors in each area can connect to the corresponding regional operating platform through the control terminal to obtain relevant information, and can also upload corresponding information.
  • the control terminal will receive information sent by the regional operating platform. Since each regional operating platform corresponds to one area, the supervisors in the corresponding area can be notified in time to deal with it.
  • a pressure monitoring unit is provided in the information communication module, and the output end of the pressure monitoring unit is connected to the input end of the alarm unit.
  • the information communication module corresponds to the information communicator, which itself has a shell.
  • the pressure felt by the pressure monitoring unit will be greatly stimulated, thereby triggering the alarm unit to issue relevant alarm instructions.
  • the main purpose is to ensure that when an accident occurs, it can be dealt with in a timely manner.

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Physics & Mathematics (AREA)
  • Software Systems (AREA)
  • Multimedia (AREA)
  • Medical Informatics (AREA)
  • Evolutionary Computation (AREA)
  • Artificial Intelligence (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Computing Systems (AREA)
  • General Health & Medical Sciences (AREA)
  • Databases & Information Systems (AREA)
  • Health & Medical Sciences (AREA)
  • Signal Processing (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Data Mining & Analysis (AREA)
  • General Engineering & Computer Science (AREA)
  • Mathematical Physics (AREA)
  • Alarm Systems (AREA)

Abstract

本发明提出一种基于图像识别技术的工业场所不安全行为识别系统，包括设备层、数据层、管理层和通信层，设备层、数据层和管理层之间通过通信层连接，且设备层包括图像采集模块、信息传达模块和控制终端，数据层包括区域数据库和云端数据库，管理层包括人员管理模块、设备管理模块以及通信管理模块，通信层包括4G通信模块、5G通信模块以及无线通信模块，本发明通过设置的信息传达模块，使得其与每个工作人员进行绑定，继而再通过引用图像识别技术的方式，通过图像采集模块采集工作人员的工作图像数据，并进行识别，继而出现不安全行为时，能够定向的通过对应工作人员的信息传达模块进行语音播报提醒，方便工作人员改正。

Description

一种基于图像识别技术的工业场所不安全行为识别系统 技术领域
本发明涉及不安全行为识别技术领域，尤其涉及一种基于图像识别技术的工业场所不安全行为识别系统。
背景技术
工业场所的工作安全是工业场所正常运行的一项重要指标，如发生工人发生任何伤害事故，都会造成工业场所的重大损失。当工业场所在安全运行时，工人难免会存在一定的疏忽大意，存在一些不安全的行为，这些行为有可能会导致产生事故发生；
由于工业场所大，作业面积广，人员复杂，现有的方式通常是扩大监管人员的数量，利用人工对工人的工作状态进行监管，但人工的方式仍然存在较大的弊端，比如不能实时监控，遇到情况时不一定能传达的及时，而集中对其进行监督管理的难度更大，很容易导致安全监管手段弱、监控盲区大等情况出现，不仅使工人的安全不能得到保证，同时还容易带来不必要的麻烦，因此本发明提出一种基于图像识别技术的工业场所不安全行为识别系统以解决现有技术中存在的问题。
发明内容
针对上述问题，本发明的目的在于提出一种基于图像识别技术的工业场所不安全行为识别系统，该种基于图像识别技术的工业场所不安全行为识别系统具有实时监控、实时定点播报的优点，解决现有技术中存在的问题。
为实现本发明的目的，本发明通过以下技术方案实现：一种基于图像识别技术的工业场所不安全行为识别系统，包括设备层、数据层、管理层和通信层，所述设备层、数据层和管理层之间通过通信层连接，且设备层包括用于采集人员工作的图像采集模块、用于传达警告信息的信息传达模块和用于监管人员操作的控制终端，所述数据层包括用于区域计算存储的区域数据库和用于云端计算存储的云端数据库，所述管理层包括用于人员信息管理的人员管理模块、用于设备信息管理的设备管理模块以及用于通信信息管理的通信管理模块，所述通信层包括4G通信模块、5G通信模块以及无线通信模块，所述信息传达模块内设有用于播报提示语音的语音播报单元、用于定位人员位置的人员定位单元、用于绑定人员信息的人员信息绑定单元、用于无线连接的无线连接单元以及可充电蓄电池单元。
进一步改进在于：所述区域数据库设有多组，且多组区域数据库均与云端数据库连接，多组所述区域数据库分别对应一个区域，且区域数据库内设有区域数据处理单元、区域数据计算单元、区域数据存储单元、区域人员信息存储单元以及区域系统操作平台。
进一步改进在于：所述云端数据库包括云端数据获取单元、云端数据计算单元、云端模型训练单元、识别模型存储单元、识别模型下放单元、云端数据存储单元以及云端操作平台。
进一步改进在于：所述区域数据库内还设有区域地区3D可视化模块，所述区域地区3D可视化模块与区域系统操作平台连接，且区域地区3D可视化模块内设有对应区域的建筑3D模型图，所述区域系 统操作平台内设有语音信息编辑模块和语音通报模块。
进一步改进在于：还包括有供电层，所述供电层包括用于提供设备电源的设备电源模块、用于清洁能源发电的太阳能发电模块以及用于充电的充电模块。
进一步改进在于：所述区域系统操作平台内还设有定时模块和密码解锁模块，所述定时模块的输出端与密码解锁模块的输入端连接。
进一步改进在于：所述区域数据库内还设有日志记录模块和日志上传模块，所述日志记录模块的输出端与日志上传模块的输入端连接，所述日志上传模块的输出端与云端数据库的输入端连接。
进一步改进在于：所述信息传达模块内还设有警报单元，所述警报单元的输出端通过无线连接单元与区域数据库连接。
进一步改进在于：所述控制终端为手机或平板电脑中的一种。
进一步改进在于：所述信息传达模块内设有压力监测单元，所述压力监测单元的输出端与警报单元的输入端连接。
本发明的有益效果为：该种基于图像识别技术的工业场所不安全行为识别系统通过设置的信息传达模块，使得其与每个工作人员进行绑定，继而再通过引用图像识别技术的方式，通过图像采集模块采集工作人员的工作图像数据，并进行识别，继而出现不安全行为时，能够定向的通过对应工作人员的信息传达模块进行语音播报提醒，方便工作人员改正，解决现有技术中监管难度大、传达不到位等情况，保证了作业人员的生命安全和财产安全，有效落实企业和各级人员的安全管理责任，并且降低工作人员在工业场所的不安全行为，即有效降 低不安全事件的发生概率，再通过区域数据库对应一个区域的方式，使得其无需计算其他区域的数据，由此降低每个区域数据库的计算量，以此提升每个区域数据库的不安全行为识别的效率和效果。
附图说明
图1是本发明的系统架构示意图。
图2是本发明的信息传达模块架构示意图。
图3是本发明的区域数据库与云端数据库连接状态示意图。
具体实施方式
为了加深对本发明的理解，下面将结合实施例对本发明做进一步详述，本实施例仅用于解释本发明，并不构成对本发明保护范围的限定。
根据图1-图3所示，本实施例提出了一种基于图像识别技术的工业场所不安全行为识别系统，包括设备层、数据层、管理层和通信层，设备层、数据层和管理层之间通过通信层连接，且设备层包括用于采集人员工作的图像采集模块、用于传达警告信息的信息传达模块和用于监管人员操作的控制终端，具体的图像采集模块对应的是全景摄像头，在本实施例中，图像采集模块设有布设有多组，即通过多组全景摄像头来监控工业场所内工作人员的工作状态，以获取对应的图像数据，方便后续对进行识别，其次，图像信息传达模块对应的是信息传达器，进一步的信息传达模块内设有用于播报提示语音的语音播报单元、用于定位人员位置的人员定位单元、用于绑定人员信息的人员信息绑定单元、用于无线连接的无线连接单元以及可充电蓄电池单 元，即在本实施例中，信息传达器具有语音播报和人员定位的功能，并且，信息传达模块是设有多组，即每个进入工业场所的工作人员均配备一个信息传达模块，同时通过人员信息绑定单元，使每个工作人员与其对应的信息传达模块进行绑定，由此，当某个工作人员被识别到出现不安全行为时，语音播报会从该工作人员的信息传达模块播放出来，从而提醒该工作人员进行改正，确保工作人员的安全性，而其中的无线连接单元则是用于信息传达模块的无线连接，提供便利性，而可充电蓄电池单元则是为信息传达模块提供了一个可充电循环使用的作用，无需更换电池，减少不必要的设备投入，进一步的人员定位单元则是用于记录对应工作人员在工业场所到的范围，方便后续追溯。
进一步的数据层包括用于区域计算存储的区域数据库和用于云端计算存储的云端数据库，在本实施例中，区域数据库设有多组，且多组区域数据库均与云端数据库连接，多组区域数据库分别对应一个区域，且区域数据库内设有用于区域内数据处理的区域数据处理单元、用于根据不安全行为模型计算的区域数据计算单元、用于存储区域数据的区域数据存储单元、用于记录该区域对应人员信息的区域人员信息存储单元以及区域系统操作平台，区域操作平台内设有登录模块，需要相关的管理人员进行登录，从而进行相应的操作，对应的，由图像采集模块采集的数据进入区域数据处理单元内，由区域数据处理单元对其进行图像处理，比如数据降噪，数据特征提取等，之后再由区域数据计算单元根据不安全行为模型进行计算，判断当前图像数据内 是否存在工作人员有不安全行为，无论是否存在，均存入区域数据存储单元内，具体的，由于工业场所自身的范围是较广的，继而布设的图像采集模块的数量也非常广，由此其产生的数据量也是非常多的，若是仅采用一台数据库对其进行数据计算，则会大大降低了系统的识别效率，导致提醒改正的信息不能及时的传递，继而采用将工业场所划分为多个区域，每一个区域对应一个区域数据库，即采用分布式计算的方式，由区域数据库进行计算，而云端数据库进行不安全行为模型的训练等操作，大大降低了每个区域数据库所需要计算的数据，从而提升了区域数据库的计算效率，即提升了不安全行为的识别效率和效果，同时，在对图像数据进行识别时，还能够识别到工作人员的脸部数据，对此，在区域数据计算单元计算的过程中，会调用区域人员信息存储单元内存储的人员信息，并进行比对，确认出现不安全行为所对应的人员信息，再根据人员信息绑定的信息传达模块，由系统自动下放指令至对应信息传达模块内的语音播报单元播报，继而提醒工作人员进行修正，播报时间为15s，由此无需监管人员二次传达，效率高；
具体的，工业场所不安全行为包括但不限于高处作业不使用安全带；高处作业安全带未挂在牢固的构件上；作业不戴安全帽、在高处平台、孔洞边缘、安全网内休息或倚坐栏杆等多种行为，因此，在出现不安全行为时，与工作人员对应的信息传达模块发出播报信息，提醒工作人员进行更正，并且区域数据库内还设有日志记录模块和日志上传模块，日志记录模块的输出端与日志上传模块的输入端连接，日 志上传模块的输出端与云端数据库的输入端连接，对于每次触发播报指令的人员信息，由日志记录模块对其进行记录，并附上对应的时间轴，方便后续追溯，以进行针对性加强安全作业知识，确保后续工作人员能够安全作业。
进一步的管理层包括用于人员信息管理的人员管理模块、用于设备信息管理的设备管理模块以及用于通信信息管理的通信管理模块，人员管理模块是总管理模块，其记录了每个区域对应的人员信息，而设备信息管理的设备管理模块则是用于管理系统内各个设备状态的，同时通信层包括4G通信模块、5G通信模块以及无线通信模块，用于提供网络连续。
具体的云端数据库包括云端数据获取单元、云端数据计算单元、云端模型训练单元、识别模型存储单元、识别模型下放单元、云端数据存储单元以及云端操作平台，其中，云端数据获取单元是与多组区域数据库进行连接，用于获取区域数据库内的数据，同时，也可以获取外部所输入的数据，继而对于不安全行为模型是需要先由人员输入工业场所的现场素材，该现场素材及对应了不同安全行为素材，从而生成对应的训练数据集，再通过云端模型训练单元利用训练数据集进行分类训练，从而得到对工业场所不安全行为识别模型，再由识别模型下放单元下放至区域数据库内，由区域数据库根据训练后的不安全行为模型对采集的图像数据进行识。
进一步的区域数据库内还设有区域地区3D可视化模块，区域地区3D可视化模块与区域系统操作平台连接，即管理人员登录区域系 统操作平台后，可以通过区域地区3D可视化模块观察到对应区域的地图信息，同时区域地区3D可视化模块内设有对应区域的建筑3D模型图，在本实施例中，区域地区3D可视化模块与图像采集模块对应，即每组图像采集模块对应的位置均在区域地区3D可视化模块内体现出，方便管理人员进行查看，具体的区域系统操作平台内设有语音信息编辑模块和语音通报模块，语音信息编辑模块是用于编辑需要发出的语音信息，而语音通报模块则是通过无线与信息传达模块连接，发出相关的语音播报信息，同时信息传达模块内还设有警报单元，警报单元的输出端通过无线连接单元与区域数据库连接，警报单元对应的是信息传达器上设有一个按钮，遇到突发事情时，工作人员按下按钮，即信息传达器(信息传达模块)发出警报信息至区域数据库内，此时是直接的反馈到区域操作平台内，继而管理人员可以直接看到，并进行后续的处理。
具体的本系统还包括有供电层，供电层包括用于提供设备电源的设备电源模块、用于清洁能源发电的太阳能发电模块以及用于充电的充电模块，即利用清洁能源发电，来达到一个节能效果，降低成本，对应的设备电源模块主要是为图像采集模块进行供电，同时通过设置的充电模块可信息传达模块进行充电，其充电方式是采用无线充电的方式进行的。
进一步的区域系统操作平台内还设有定时模块和密码解锁模块，定时模块的输出端与密码解锁模块的输入端连接，为了防止非管理人员的操作，在本实施例中，通过设置的定时模块进行定时，定时时间 30min，到达时间后，区域系统操作平台出现锁定状态，需要通过密码解锁模块输入正确的密码，才可以进行操作，
进一步的控制终端为手机或平板电脑中的一种，即每个区域的监管人员均可以通过控制终端与对应区域操作平台进行连接，从而获取相关信息，也可以上传对应的信息，遇到紧急情况时，虽然系统会优先识别到，但无法进行救援，继而在出现意外情况时，控制终端会收到由区域操作平台发出的信息，因每个区域操作平台是对应一个区域的，继而能够及时的通知对应区域的监管人员，前去处理。
具体的在信息传达模块内设有压力监测单元，压力监测单元的输出端与警报单元的输入端连接，由于在本实施例中，信息传达模块对应的是信息传达器，其自身是具有外壳的，当信息传达模块自身出现问时，比如受力破碎等情况，压力监测单元其感受到的压力会出现巨大的拨动，由此来触发警报单元，使其发出相关的警报指令，主要的目的就是为了确保在发意外情况时，能够及时的进行处理。
以上显示和描述了本发明的基本原理、主要特征和优点。本行业的技术人员应该了解，本发明不受上述实施例的限制，上述实施例和说明书中描述的只是说明本发明的原理，在不脱离本发明精神和范围的前提下，本发明还会有各种变化和改进，这些变化和改进都落入要求保护的本发明范围内。本发明要求保护范围由所附的权利要求书及其等效物界定。

Claims (10)

  1. 一种基于图像识别技术的工业场所不安全行为识别系统，其特征在于：包括设备层、数据层、管理层和通信层，所述设备层、数据层和管理层之间通过通信层连接，且设备层包括用于采集人员工作的图像采集模块、用于传达警告信息的信息传达模块和用于监管人员操作的控制终端，所述数据层包括用于区域计算存储的区域数据库和用于云端计算存储的云端数据库，所述管理层包括用于人员信息管理的人员管理模块、用于设备信息管理的设备管理模块以及用于通信信息管理的通信管理模块，所述通信层包括4G通信模块、5G通信模块以及无线通信模块，所述信息传达模块内设有用于播报提示语音的语音播报单元、用于定位人员位置的人员定位单元、用于绑定人员信息的人员信息绑定单元、用于无线连接的无线连接单元以及可充电蓄电池单元。
  2. 根据权利要求1所述的一种基于图像识别技术的工业场所不安全行为识别系统，其特征在于：所述区域数据库设有多组，且多组区域数据库均与云端数据库连接，多组所述区域数据库分别对应一个区域，且区域数据库内设有区域数据处理单元、区域数据计算单元、区域数据存储单元、区域人员信息存储单元以及区域系统操作平台。
  3. 根据权利要求1所述的一种基于图像识别技术的工业场所不安全行为识别系统，其特征在于：所述云端数据库包括云端数据获取单元、云端数据计算单元、云端模型训练单元、识别模型存储单元、识别模型下放单元、云端数据存储单元以及云端操作平台。
  4. 根据权利要求2所述的一种基于图像识别技术的工业场所不 安全行为识别系统，其特征在于：所述区域数据库内还设有区域地区3D可视化模块，所述区域地区3D可视化模块与区域系统操作平台连接，且区域地区3D可视化模块内设有对应区域的建筑3D模型图，所述区域系统操作平台内设有语音信息编辑模块和语音通报模块。
  5. 根据权利要求1所述的一种基于图像识别技术的工业场所不安全行为识别系统，其特征在于：还包括有供电层，所述供电层包括用于提供设备电源的设备电源模块、用于清洁能源发电的太阳能发电模块以及用于充电的充电模块。
  6. 根据权利要求2所述的一种基于图像识别技术的工业场所不安全行为识别系统，其特征在于：所述区域系统操作平台内还设有定时模块和密码解锁模块，所述定时模块的输出端与密码解锁模块的输入端连接。
  7. 根据权利要求1所述的一种基于图像识别技术的工业场所不安全行为识别系统，其特征在于：所述区域数据库内还设有日志记录模块和日志上传模块，所述日志记录模块的输出端与日志上传模块的输入端连接，所述日志上传模块的输出端与云端数据库的输入端连接。
  8. 根据权利要求1所述的一种基于图像识别技术的工业场所不安全行为识别系统，其特征在于：所述信息传达模块内还设有警报单元，所述警报单元的输出端通过无线连接单元与区域数据库连接。
  9. 根据权利要求1所述的一种基于图像识别技术的工业场所不安全行为识别系统，其特征在于：所述控制终端为手机或平板电脑中的一种。
  10. 根据权利要求8所述的一种基于图像识别技术的工业场所不安全行为识别系统，其特征在于：所述信息传达模块内设有压力监测单元，所述压力监测单元的输出端与警报单元的输入端连接。
PCT/CN2023/086167 2022-12-16 2023-04-04 一种基于图像识别技术的工业场所不安全行为识别系统 WO2024124750A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
ZA2023/09548A ZA202309548B (en) 2022-12-16 2023-10-12 A system for recognizing unsafe behaviors in industrial workplaces based on image recognition technology

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202211624448.1 2022-12-16
CN202211624448.1A CN116311021A (zh) 2022-12-16 2022-12-16 一种基于图像识别技术的工业场所不安全行为识别系统

Publications (1)

Publication Number Publication Date
WO2024124750A1 true WO2024124750A1 (zh) 2024-06-20

Family

ID=86783997

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2023/086167 WO2024124750A1 (zh) 2022-12-16 2023-04-04 一种基于图像识别技术的工业场所不安全行为识别系统

Country Status (3)

Country Link
CN (1) CN116311021A (zh)
WO (1) WO2024124750A1 (zh)
ZA (1) ZA202309548B (zh)

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2005182652A (ja) * 2003-12-22 2005-07-07 Chugoku Electric Power Co Inc:The 遠隔監視方法及び遠隔監視システム
CN103093528A (zh) * 2012-12-28 2013-05-08 中山大学 一种小区车辆和人员智能监控系统
CN113298444A (zh) * 2021-07-15 2021-08-24 若普自动化技术（北京）有限公司 基于私有云的电力安全智能管控平台
CN114067472A (zh) * 2021-11-29 2022-02-18 广东电网有限责任公司 一种变电站进站授权管理系统及方法

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2005182652A (ja) * 2003-12-22 2005-07-07 Chugoku Electric Power Co Inc:The 遠隔監視方法及び遠隔監視システム
CN103093528A (zh) * 2012-12-28 2013-05-08 中山大学 一种小区车辆和人员智能监控系统
CN113298444A (zh) * 2021-07-15 2021-08-24 若普自动化技术（北京）有限公司 基于私有云的电力安全智能管控平台
CN114067472A (zh) * 2021-11-29 2022-02-18 广东电网有限责任公司 一种变电站进站授权管理系统及方法

Also Published As

Publication number Publication date
ZA202309548B (en) 2024-05-30
CN116311021A (zh) 2023-06-23

Similar Documents

Publication Publication Date Title
CN109767121A (zh) 一种变电站远方智能作业管控系统及方法
CN209543518U (zh) 一种对河道非法采砂行为进行自动监控的装置
CN205990117U (zh) 智能安防电梯控制系统
CN109149765A (zh) 一种分布式能源站无人值守方法
CN101458502A (zh) 电力系统标准化作业智能控制方法及专用装置
CN112869278A (zh) 一种用于工地的安全帽检测系统
CN111416831A (zh) 一种用于海上风电场的风机平台作业监管系统
CN110007170A (zh) 充电桩故障自动处理系统
WO2024124750A1 (zh) 一种基于图像识别技术的工业场所不安全行为识别系统
CN115169777A (zh) 一种安全作业智能监管系统
CN110246062A (zh) 一种物业管理监控管理系统
CN113850459A (zh) 一种火电厂作业票执行全过程管理系统
CN106532690A (zh) 一种变电站二次设备状态在线校核及智能诊断方法
CN111510685A (zh) 一种基于三维模型的信息输出方法及装置
CN104269007A (zh) 电力杆塔电子标识牌
CN112508384A (zh) 一种电力调度操作票可视化自动监控设备及方法
CN111754372A (zh) 一种基于物联网的水电站水淹厂房事故感知决策指挥系统
CN115953278A (zh) 一种基于bim的市政工程智慧工地管理系统
CN207352670U (zh) 一种施工人员安全布控系统
CN114205562B (zh) 智能终端及其工作方法
CN115641608A (zh) 基于边缘智能的配网作业个人防护用品智能识别设备
CN115396751A (zh) 具备医疗设备状态智能检测的机柜安防系统及控制方法
CN107545610A (zh) 一种施工人员安全布控系统
CN115022598A (zh) 水电安全人工智能视频监控系统
CN210201968U (zh) 一种智慧工地oa项目管理系统