31 results about "Safety behaviour" patented technology

The invention discloses a method for identifying unsafe behaviors of coal mine personnel based on human posture estimation. Aiming at the limitation of human-centered video surveillance mode for monitoring the unsafe behavior of underground personnel, Simultaneous monitoring of multiple scenarios is difficult, the result of manual monitoring can not be handled in time, Intelligent analysis technology is introduced into the coal mine video surveillance system, and the attitude information of coal mine personnel is extracted by Hourglass Networks with Hard Mining (HMN) based on the difficult samples of generated antagonism training. Then, according to the extracted human posture information in the video surveillance trajectory to determine whether the behavior of coal mine personnel abnormal, precisely detected and issued an alarm prompt, to prevent trouble in advance, to ensure coal mine safety in production.

The invention discloses a dangerous chemical safety production risk monitoring and early warning system and method. The system comprises a dangerous chemical safety production risk monitoring and early warning system. The method comprises the following steps: extracting and summarizing monitoring information; collecting and transmitting monitoring information; analyzing and early warning; and performing violation / abnormity alarm processing. Real-time online monitoring and early warning are realized on the basis of the Internet of Things technology, Internet of Things monitoring data are transmitted to the cloud for centralized processing, hidden dangers can be found and eliminated in time, major risks are prevented, and major and extra-major accidents are restrained; through intelligent video analysis, unsafe behaviors of people and unsafe states of objects are automatically detected; and meanwhile, interconnection and intercommunication with other information systems are realized.

The invention discloses a global supervision system for equipment and workers. The system comprises acquisition equipment and a worker supervision cloud platform. The acquisition equipment is connected with the worker supervision cloud platform through the NB-IoT communication module; the worker supervision cloud platform comprises an attendance module, an equipment operation and maintenance management module, a regional terminal and a comprehensive supervision platform, and the acquisition equipment comprises an intelligent safety helmet, a sensor module, a positioning module I, an identity recognition module, a processing module, a help seeking module, an information acquisition module and a power supply module; the attendance module comprises a positioning module II, a region module, atime recording unit and an information generation unit which are connected in sequence; the system has the advantages that real-time body movement information and real-time health sign information areobtained, real-time working condition information of equipment is obtained through the equipment operation and maintenance management module, the monitoring center analyzes and judges whether body movements of constructors are unsafe behaviors or non-standard operations or not, and early warning or video recording is conducted.

The invention discloses a method for evaluating the construction safety of a construction project, and the method comprises the steps: determining a personnel scoring index which comprises a pluralityof working parameters of personnel; confirming an equipment scoring index, wherein the equipment scoring index comprises a plurality of physical parameters of equipment; confirming personnel behaviorscoring indexes including a plurality of unsafe behaviors of personnel, the unsafe behaviors being obtained by substituting unsafe behavior data into the planned behavior theoretical model; carryingout multi-layer model construction on the personnel scoring indexes, the equipment scoring indexes and the personnel behavior scoring indexes; obtaining and substituting actual data into the multi-layer model, carrying out Logistic regression analysis on the result of the actual data, and obtaining the weight of each working parameter in the personnel scoring index, the weight of each physical parameter in the equipment scoring index and the weight of each unsafe behavior in the personnel behavior scoring index. According to the evaluation method for the construction safety of the constructionengineering project, the evaluation mode is reasonable, the influence of human factors is reduced, and evaluation standards are unified.

The invention belongs to the technical field of computer vision, and particularly relates to a method for identifying unsafe behaviors of power plant workers. The method specifically comprises the steps of presetting a plurality of working personnel unsafe behaviors, obtaining a field video image, analyzing the video image through a convolutional neural network of an improved YOLOv4 algorithm, and identifying the unsafe behaviors, including unsafe behavior analysis of a single-frame image, unsafe behavior analysis of a video sequence, and unsafe behavior analysis of a specific task and a scene. The method has the beneficial effects that the method is based on a computer vision technology, and unsafe behaviors are recognized in combination with unsafe behavior semantic information. Compared with an existing scheme, the method has the advantages of low price, high efficiency, real-time detection and no dead angle leakage, and a large amount of manpower, material resources and financial resources are saved. The method can be popularized and applied to unsafe behavior identification scenes in all industrial fields.

The invention discloses an escalator intelligent monitoring and management system based on computer vision. The escalator intelligent monitoring and management system comprises a plurality of fixed vision sensing devices, a plurality of infrared vision sensors, a tracking type vision sensing device, a photographic device, a 5G communication device and a terminal processor. According to the escalator intelligent monitoring and management system, an escalator intelligent monitoring system is connected through a 5G communication device, environmental parameters and personnel unsafe behaviors of an escalator are monitored in real time through the fixed visual sensing devices, the tracking type visual sensing device and the photographic device which are arranged near the escalator, so that linkage management of the monitoring system is realized, and the intelligent level of the system is improved; the early warning level of the escalator is determined by collecting environment information data in an escalator area in real time and recognizing unsafe behaviors of personnel in real time, and when the early warning level of the escalator is an abnormal level, an alarm module, an early warning prompt module and a safety early warning module can send out alarm signals, so that accurate early warning of abnormal operation of the escalator is achieved, and safety accidents of the escalator can be effectively prevented.

The invention discloses a method for automatically detecting unsafe behaviors of workers based on machine vision. The method comprises the steps that S10, acquiring original images of the unsafe behaviors of workers; S20, manufacturing label images of the unsafe behaviors; S30, constructing a target detection model, and then training and verifying the model; and S40, obtaining a real-time monitoring video, inputting the obtained image frame into the target detection model for detection, automatically identifying unsafe behaviors of workers, and receiving an identification result through monitoring software. According to the method for automatically identifying the unsafe behaviors of the workers, the principle is that the unsafe behavior database of the workers is established, the unsafe actions of the workers are automatically identified through a deep learning algorithm, results are displayed, and therefore the occurrence rate of the unsafe behaviors of the workers is reduced. Compared with a traditional monitoring method, the method has the advantages of being automatic, good in monitoring effect, easy to operate and the like.

The invention provides a multi-scene sailor unsafe behavior detection method based on improved YOLOv3. Water traffic accidents directly or indirectly caused by human factors account for about 80% of the total number of the water traffic accidents. According to the invention, the improved YOLOv3 network is utilized to detect six types of unsafe behaviors of sailors in different scenes, such as off-duty, smoking, mobile phone playing, conversation, sleeping and lookout negligence, and the detected unsafe behaviors are calibrated and reminded. The method comprises the following steps: firstly, adding a CLAHE algorithm into a data enhancement part; secondly, changing the original three-scale detection into four-scale detection, and introducing a negative feedback mechanism into a feature fusion network part, so the features of all scales are better fused; and finally, improving a non-maximum suppression algorithm to carry out Gaussian weighting on a candidate frame to obtain a more accurate result. According to the invention, unsafe behaviors of the sailors can be accurately detected and the sailors can be reminded, so the occurrence rate of marine accidents caused by human factors is reduced and the safe navigation capability of a ship is improved.

The invention relates to the technical field of coal mine safety production, and discloses a coal mine underground personnel unsafe behavior identification method comprising the following steps: 1, carrying out alcohol detection on personnel entering a well; step 2, checking articles carried by the personnel entering the well and clothes worn by the personnel entering the well; the method comprises the following steps of 1, acquiring a plurality of frames of continuous images of a plurality of preset unsafe behaviors through personnel simulation, 2, detecting image information in the step 3, and extracting posture characteristics of personnel in the images, and 3, constructing and extracting undirected skeleton behavior characteristics capable of describing human body behaviors, step 6, analyzing and classifying the behavior characteristics in the step 5, and step 7, shooting and comparing the behavior images of the underground personnel, and performing background alarm according to the comparison result. Through the method, the unsafe behaviors of the underground personnel can be identified in real time, the identification effect is good, the requirements of background monitoring personnel are reduced, and the safety of coal mine production is effectively improved.

The invention provides a mine team worker unsafe behavior propagation process extraction and analysis method. The method comprises a technology for achieving coal mine underground work group worker unsafe behavior group emergence propagation warning based on an Agent modeling and simulation method (AMBS), and comprises the following steps: 1, establishing an information processing process model ofworker unsafe behavior emergence propagation; 2, extracting and depicting system elements such as a main body, a structure and an internal communication connection mode of a team group system, and constructing an unsafe behavior propagation process concept model; 3, designing a worker main body Agent model; 4, designing a worker unsafe behavior propagation decision mechanism based on a cognitiveprocess; and 5, carrying out simulation realization. On the basis of fully depicting system elements in the team operation process, the dynamic propagation evolution process of the unsafe behaviors ofthe workers is dynamically presented, a warning effect is achieved on the safety production process of the workers, and a basis is provided for safety management decision making.

The invention relates to the technical field of behavior recognition, and particularly discloses an intelligent recognition method for unsafe behaviors of an oil and gas laboratory, and the method comprises the steps: constructing a face recognition platform through a Retinaface network and an improved Facenet network, intercepting a face part through the Retinaface network, transmitting the face part to the improved Facenet network, comparing the face part with an existing face database, and judging the personnel entering and exiting the laboratory. The Facenet network is improved by replacing a backbone extraction network with a MoilentV1 network and using Triplet Loss and Cross-Entry Loss as loss functions, the accuracy of the network is reduced by 1.55%, and the FPS (frame number per second) is improved by 25. An improved YOLOv4-tiny network is adopted to identify unsafe behaviors of protective wearing, unsafe behaviors of dangerous area invasion, unsafe behaviors of experimenters and unsafe behaviors of illegal use of equipment, a K-means clustering method is used in a first part to select a proper prior box, an ECA attention module is added in a second part to carry out improvement, and the safety of the equipment is improved. Therefore, the feature extraction capability of the network on small targets is improved. Compared with an original YOLOv4-tiny network, the average precision mean value is improved by 21.17%.

The invention discloses a staff safety behavior capability assessment system and method based on a VR technology. The beneficial effects of the invention are that the method is implemented; Accidentsin real operation are achieved through virtual reality, four first-level safety behavior capability assessment indexes are selected by staff in the operation process or the accident handling process,the four first-level safety behavior capability assessment indexes comprise safety physiology, safety psychology, safety knowledge and safety emergency, and the overall safety behavior capability of the staff is further assessed through the four indexes.

The invention belongs to the technical field of risk assessment systems, and particularly relates to a building construction quality safety online risk assessment system which comprises an engineeringinformation module, a risk assessment module, a risk grade assessment module, a dynamic assessment module and an output module, wherein the engineering information module is used for surveying geographical, hydrological and meteorological conditions of an evaluation object on site and collecting construction-related data of engineering; the risk assessment module comprises overall risk assessmentand special risk assessment, and carries out comprehensive risk analysis on the overall risk assessment and the special risk assessment; and the risk grade assessment module applies an LEC evaluationmethod according to the comprehensive risk analysis result to analyze unsafe behaviors of people and unsafe states of objects which may cause accidents in the engineering construction process from the two aspects of generality and personality respectively for one-by-one evaluation. The building construction quality safety online risk assessment system defines the types and harm degrees of variousdangerous sources, provides feasible safety countermeasures from the aspects of safety technology, organization and management and the like, and achieves safety production.

The invention discloses an intelligent work ticket and operation ticket system, which comprises an on-site investigation module for realizing on-site investigation risk output and safety measure drilling; the flow pushing module is used for realizing the processes of starting, signing, issuing, receiving, permitting, environment change early warning, interruption, change, delay and termination in the process of operation order starting, signing, issuing, receiving and permitting; the safety measure implementation and recovery early warning module is used for implementing safety measures and safety early warning; the operating personnel risk identification early warning module automatically matches and identifies personnel safety credit information; the tool and tool management and control module is used for managing and controlling tools and tools; the personal intelligent equipment early warning module is used for accessing related equipment warning signals to the intelligent work ticket and operation ticket system; the video monitoring and early warning module identifies unsafe behaviors of personnel and unsafe states of objects in the operation process and uploads related warning signals to the intelligent work ticket and operation ticket system; and the intrinsic safety capability and the working efficiency of field operation are improved.

By means of the mode, a high-definition camera can be used for recognizing the unsafe behaviors of people in engineering construction and the unsafe state of a construction site through training, andthe corresponding relation between the behaviors / states and the recognition algorithm is established. Remote supervision of a construction site is realized, the real-time performance of supervision isimproved, the sudden occurrence of safety accidents is reduced, and the adverse consequences caused by the sudden occurrence of the safety accidents are reduced. Generally speaking, the work of supervisors is assisted and optimized on the basis of a machine and an intelligent algorithm, so that the labor intensity of the supervisors is reduced, the working efficiency of the supervisors is improved, and the emergency performance of prevention and treatment of engineering constructors is greatly improved.

The invention discloses an omnibearing monitoring, recording and correcting system for unsafe behaviors of building construction. The omnibearing monitoring, recording and correcting system comprisesa control terminal, a first monitoring terminal and a second monitoring terminal, wherein the control terminal is convenient for supervisors to observe staff and site conditions; the first monitoringterminal wirelessly interacts with the control terminal and is used for detecting a human body state, and the second monitoring terminal wirelessly interacts with the control terminal and is used fordetecting a site construction state; the first monitoring terminal comprises a heart rate detection part worn on a worker and used for detecting the heart rate of the worker, and a safety helmet wornon the head of the worker. The safety helmet comprises a helmet body and an isolation plate arranged at the top in the helmet body. The device is designed for overcoming the defects of an existing device, staff can be positioned, the physical condition of the staff can be detected, whether the staff wear the device or not can be detected, the bending state of workers can be detected, and thereforethe workers can be prompted and corrected in a targeted mode, the construction quality is guaranteed, and practicability is high.

Systems and methods are disclosed for spatially and temporally mapping safer and riskier and wellbeing areas based on perceived safety / risk and wellbeing, related to perceived viral mitigating health behaviors, known built environment wellbeing features, and predictively modelling movement through locations, and building environment features and operations contributing to viral spread.

The invention discloses an intelligent video analysis system and method. The system comprises a scene analysis system and a linkage comprehensive analysis early warning system, and the scene analysis system is used for judging whether unsafe behaviors exist in different operation scenes and transmitting scene data to the linkage comprehensive analysis early warning system; and the linkage comprehensive analysis early warning system is used for analyzing scene data and carrying out early warning on the condition that unsafe behaviors exist in the operation scene. According to the method, the problems that in a complex monitoring scene with an indefinite detection target and an unfixed detection range, the video detection accuracy is low, positioning cannot be used due to multiple positioning blind areas, the hardware resource consumption is great, the intelligent degree is low and the like in the prior art are solved.

The invention discloses a subway passenger unsafe behavior risk assessment system. The system comprises a behavior data acquisition module, a behavior state display module, a behavior identification module and a behavior risk early warning module, wherein the behavior data acquisition module comprises Kinect, and the behavior data acquisition module can acquire skeleton data and related depth information of unsafe behaviors of passengers through the Kinect to obtain passenger position data. The subway passenger unsafe behavior risk assessment system is provided with four functional modules. The method has the beneficial effects that the convenience and the accuracy of judging the unsafe behaviors of passengers by subway management personnel are improved, and the risk caused by the behaviors is used as an early warning index, so that decision-making personnel can accurately estimate the risk situation and quickly make emergency response.

The invention provides an automatic identification method and device for unsafe behaviors of subway construction operators, relates to the technical field of subway construction operation, and aims to effectively reduce the occurrence frequency of the unsafe behaviors of the operators and improve the safety guarantee of the operators in a construction site. The method comprises the following steps: acquiring a first type of unsafe behaviors of subway construction workers based on a detection sensor; acquiring a second type of unsafe behaviors of subway construction workers based on an artificial intelligence image processing method; based on the first type of unsafe behaviors and the second type of unsafe behaviors, constructing a subway construction worker unsafe behavior evaluation model, and training the evaluation model to form a trained identification model; and automatically identifying a new unsafe behavior based on the trained identification model and carrying out early warning. The device for automatically identifying the unsafe behaviors of the subway construction workers is applied to the method for automatically identifying the unsafe behaviors of the subway construction workers.

The invention discloses a detection method and device for identifying a personal safety behavior portrait, and the method comprises the steps: carrying out the point set extraction of a dangerous source region in a video image, carrying out the skeleton point set extraction of a person in the video image, and judging whether the person is in the dangerous source region or not; if the person is in the dangerous source area, judging whether the person has a smoking behavior or not; if the person has the smoking behavior, calculating the gradient of a certain pixel point of smoke in the obtained video image of the mouth of the person, and constructing a histogram in the gradient direction based on the gradient of the certain pixel point; and comparing the constructed histogram in the gradient direction with a histogram in a preset smoking smoke histogram database. The smoking smoke histogram database is established based on big data, the obtained histogram of the mouth video image of the person is compared with the smoking smoke histogram in the database, whether the person smokes in the dangerous area or not is determined, the behavior of the person in the dangerous area is detected, and the detection task load can be reduced.

The invention relates to the technical field of segmental beams, in particular to a process for quickly disassembling and assembling an integrated segmental beam steering block formwork, which comprises the following steps of: installing a bottom corner outer formwork, installing a top end outer formwork, then installing a movable inner formwork in a space formed by the bottom corner outer formwork and the top end outer formwork, and filling the space with a steel reinforcement framework. The movable inner formwork is fed into the outer formwork cavity through the C-shaped inner formwork, the movable inner formwork is fixed into the outer formwork cavity through the hydraulic system, then the C-shaped inner formwork retreats, and finally the outer side plate is installed. According to the process, a segmental beam can be integrally formed, and modular splicing is not required. The unsafe state of unsafe behaviors of people is completely eradicated, potential safety hazards are eliminated, the effect is high, the mold closing time is shortened, the demolding effect of the precast beam section is ideal, and the requirement of the construction technology is met.

The invention provides a ship deck cable database construction method and system, and the method comprises the steps: obtaining video data of unsafe behaviors; preprocessing the video data; screening the preprocessed data; carrying out labeling on the screened images by adopting the labmeli; and storing the basic information of the marked unsafe behavior database image in a picture data table to form a deck cable unsafe behavior database. An image database of unsafe behaviors of sailors is collected, unsafe behavior characteristics of each database picture are extracted, and the database picture characteristics and unsafe behavior indexes are stored in a picture characteristic table; and unsafe behaviors of the sailors operating the deck cable are monitored through algorithms such as deep learning and a neural network, and the sailors can obtain early warning through bracelet vibration on the basis of support of the project of a sailors unsafe behavior and state monitoring and early warning platform.

The invention discloses a method for recognizing unsafe behaviors of coal mine workers based on human body postures. The method comprises the following steps: firstly, presetting a plurality of unsafebehaviors of coal mine workers, acquiring video information of the unsafe behaviors, determining positions and categories of human body joints in the video information through a human body detector,and calculating affinity vector fields of each group of adjacent human body joints in the unsafe behaviors to serve as threshold values; secondly, acquiring a working image of a coal mine worker through a camera, and predicting the position and category of a human body joint in the working image by using a regression method; and finally, calculating an affinity vector field of adjacent human bodyjoints in the working image, comparing the affinity vector field with the threshold value, judging that the worker behavior violates rules if the affinity vector field and the threshold value are within a certain difference rang, and giving an alarm; and if the difference is not within the range, judging the behavior to be not illegal. An alarm can be given on the three-violation behavior of coalmine workers, and the occurrence of the three-violation behavior can be reduced.

The invention discloses a gas cylinder intelligent safety training system and method based on a sensing technology, and belongs to the field of enterprise safety training. The system comprises a personnel information acquisition and registration module which inputs the personal information of students to support the operation of other modules and satisfy the establishment of employee safety training files by enterprises; a practical operation data collection module which is used for collecting practical operation data of students through various sensors; a data analysis and comparison module which is used for comparing and analyzing the data collected by the practical operation data collection module with the national standard; and a video teaching module which is used for checking personal scores and score deduction items of the students and automatically generating corresponding learning contents according to the options fed back by the data analysis and comparison module. By the adoption of the intelligent safety training system, bad habits of safety behaviors of operators are automatically collected and recorded, targeted video teaching is conducted, and the effects of process-based and integrated behavior evaluation and safety training can be achieved.

The invention discloses an intelligent monitoring and management system for an escalator. The intelligent monitoring and management system comprises a plurality of fixed camera devices, a plurality of infrared sensing devices, a movable visual sensor, a photographing device, a wireless communication device and a central processing unit. According to the system, the escalator intelligent monitoring system is connected through the wireless communication device, and environmental parameters and personnel unsafe behaviors of the escalator are monitored in real time through the fixed camera devices, the movable visual sensor and the photographing device which are arranged near the escalator, so that linkage management of the monitoring system is realized, the intelligent level of the system is improved, the early warning level of the escalator is determined by collecting environment information data in an escalator area in real time and recognizing unsafe behaviors of people in real time, and when the early warning level of the escalator is an abnormal level, an alarm module and an unsafe state early warning module can send out alarm signals, therefore, accurate early warning of abnormal operation of the escalator is achieved, and safety accidents of the escalator can be effectively prevented.