CN113822197A - Work dressing identification method and device, electronic equipment and storage medium - Google Patents
Work dressing identification method and device, electronic equipment and storage medium Download PDFInfo
- Publication number
- CN113822197A CN113822197A CN202111114346.0A CN202111114346A CN113822197A CN 113822197 A CN113822197 A CN 113822197A CN 202111114346 A CN202111114346 A CN 202111114346A CN 113822197 A CN113822197 A CN 113822197A
- Authority
- CN
- China
- Prior art keywords
- safety
- uniform
- loss function
- human body
- module
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
Abstract
The invention discloses a method and a device for identifying working dressing, electronic equipment and a storage medium, wherein the method comprises the following steps: s1, obtaining video frame data of an electric power operation site, S2, recognizing a human body according to the video frame data to obtain human body parameters, S3, performing uniform recognition comprehensive processing based on the human body parameters to obtain comprehensive loss function values of safety clothes, S4, performing safety belt recognition based on the human body parameters to obtain corresponding parameters and loss function values of safety belts, S5, analyzing the comprehensive loss function values of the safety clothes and the corresponding parameters and loss function values of the safety belts to obtain safety state data, S6, and determining whether the clothes of an operator are in compliance or not based on the safety state data. The invention provides a working dressing identification method, which manages and controls the normative dressing of working clothes and safety belts of personnel in the working process to the maximum extent and improves the normative and safety of construction operation.
Description
Technical Field
The present invention relates to the technical field of work dressing security, and in particular, to a work dressing identification method and apparatus, an electronic device, and a storage medium.
Background
With the rapid development of social economy and the continuous improvement of the technological level, the electric power industry of China is rapidly developed in recent years.
In the electric power production process, dress to the operation personnel and wear and have strict requirement, need real-time supervision field operation personnel when the operation of ascending a height, do not have and wear suitable work clothes, wear the safety belt, moreover in whole course of the work: the sleeves and the trouser legs cannot be rolled up, the working clothes are kept intact, and safety belts are used in compliance during the climbing operation process. In the operation process, if the working clothes and the safety belts of the field operation personnel are not in compliance, an alarm is sent out in time, and the images of the personnel are recorded. The field operation of electric power production is mostly carried out in the field, the field illumination condition changes irregularly, the working environment is very different, the working position changes constantly, and the human posture changes constantly, which puts forward severe requirements on the detection and identification algorithms of the wearing compliance of the working clothes and the safety belts of personnel.
Therefore, in order to solve the technical problem of potential safety hazard caused by low safety identification rate of the existing dressing identification method in the safety operation of workers, it is urgently needed to construct a working dressing identification method.
Disclosure of Invention
The invention provides a working dressing identification method and device, electronic equipment and a storage medium, and solves the technical problem of potential safety hazard caused by low safety identification rate of the existing dressing identification method.
In a first aspect, the present invention provides a method for identifying a working dressing, comprising:
s1, acquiring video frame data of the electric power operation site;
s2, identifying the human body according to the video frame data to obtain human body parameters;
s3, performing uniform identification comprehensive treatment based on the human body parameters to obtain a safety uniform comprehensive loss function value;
s4, carrying out safety belt identification based on the human body parameters to obtain corresponding parameters of the safety belt and a loss function value;
s5, analyzing the safety suit comprehensive loss function value, the safety belt corresponding parameter and the loss function value to obtain safety state data;
and S6, determining whether the dress of the operator is in compliance or not based on the safety state data.
Optionally, the step S2 includes:
s21, converting the format of the video frame data to obtain picture numerical data;
s22, inputting the picture numerical data into a preset network neural model to obtain human body parameters;
s23, judging whether the human body parameters are correct or not; if yes, go to step S3; if not, the process returns to step S21.
Optionally, the step S3 includes:
s31, dividing the uniform categories based on the human body parameters to obtain uniform category classification results and uniform category loss function values;
s32, performing color classification calculation on the uniform to obtain a uniform color classification result and a uniform color loss function value;
and S33, performing comprehensive calculation of the safety uniform based on the uniform classification result, the uniform classification loss function value, the uniform color classification result and the uniform color loss function value to obtain a numerical value of the comprehensive loss function of the safety uniform.
Optionally, the step S4 includes:
s41, carrying out safety belt graph segmentation and extraction on the human body parameters to obtain safety belt picture data;
and S42, inputting the safety belt picture data into a preset safety belt identification model to obtain corresponding parameters and loss function values of the safety belt.
In a second aspect, the present invention provides a working dressing identification apparatus comprising:
the acquisition module is used for acquiring video frame data of an electric power operation site;
the parameter module is used for identifying a human body according to the video frame data to obtain human body parameters;
the safety clothing module is used for carrying out uniform identification comprehensive treatment based on the human body parameters to obtain a safety clothing comprehensive loss function value;
the safety belt module is used for carrying out safety belt identification based on the human body parameters to obtain corresponding parameters and loss function values of the safety belt;
the analysis module is used for analyzing the safety suit comprehensive loss function value, the corresponding parameters of the safety belt and the loss function value to obtain safety state data;
and the safety module is used for determining whether the dressing of the operator is in compliance or not based on the safety state data.
Optionally, the parameter module comprises:
the conversion submodule is used for carrying out format conversion on the video frame data to obtain picture numerical data;
the parameter submodule is used for inputting the picture numerical data into a preset network neural model to obtain human body parameters;
the judgment submodule is used for judging whether the human body parameters are correct or not; if yes, executing the step of a safety service module; if not, returning to the step of executing the transformation module.
Optionally, the safety garment module comprises:
the category submodule is used for carrying out uniform category division based on the human body parameters to obtain uniform category classification results and uniform category loss function values;
the color submodule is used for carrying out color classification calculation on the uniform to obtain a uniform color classification result and a uniform color loss function value;
and the comprehensive submodule is used for carrying out comprehensive calculation on the safety uniform based on the uniform classification result, the uniform classification loss function value, the uniform color classification result and the uniform color loss function value to obtain a numerical value of the comprehensive loss function of the safety uniform.
Optionally, the seat belt module comprises:
the extraction submodule is used for carrying out safety belt graph segmentation extraction on the human body parameters to obtain safety belt picture data;
and the input submodule is used for inputting the safety belt picture data into a preset safety belt identification model to obtain corresponding parameters and loss function values of the safety belt.
In a third aspect, the present application provides an electronic device comprising a processor and a memory, wherein the memory stores computer readable instructions, and the computer readable instructions, when executed by the processor, perform the steps of the method as provided in the first aspect.
In a fourth aspect, the present application provides a storage medium having stored thereon a computer program which, when executed by a processor, performs the steps of the method as provided in the first aspect above.
According to the technical scheme, the invention has the following advantages: the invention provides a working dressing identification method, which comprises the steps of obtaining video frame data of an electric power working site through S1, S2, identifying a human body according to the video frame data to obtain human body parameters, S3, carrying out uniform identification comprehensive treatment based on the human body parameters to obtain a safety uniform comprehensive loss function value, S4, carrying out safety belt identification based on the human body parameters to obtain corresponding parameters and loss function values of a safety belt, S5, analyzing the safety uniform comprehensive loss function value and the corresponding parameters and loss function values of the safety belt to obtain safety state data, S6, determining whether the dressing of an operator is in compliance or not based on the safety state data, and solving the technical problem of potential safety hazards caused by low safety identification rate of the existing dressing identification method through a working dressing identification method, personnel's work clothes and safety belt normative dress in the at utmost management and control operation process improve the normative and the security of construction operation.
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, and 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 these drawings without inventive exercise.
FIG. 1 is a flowchart illustrating a first embodiment of a method for identifying a working garment according to the present invention;
FIG. 2 is a flowchart illustrating a second embodiment of a method for identifying a working garment of the present invention;
FIG. 3 is a block diagram of a power job site personnel dressing safety identification system according to the present invention;
FIG. 4 is a block diagram of a judgment system in the dressing security comprehensive analysis method according to the present invention;
FIG. 5 is a flowchart illustrating a method for calculating a loss function value for a safety apparel category in a method for identifying a work apparel according to the present invention;
FIG. 6 is a flowchart illustrating the steps of a method for calculating a color loss function value of a safety garment in a method for identifying a work garment according to the present invention;
FIG. 7 is a flowchart illustrating the steps of a method for calculating a belt loss function value in a method for identifying a work garment of the present invention;
fig. 8 is a block diagram illustrating an embodiment of a working clothing recognition apparatus according to the present invention.
Detailed Description
The embodiment of the invention provides a working dressing identification method, a working dressing identification device, electronic equipment and a storage medium, which are used for solving the technical problem of potential safety hazard caused by lower safety identification rate of the existing dressing identification method.
In order to make the objects, features and advantages of the present invention more obvious and understandable, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention, and it is obvious that the embodiments described below are only a part of the embodiments of the present invention, and not all of the embodiments. 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.
In a first embodiment, referring to fig. 1, fig. 1 is a flowchart illustrating a first working clothing identification method according to a first embodiment of the present invention, including:
step S101, video frame data of an electric power operation site are obtained;
step S102, identifying a human body according to the video frame data to obtain human body parameters;
step S103, performing uniform identification comprehensive treatment based on the human body parameters to obtain a safety uniform comprehensive loss function value;
step S104, based on the human body parameters, carrying out safety belt identification to obtain corresponding parameters and loss function values of the safety belts;
step S105, analyzing the comprehensive loss function value of the safety suit, and the corresponding parameters and loss function values of the safety belt to obtain safety state data;
it should be noted that the safety state data includes a code for identifying the state, alarm information, and a human body figure, and is used to ensure the dressing safety of the operator.
And step S106, determining whether the dress of the operator is in compliance or not based on the safety state data.
In the working dressing identification method provided by the embodiment of the invention, video frame data of an electric power working site is obtained through S101, S102, a human body is identified according to the video frame data to obtain human body parameters, S103, uniform identification comprehensive processing is carried out based on the human body parameters to obtain a safety uniform comprehensive loss function value, S104, safety belt identification is carried out based on the human body parameters to obtain parameters and a loss function value corresponding to the safety belt, S105, the safety uniform comprehensive loss function value and the parameters and the loss function value corresponding to the safety belt are analyzed to obtain safety state data, S106, whether dressing of an operator is in compliance or not is determined based on the safety state data, and the technical problem of potential safety hazards caused by low safety identification rate of the existing dressing identification method is solved through the working dressing identification method, personnel's work clothes and safety belt normative dress in the at utmost management and control operation process improve the normative and the security of construction operation.
In a second embodiment, referring to fig. 2, fig. 2 is a flowchart illustrating a method for identifying a working wear of the present invention, including:
step S201, acquiring video frame data of an electric power operation site;
in an embodiment of the present invention, please refer to fig. 3, fig. 3 is a block diagram illustrating a structure of a personnel dressing safety recognition system of an electric power operation site according to the present invention, wherein 301 is an intelligent processing module, 302 is an image decoding module, 303 is a communication module, 304 is a power adapter module, and 305 is a removable storage module; the intelligent processing module 301 is respectively connected with the image decoding module 302, the communication module, the power supply adapting module and the removable storage module 305; the dressing safety identification system for the personnel on the electric power operation site can be used for an image identification edge computing terminal; the method comprises the steps of obtaining video stream data of the power operation site shot by a portable camera through calling a communication interface of an image identification edge computing terminal (namely, a power operation site personnel dressing safety identification system), and coding and compressing the video stream data through an image decoding module to convert the video stream data into video frame data.
In specific implementation, the dressing safety identification system for the personnel in the electric power operation field is applied to an image identification edge computing terminal, the image identification edge computing terminal and a portable camera are connected to a communication module through a network cable or WIFI (wireless fidelity) to form a local area network, and all devices in a monitoring field can communicate in the local area network without depending on other communication means.
The image decoding module 302 is configured to decode the video stream data of the power operation site captured by the portable camera into image frame data.
The intelligent processing module 301 is used for calculating a dressing identification algorithm of workers on the electric power operation site at a high speed and analyzing whether the image frame data is a worker wearing a garment or a worker wearing a safety belt which is not in compliance in real time.
And the communication module 303 is configured to transmit an analysis result of the intelligent processing chip to the electric power operation safety supervision platform. The communication module has a 4G/5G communication function and provides a wide area communication link for the image recognition edge computing terminal, the camera and other equipment to report data to the master station.
The removable storage module 304 is used for locally storing the analysis result of the intelligent processing chip.
The image recognition edge computing terminal can normally operate under the condition of lacking 4G communication, reported data can be temporarily stored in the local east Asia removable storage module, and reporting is carried out after the communication condition is allowed.
The intelligent processing module 301 can mainly perform the following functions: firstly, decoding the video stream data of the operation site into image frame data through an image decoding module and then inputting the image frame data into an intelligent processing chip, carrying out high-speed worker identification operation on the image frame data by the intelligent processing chip, and analyzing whether the image frame data is not in compliance when being worn by a worker in real time; secondly, the intelligent processing chip connects the operation recognition analysis result to the communication module through the signal interaction end and transmits the operation recognition analysis result to the background master station; and thirdly, the intelligent processing chip transmits the recognition and analysis result to the movable storage module through the signal output end for locally storing the recognition result.
The intelligent processing chip is connected with the display touch module through an interface of HDMI + USB, and the intelligent processing chip is connected with the audio output module through an interface I2S.
The image decoding module 302 has two interfaces of LAN and USB, supports two portable cameras at different angles to be connected to the operation site, and can simultaneously access at most two paths of video stream data of the power operation site for decoding.
The removable memory module 304 has an external slot of UHS-I interface type, and can support the insertion of an SD card with a maximum capacity of 128 GB.
The communication module 302 is provided with an external antenna, and the communication modes include WIFI and 4G.
The power adapter module 303 has an external jack of the PJ2.5 interface type.
The intelligent processing module 301 integrates a GPU and a CPU; NVIDIA (England) Jetson AGX Xavier cards or Hi35XX series multi-core heterogeneous processors of Haisi may be used.
Step S202, format conversion is carried out on the video frame data to obtain picture numerical data;
in the embodiment of the invention, the video frame data format is converted into the picture numerical data.
In a specific implementation, data in a BGR format of a read picture is converted into an RGB format through a color conversion function formula, the obtained BGR format is converted into a (C, H, W) format, and normalization processing is performed. Converting the reading into a Tensor format of which the RGB format is converted into a shape of (C, H, W), and dividing the numerical value by 255 to be normalized to a numerical value between [0,1.0] to obtain picture numerical value data.
Step S203, inputting the picture numerical data into a preset network neural model to obtain human body parameters;
in the embodiment of the invention, the converted picture numerical data is input into a preset network neural model for reasoning and identifying to obtain the human body parameters, and the human body parameters are output to the storage unit.
Step S204, judging whether the human body parameters are correct or not; if yes, go to step S205; if not, returning to execute the step S202;
in the embodiment of the invention, the identified human body parameters are judged, if the human body parameters are correct, the next step is executed, and if not, the format conversion step is executed again.
In the specific implementation, the identified human body parameters are judged, if the human body parameters are null, the human body parameters are reset, an N instruction is sent out, and the human body identification is restarted; if not, sending out a 'Y' command, and outputting the calculated human body parameters to the following steps through a change-over switch.
Step S205, on the basis of the human body parameters, performing uniform category division to obtain uniform category classification results and uniform category loss function values;
in the embodiment of the invention, the uniform categories are divided based on the human body parameters to obtain uniform category classification results and uniform category loss function values.
In a specific implementation, a pre-trained uniform class recognition model loaded in a program model is extracted. All pixels belonging to the uniform class are assigned the same color/value in the mask.
And converting the BGR format picture of the human body parameters read into RGB format through a color conversion function formula, and converting the obtained uniform RGB format data into picture data with a given size.
The obtained picture data of a given size is converted into a (C, H, W) format and normalized. The read-in scaled RGB format data is converted to a Tensor format shaped as (C, H, W) and the value divided by 255 is normalized to a value between 0, 1.0. Further according to the following calculation formula: the normalized calculation is performed for image (image-mean)/std.
And sending the picture data to a uniform type identification model for identification, sending the converted picture numerical data to the uniform type identification model for iterative optimization reasoning calculation, and finally calculating the minimum loss function value for uniform identification.
Step S206, color classification calculation is carried out on the uniform to obtain a uniform color classification result and a uniform color loss function value;
in the embodiment of the invention, the uniform is subjected to color classification calculation to obtain a uniform color classification result and a uniform color loss function value.
In a specific implementation, a pre-trained uniform color classification model loaded in a program model is extracted. Pixels belonging to different colors of the uniform are all assigned the same color value in the mask.
Converting the read human body parameters from BGR into HSV format pictures, and calculating the function of an image HSV (histogram).
Normalizing the histogram function of the picture, normalizing the obtained histogram function of the picture through the normalization function, and dividing the value by 255 to be normalized to a value between [0,1.0 ].
And sending the picture data to a uniform model for identification, sending the converted picture numerical value to a uniform color classification model for reasoning calculation, and identifying each parameter of uniform color.
And calculating a loss function, introducing the obtained parameters of the uniform colors into a uniform color classification model, performing iterative optimization reasoning calculation, and finally calculating the minimum loss function value for uniform color identification.
Step S207, performing comprehensive calculation on the safety uniform based on the uniform classification result, the uniform classification loss function value, the uniform color classification result and the uniform color loss function value to obtain a numerical value of a comprehensive safety uniform loss function;
in the embodiment of the invention, the safety suit is comprehensively calculated to obtain the safety suit comprehensive loss function value.
In the concrete implementation, firstly, a safety clothing classification loss function is formed and multiplied by a safety clothing weight coefficient, and then a clothing color classification loss function is formed and multiplied by a clothing color weight coefficient; from this, the integrated computation output value is (safety classification loss function × safety weighting factor) + (clothing color classification loss function × clothing color weighting factor).
Step S208, carrying out safety belt graph segmentation and extraction on the human body parameters to obtain safety belt picture data;
in the embodiment of the invention, human body parameters are read, safety belt graph segmentation and extraction are carried out, and the extracted picture data is subjected to scaling and normalization processing to obtain the safety belt picture data.
In the specific implementation, the BGR format data of the human body parameter picture is read in and converted into the RGB format through a color conversion function formula.
And converting the obtained safety belt RGB format data into picture data with a given size.
Converting the obtained picture data with given size into (C, H, W) format and normalizing, converting the read-in scaled RGB format data into Tensor format shaped as (C, H, W), and dividing the value by 255 to normalize the value between [0,1.0 ]. Further according to the following calculation formula: and (5) carrying out normalization calculation on the image-mean/std to obtain the safety belt picture data.
Step S209, inputting the safety belt picture data into a preset safety belt identification model to obtain corresponding parameters and loss function values of the safety belt;
in the embodiment of the invention, the picture data is transmitted to a safety belt model for identification to obtain corresponding parameters and loss function values of the safety belt.
In the specific implementation, the converted safety belt picture data is sent to a safety belt recognition model for reasoning calculation, and each parameter of a safety belt result is recognized and output to a storage unit.
And calculating a loss function, sending the converted safety belt picture data to a safety belt identification model network for iterative optimization reasoning calculation, and finally calculating the minimum safety belt identification loss function value.
Step S210, analyzing the safety suit comprehensive loss function value, the corresponding parameters of the safety belt and the loss function value to obtain safety state data;
in the embodiment of the invention, the comprehensive loss function value of the safety suit, the corresponding parameters of the safety belt and the loss function value are analyzed according to different working conditions required by field operation to obtain safety state data; the safety state data comprises codes for identifying the state, alarm information and human body figures and is used for ensuring the dressing safety of operators.
Step S211, determining whether the dress of the operator is in compliance based on the safety state data;
in the embodiment of the invention, the safety state data is used for ensuring the dressing safety of the operator.
In a specific implementation, please refer to fig. 4, fig. 4 is a block diagram of a determining system in a comprehensive dressing safety analysis method according to the present invention, which includes a 1 st reading module (51), a 2 nd reading module (52), a 1 st not logic module (53), a 2 nd not logic module (55), a 1 st parameter setting module (54), a 2 nd parameter setting module (56), a 3 rd parameter setting module (82), a 4 th parameter setting module (83), a 5 th parameter setting module (84), a 6 th parameter setting module (85), a 7 th parameter setting module (86), an 8 th parameter setting module (87), a 1 st high limit alarm module (58), a 2 nd high limit alarm module (59), a 1 st and logic module (60), a 2 nd and logic module (61), a 3 rd and logic module (62), a 4 th and logic module (63), A 5 th AND logic module (68), a 6 th AND logic module (69), a 7 th AND logic module (70), an 8 th AND logic module (71), a 9 th AND logic module (72), a 10 th AND logic module (73), a 1 st mode selector switch (65), a 2 nd mode selector switch (66), a 3 rd mode selector switch (67), a 1 st OR logic module (64), a 2 nd OR logic module (75), a 3 rd OR logic module (76), a 1 st selector module (76), a 2 nd selector module (77), a 3 rd selector module (78), a 4 th selector module (79), a 5 th selector module (80), a 6 th selector module (81), and the like.
The input of the 1 st reading module (51) is terminated "step S4: seat belt identification "output, read" step S4: the safety belt identification outputs the confidence coefficient, the identification result (for example, the compliance is 0, and the non-compliance is 1) classification signal (for short, classification signal) and the like through decoding. The input end of the 1 st 'not' logic module (53) is connected with the classified signal output end of the module (51), when the safety belt is identified as being in compliance, the signal '0' is sent to the module (53), and the signal '1' is output through 'not' logic operation and sent to the 1 st input end of the 2 nd 'and' logic module (61); (51) the module's classification signal output is also directly connected to the 1 st input of the 1 st and logic module (60). The input end of the 1 st high limit alarm module (58) is connected with the confidence coefficient output end of the 1 st reading module (51), and the 1 st parameter setting module (54) transmits the high limit threshold value to the high limit setting end thereof, so that when the confidence coefficient of the safety belt identification is greater than the high limit, the module (58) gives a signal '1'; this signal is applied to the 2 nd input of the 1 st and logic block (60), so that when step S4: when the result of the safety belt identification is 'non-compliance' and the 'identification confidence coefficient is greater than the high-limit threshold value', the module (60) outputs a signal '1', and sends out a 'alarm-on' signal; this signal of the block (58) is also sent to the 2 nd input of the 2 nd and logic block (61), so that when "step S4: when the result of the safety belt identification is 'compliance' and the 'identification confidence coefficient is greater than the high-limit threshold value', the module (61) outputs a signal '1', and a 'no alarm' signal is sent out.
The input of the 2 nd reading module (52) terminates "step S3: the output end of the uniform identification integrated process reads "step S3: the parameters sent by the uniform identification comprehensive processing are decoded to respectively output confidence degrees, identification results (for example, the compliance is 0, and the non-compliance is 1), classification signals (for short, classification signals) and the like. The input end of the 2 nd 'not' logic module (55) is connected with the classified signal output end of the module (52), when the uniform identification comprehensive processing uniform identifies that the wearing is in compliance, the signal '0' is sent to the module (55), and 1 is output through 'not' logic operation. The input end of the 2 nd high limit alarm module (59) is connected with the confidence coefficient output end of the 2 nd reading module (52), and the 2 nd parameter setting module (56) transmits the high limit threshold value to the high limit setting end thereof, so that when the confidence coefficient of uniform identification is greater than the high limit, the module (59) gives a signal '1'; this signal is applied to the 2 nd input of the 3 rd and logic block (62), so that when "S3: when the result of uniform identification comprehensive processing is 'compliance' and 'identification confidence coefficient is greater than a high-limit threshold value', a signal '1' is output by a (62) module; this signal from block (59) is also applied to the 2 nd input of the 4 th and logic block (63), the 1 st input of block (63) being connected directly to the classification signal output of block (52), so that when "step S3: when the result of the uniform identification integration process is "non-compliance" and "identification confidence is greater than the upper threshold", a signal "1" is output by the module (63).
The 1 st input end of the 5 th 'and' logic module (68) is connected with the output of the 1 st mode (only wearing the working clothes) change-over switch (65), the 2 nd input end of the module (68) is connected with the output of the 3 rd 'and' logic module (62), thus when the 'uniform wearing compliance' appears under the 'working clothes wearing' working condition, (68) the module outputs a signal '1', and sends out a 'no alarm' signal.
The 1 st input end of the 10 th 'and' logic module (73) is connected with the output of the 1 st mode (only wearing the working clothes) change-over switch (65), the 2 nd input end of the module (73) is connected with the output of the 4 th 'and' logic module (63), thus when the 'non-compliance of uniform wearing' occurs under the working clothes wearing condition, the module (73) outputs a signal '1', and sends an 'alarm' signal.
The 1 st input end of the 6 th 'and' logic module (69) is connected with the output of a 2 nd mode (wearing work clothes and safety belts) change-over switch (66), the 2 nd input end of the module (69) is connected with the output of the 2 nd 'and' logic module (61), the 3 rd input end of the module (69) is connected with the output of the 3 rd 'and' logic module (62), thus, when the 'safety belt wearing compliance' appears under the working condition of wearing work clothes and safety belts and the 'uniform wearing compliance', the module (69) outputs a signal '1', and sends out a 'no alarm' signal.
The 1 st input end of the 1 st OR logic module (64) is connected with the output of the 1 st AND logic module (60), the 2 nd input end is connected with the output of the 4 th AND logic module (63), and the output of the 2 nd OR logic module is connected with the 2 nd input end of the 9 th AND logic module (72); the 1 st input end of the 9 th 'and' logic module (72) is connected with the output of the 2 nd mode (wearing work clothes and safety belts) change-over switch (66). Thus, when the working clothes and the safety belts are worn, the safety belts are not worn in a standard way or the uniform is not worn in a standard way, the module (72) outputs a signal (1) to send out an alarm signal.
The 1 st input end of the 7 th 'and' logic module (70) is connected with the output of a 3 rd mode (wearing a safety belt) change-over switch (67), the 2 nd input end of the module (70) is connected with the output of the 2 nd 'and' logic module (61), so that when the 'safety belt wearing compliance' occurs under the 'wearing safety belt' working condition, the module (70) outputs a signal '1', and sends out a 'no alarm' signal.
The 1 st input end of the 8 th 'and' logic module (71) is connected with the output of a 3 rd mode (wearing a safety belt) change-over switch (67), the 2 nd input end of the module (71) is connected with the output of the 1 st 'and' logic module (60), so that when the 'safety belt wearing is not normal' under the 'wearing safety belt' working condition, the module (71) outputs a signal '1', and sends an 'alarm' signal.
The 1 st input end of the 2 nd or logic module (75) is connected with the output of the 5 th and logic module (68), the 2 nd input end is connected with the output of the 6 th and logic module (69), and the 3 rd input end is connected with the output of the 7 th and logic module (70), so that under the above 3 working conditions, when no alarm signal exists, a normal state signal is sent out, and the normal state signal is output to a picture through the 5OG end to be displayed in a corresponding color.
The 1 st input end of the 3 rd OR logic module (76) is connected with the output of the 8 th AND logic module (71), the 2 nd input end is connected with the output of the 9 th AND logic module (72), and the 3 rd input end is connected with the output of the 10 th AND logic module (73), so that under the above 3 working conditions, when an alarm signal is generated, an abnormal state signal is sent out and is output to a picture through the 5OR end to be displayed in a corresponding color.
The A input of the 1 st switching module (76) is connected with 0, the B input thereof is connected with the 3 rd constant value module (82) (coding: 110), the control thereof is connected with the output end of the 5 th AND logic module (68), the A input of the 2 nd switching module (77) is connected with the output end of the 1 st switching module (76), the B input thereof is connected with the 4 th constant value module (83) (coding: 130), the control thereof is connected with the output end of the 6 th AND logic module (69), the A input of the 3 rd switching module (78) is connected with the output end of the 2 nd switching module (77), the B input thereof is connected with the 5 th constant value module (84) (coding: 120), the control thereof is connected with the output end of the 7 th AND logic module (70), the A input of the 4 th switching module (79) is connected with the output end of the 3 rd switching module (78), and the B input thereof is connected with the 6 th constant value module (85) (coding: 121), The control end of the switch is connected with the output end of the 8 th AND logic module (71), the A input end of the 5 th switch module (80) is connected with the output end of the 4 th switch module (79), the B input end of the switch is connected with the 7 th constant value module (86) (code: 131), the control end of the switch is connected with the output end of the 9 th AND logic module (72), the A input end of the 6 th switch module (81) is connected with the output end of the 5 th switch module (80), the B input end of the switch is connected with the 8 th constant value module (87) (code: 110), the control end of the switch is connected with the output end of the 10 th AND logic module (73), and the output of the switch is used as the identification state code and sent to the output 5 OC.
Thus, in the initial state, no recognition result is generated, all the switching modules are switched to the A end, and a 0 signal is sent to the 5OC output. When the result of the identification causes the 5 th AND logic module to output 1, the output of the 1 st switching module (76) outputs the original value of 0 to the A end, the output is switched to the code of 'working clothes not alarming' (such as 110) set by the B end output module (82), and the other switching modules are switched to the A end, and the code set by the module (82) is sent to the 5OC output. When the result of the identification causes the 6 th AND logic module to output 1, the output of the 2 nd switching module (77) is switched to the original output connected to the A end into the code (such as 130) set by the output module (83) connected to the B end, wherein the code is neither working clothes nor safety belts alarming, and the other switching modules are switched to the A end, and the code set by the module (83) is sent to the 5OC output. When the 7 th AND logic module outputs 1 as a recognition result, the output of the 3 rd switching module (78) is switched to the output originally connected to the A end into the safety belt non-alarm code (such as 120) set by the B end output module (84), and the other switching modules are switched to the A end and send the code set by the module (84) to the 5OC output. When the 8 th AND logic module outputs 1 as the recognition result, the output of the 4 th switching module (79) is switched to the safety belt alarm code (such as 121) set by the B-end output module (85) from the output of the original A end, and the other switching modules are switched to the A end, so that the code set by the module (85) is sent to the 5OC output. When the result of the identification causes the 9 th AND logic module to output 1, the output of the 5 th switching module (80) is switched to the output originally connected to the A end into the safety belt alarm code (such as 131) set by the B end output module (86), and the other switching modules are switched to the A end and send the code set by the module (86) to the 5OC output. When the 10 th AND logic module outputs 1 as the recognition result, the output of the 6 th switching module (81) switches the output originally connected to the A end into the safety belt alarm code (such as 111) set by the B end output module (87), and the other switching modules are switched to the A end, and then the code set by the module (87) is sent to the 5OC output. Thus, corresponding codes are formed according to different recognition results and are output through the 5 OC.
Further, step S2 is also connected: and reading human body parameters at the output end OF human body identification, and then outputting the intercepted human body pictures (masks) through the 5 OF.
In the working clothing identification method provided by the embodiment of the invention, video frame data of an electric power working site is obtained through S1, S2 is carried out, a human body is identified according to the video frame data to obtain human body parameters, S3 is carried out, uniform identification comprehensive treatment is carried out based on the human body parameters to obtain safety clothing comprehensive loss function values, S4 is carried out, safety belt identification is carried out based on the human body parameters to obtain corresponding parameters and loss function values of the safety belt, S5 is carried out, the safety clothing comprehensive loss function values and the corresponding parameters and loss function values of the safety belt are analyzed to obtain safety state data, S6 is carried out, whether clothing of an operator is in compliance or not is determined based on the safety state data, and the technical problem that the safety hidden danger is caused by the low safety identification rate of the existing clothing identification method at present is solved through the working clothing identification method, personnel's work clothes and safety belt normative dress in the at utmost management and control operation process improve the normative and the security of construction operation.
Referring to fig. 5, fig. 5 is a flowchart illustrating a method for calculating a loss function value of a safety clothing category in a working clothing identification method according to the present invention, including:
step S501, extracting a preset uniform category identification model;
in the embodiment of the invention, a preset uniform class recognition model loaded in a program model and pre-trained is extracted. All pixels belonging to the uniform class are assigned the same color/value in the mask.
Step S502, converting the human body parameter format into preliminary uniform type picture data;
in the embodiment of the invention, the picture in the BGR format of the human body parameters is read, the RGB format is converted through a color conversion function, the obtained uniform RGB format data is converted into the picture data with the given size, and the preliminary uniform category picture data is obtained.
Step S503, carrying out format conversion on the preliminary uniform category picture data to obtain uniform category picture data;
in the embodiment of the present invention, the obtained picture data of a given size is converted into the (C, H, W) format and normalized. The read-in scaled RGB format data is converted to a Tensor format shaped as (C, H, W) and the value divided by 255 is normalized to a value between 0, 1.0. Further according to the following calculation formula: and (5) carrying out normalization calculation on the image-mean/std to obtain uniform type picture data.
Step S504, inputting the uniform class picture data into the uniform class identification model to obtain a uniform class classification result and a uniform class loss function value;
in the embodiment of the invention, the picture is sent to the uniform model for recognition, the converted picture numerical data is sent to the uniform type recognition model for iterative optimization reasoning calculation, and finally the minimum loss function numerical value for uniform recognition is calculated.
According to the method for calculating the safety uniform category loss function value in the work clothing identification method, provided by the embodiment of the invention, the human body model data is converted into the picture data, and the picture data is input into the preset uniform category identification model to obtain the uniform category classification result and the uniform category loss function value, so that the normative wearing of the work uniform of the personnel in the operation process is controlled to the maximum extent, the normative and the safety of the construction operation are improved, the safety risk is reduced, and the supervision efficiency is improved.
Referring to fig. 6, fig. 6 is a flowchart illustrating a method for calculating a color loss function value of a safety suit in a working clothing identification method according to the present invention, including:
step S601, extracting a preset uniform color classification model;
in the embodiment of the invention, a preset uniform color classification model loaded in a program model and pre-trained is extracted. Pixels belonging to different colors of the uniform are all assigned the same color value in the mask.
Step S602, converting the human body parameters to obtain a function of an image histogram;
in the embodiment of the invention, the read human body parameters are converted into HSV from BGR, the processed human body parameters are read, the picture in the BGR format is read into the picture in the HSV format through a color conversion function, and the function of the HSV (histogram) of the image is calculated through calculating the histogram function of the image.
Step S603, carrying out normalization processing on the function of the image histogram to obtain uniform color classification picture data;
in the embodiment of the invention, the histogram function of the picture is normalized, the obtained histogram function of the picture is normalized through the normalization function, and the numerical value is divided by 255 and normalized to the numerical value between [0,1.0], so that uniform color classification picture data is obtained.
Step S604, inputting uniform color classification picture data into the uniform color classification model to obtain uniform color classification parameters;
in the embodiment of the invention, the picture data is sent to the uniform model for identification, the converted picture numerical value is sent to the uniform color classification model for reasoning calculation, and each parameter of uniform color is identified.
Step S605, inputting the uniform color classification parameters into the uniform color classification model, and calculating to obtain a uniform color classification result and a uniform color loss function value;
in the embodiment of the invention, the loss function is calculated, the obtained parameters of the uniform color are used and brought into the uniform color classification model, iterative optimization reasoning calculation is carried out, and finally the minimum loss function value for uniform color identification is calculated.
According to the method for calculating the safety uniform category loss function value in the work clothing identification method, provided by the embodiment of the invention, the human body model data is converted into the picture data, and the picture data is input into the preset uniform color classification model to obtain the uniform color classification result and the uniform color loss function value, so that the normative wearing of the work uniform of the personnel in the operation process is controlled to the maximum extent, the normative and the safety of the construction operation are improved, the safety risk is reduced, and the supervision efficiency is improved.
Referring to fig. 7, fig. 7 is a flowchart illustrating a method for calculating a seat belt loss function value in a working clothing identification method according to the present invention, including:
step S701, extracting a preset safety belt identification model;
in the embodiment of the invention, the preset safety belt identification model is extracted from the preset model.
Step S702, converting the human body parameters into preliminary safety belt picture data by using a color conversion function formula;
in the embodiment of the invention, the BGR format data of the human body parameter picture is read and converted into the RGB format through the color conversion function formula.
And converting the obtained RGB format data of the safety belt into picture data with a given size to obtain primary safety belt picture data.
Step S703, carrying out normalization processing on the preliminary safety belt picture data to obtain safety belt picture data;
in the embodiment of the invention, the obtained picture data with a given size is converted into a (C, H, W) format and subjected to normalization processing, the read-in scaled RGB format data is converted into a Tensor format shaped as (C, H, W), and the value is divided by 255 and normalized to a value between [0,1.0 ]. Further according to the following calculation formula: and (5) carrying out normalization calculation on the image-mean/std to obtain the safety belt picture data.
Step S704, inputting the safety belt picture data into a preset safety belt identification model to obtain corresponding parameters and loss function values of the safety belt;
in the embodiment of the invention, the converted safety belt picture data is sent to a safety belt identification model network for reasoning calculation, and each parameter of a safety belt result is identified and output to a storage unit.
And calculating a loss function, sending the converted safety belt picture data to a safety belt identification model network for iterative optimization reasoning calculation, and finally calculating the minimum safety belt identification loss function value.
In the embodiment of the invention, the human body model data is converted into the picture data, and the picture data is input into the preset safety belt identification model to obtain the corresponding parameters of the safety belt and the safety belt loss function value, so that the normative wearing of the safety belt of a person in the operation process is controlled to the maximum extent, the normative and the safety of construction operation are improved, the safety risk is reduced, and the supervision efficiency is improved.
Referring to fig. 8, fig. 8 is a block diagram illustrating an embodiment of a working clothing identification device according to the present invention, including:
an obtaining module 801, configured to obtain video frame data of an electric power operation site;
a parameter module 802, configured to identify a human body according to the video frame data to obtain a human body parameter;
a safety clothing module 803, configured to perform comprehensive uniform identification processing based on the human body parameters to obtain a safety clothing comprehensive loss function value;
a safety belt module 804, configured to perform safety belt identification based on the human body parameter, so as to obtain a corresponding parameter and a loss function value of the safety belt;
an analysis module 805, configured to analyze the safety suit comprehensive loss function value, and parameters and loss function values corresponding to the safety belt, to obtain safety state data;
a safety module 806 for determining whether the worker's dressing is in compliance based on the safety status data.
In an alternative embodiment, the parameter module 802 includes:
the conversion submodule is used for carrying out format conversion on the video frame data to obtain picture numerical data;
the parameter submodule is used for inputting the picture numerical data into a preset network neural model to obtain human body parameters;
the judgment submodule is used for judging whether the human body parameters are correct or not; if yes, executing the step of a safety service module; if not, returning to the step of executing the transformation module.
In an alternative embodiment, the safety clothing module 803 includes:
the category submodule is used for carrying out uniform category division based on the human body parameters to obtain uniform category classification results and uniform category loss function values;
the color submodule is used for carrying out color classification calculation on the uniform to obtain a uniform color classification result and a uniform color loss function value;
and the comprehensive submodule is used for carrying out comprehensive calculation on the safety uniform based on the uniform classification result, the uniform classification loss function value, the uniform color classification result and the uniform color loss function value to obtain a numerical value of the comprehensive loss function of the safety uniform.
In an alternative embodiment, the seat belt module 804 includes:
the extraction submodule is used for carrying out safety belt graph segmentation extraction on the human body parameters to obtain safety belt picture data;
and the input submodule is used for inputting the safety belt picture data into a preset safety belt identification model to obtain corresponding parameters and loss function values of the safety belt.
An embodiment of the present invention further provides an electronic device, which includes a memory and a processor, where the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of the method for identifying a work clothing according to any of the above embodiments.
The embodiment of the invention also provides a computer storage medium, on which a computer program is stored, and when the computer program is executed by the processor, the method for identifying the working clothing according to any one of the embodiments is implemented.
It is clear to those skilled in the art that, for convenience and brevity of description, the specific working processes of the above-described systems, apparatuses and units may refer to the corresponding processes in the foregoing method embodiments, and are not described herein again.
In the embodiments provided in the present application, it should be understood that the method, apparatus, electronic device and storage medium disclosed in the present application may be implemented in other ways. For example, the above-described apparatus embodiments are merely illustrative, and for example, the division of the units is only one logical division, and other divisions may be realized in practice, for example, a plurality of units or components may be combined or integrated into another system, or some features may be omitted, or not executed. In addition, the shown or discussed mutual coupling or direct coupling or communication connection may be an indirect coupling or communication connection through some interfaces, devices or units, and may be in an electrical, mechanical or other form.
The units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of the embodiment.
In addition, functional units in the embodiments of the present invention may be integrated into one processing unit, or each unit may exist alone physically, or two or more units are integrated into one unit. The integrated unit can be realized in a form of hardware, and can also be realized in a form of a software functional unit.
The integrated unit, if implemented in the form of a software functional unit and sold or used as a stand-alone product, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present invention may be embodied in the form of a software product, which is stored in a readable storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned readable storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and other various media capable of storing program codes.
The above-mentioned embodiments are only used for illustrating the technical solutions of the present invention, and not for limiting the same; 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 (10)
1. A method for identifying a working garment, comprising:
s1, acquiring video frame data of the electric power operation site;
s2, identifying the human body according to the video frame data to obtain human body parameters;
s3, performing uniform identification comprehensive treatment based on the human body parameters to obtain a safety uniform comprehensive loss function value;
s4, carrying out safety belt identification based on the human body parameters to obtain corresponding parameters of the safety belt and a loss function value;
s5, analyzing the safety suit comprehensive loss function value, the safety belt corresponding parameter and the loss function value to obtain safety state data;
and S6, determining whether the dress of the operator is in compliance or not based on the safety state data.
2. The method for identifying active dresses as in claim 1, wherein said step S2 comprises:
s21, converting the format of the video frame data to obtain picture numerical data;
s22, inputting the picture numerical data into a preset network neural model to obtain human body parameters;
s23, judging whether the human body parameters are correct or not; if yes, go to step S3; if not, the process returns to step S21.
3. The method for identifying a working garment according to claim 1 or 2, wherein the step S3 includes:
s31, dividing the uniform categories based on the human body parameters to obtain uniform category classification results and uniform category loss function values;
s32, performing color classification calculation on the uniform to obtain a uniform color classification result and a uniform color loss function value;
and S33, performing comprehensive calculation of the safety uniform based on the uniform classification result, the uniform classification loss function value, the uniform color classification result and the uniform color loss function value to obtain a numerical value of the comprehensive loss function of the safety uniform.
4. The method for identifying active dresses as in claim 1, wherein said step S4 comprises:
s41, carrying out safety belt graph segmentation and extraction on the human body parameters to obtain safety belt picture data;
and S42, inputting the safety belt picture data into a preset safety belt identification model to obtain corresponding parameters and loss function values of the safety belt.
5. An active wear identification device, comprising:
the acquisition module is used for acquiring video frame data of an electric power operation site;
the parameter module is used for identifying a human body according to the video frame data to obtain human body parameters;
the safety clothing module is used for carrying out uniform identification comprehensive treatment based on the human body parameters to obtain a safety clothing comprehensive loss function value;
the safety belt module is used for carrying out safety belt identification based on the human body parameters to obtain corresponding parameters and loss function values of the safety belt;
the analysis module is used for analyzing the safety suit comprehensive loss function value, the corresponding parameters of the safety belt and the loss function value to obtain safety state data;
and the safety module is used for determining whether the dressing of the operator is in compliance or not based on the safety state data.
6. The active wear identification device of claim 5, wherein the parameter module comprises:
the conversion submodule is used for carrying out format conversion on the video frame data to obtain picture numerical data;
the parameter submodule is used for inputting the picture numerical data into a preset network neural model to obtain human body parameters;
the judgment submodule is used for judging whether the human body parameters are correct or not; if yes, executing the step of a safety service module; if not, returning to the step of executing the transformation module.
7. The active apparel identification device of claim 5or 6 wherein the safety garment module comprises:
the category submodule is used for carrying out uniform category division based on the human body parameters to obtain uniform category classification results and uniform category loss function values;
the color submodule is used for carrying out color classification calculation on the uniform to obtain a uniform color classification result and a uniform color loss function value;
and the comprehensive submodule is used for carrying out comprehensive calculation on the safety uniform based on the uniform classification result, the uniform classification loss function value, the uniform color classification result and the uniform color loss function value to obtain a numerical value of the comprehensive loss function of the safety uniform.
8. The active wear identification device of claim 5, wherein the seat belt module comprises:
the extraction submodule is used for carrying out safety belt graph segmentation extraction on the human body parameters to obtain safety belt picture data;
and the input submodule is used for inputting the safety belt picture data into a preset safety belt identification model to obtain corresponding parameters and loss function values of the safety belt.
9. An electronic device comprising a processor and a memory, the memory storing computer readable instructions that, when executed by the processor, perform the method of any of claims 1-4.
10. A storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, performs the method according to any of claims 1-4.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111114346.0A CN113822197A (en) | 2021-09-23 | 2021-09-23 | Work dressing identification method and device, electronic equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111114346.0A CN113822197A (en) | 2021-09-23 | 2021-09-23 | Work dressing identification method and device, electronic equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113822197A true CN113822197A (en) | 2021-12-21 |
Family
ID=78915227
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111114346.0A Pending CN113822197A (en) | 2021-09-23 | 2021-09-23 | Work dressing identification method and device, electronic equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113822197A (en) |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111242209A (en) * | 2020-01-08 | 2020-06-05 | 久微信息技术无锡有限公司 | Uniform identification method |
| CN111814636A (en) * | 2020-06-29 | 2020-10-23 | 北京百度网讯科技有限公司 | Safety belt detection method and device, electronic equipment and storage medium |
| CN112132040A (en) * | 2020-09-24 | 2020-12-25 | 明见(厦门)软件开发有限公司 | Vision-based safety belt real-time monitoring method, terminal equipment and storage medium |
| CN112183472A (en) * | 2020-10-28 | 2021-01-05 | 西安交通大学 | Method for detecting whether test field personnel wear work clothes or not based on improved RetinaNet |
| CN112232142A (en) * | 2020-09-27 | 2021-01-15 | 浙江大华技术股份有限公司 | Safety belt identification method and device and computer readable storage medium |
| CN112633308A (en) * | 2020-09-15 | 2021-04-09 | 北京华电天仁电力控制技术有限公司 | Detection method and detection system for whether power plant operating personnel wear safety belts |
| CN113297900A (en) * | 2021-04-02 | 2021-08-24 | 中国地质大学（武汉） | Method, device, equipment and storage medium for identifying video stream safety helmet based on YOLO |
| CN113297913A (en) * | 2021-04-26 | 2021-08-24 | 云南电网有限责任公司信息中心 | Method for identifying dressing specification of distribution network field operating personnel |
-
2021
- 2021-09-23 CN CN202111114346.0A patent/CN113822197A/en active Pending
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111242209A (en) * | 2020-01-08 | 2020-06-05 | 久微信息技术无锡有限公司 | Uniform identification method |
| CN111814636A (en) * | 2020-06-29 | 2020-10-23 | 北京百度网讯科技有限公司 | Safety belt detection method and device, electronic equipment and storage medium |
| CN112633308A (en) * | 2020-09-15 | 2021-04-09 | 北京华电天仁电力控制技术有限公司 | Detection method and detection system for whether power plant operating personnel wear safety belts |
| CN112132040A (en) * | 2020-09-24 | 2020-12-25 | 明见(厦门)软件开发有限公司 | Vision-based safety belt real-time monitoring method, terminal equipment and storage medium |
| CN112232142A (en) * | 2020-09-27 | 2021-01-15 | 浙江大华技术股份有限公司 | Safety belt identification method and device and computer readable storage medium |
| CN112183472A (en) * | 2020-10-28 | 2021-01-05 | 西安交通大学 | Method for detecting whether test field personnel wear work clothes or not based on improved RetinaNet |
| CN113297900A (en) * | 2021-04-02 | 2021-08-24 | 中国地质大学（武汉） | Method, device, equipment and storage medium for identifying video stream safety helmet based on YOLO |
| CN113297913A (en) * | 2021-04-26 | 2021-08-24 | 云南电网有限责任公司信息中心 | Method for identifying dressing specification of distribution network field operating personnel |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106027978B (en) | A kind of video monitoring abnormal behaviour method of smart home endowment | |
| CN112560745B (en) | Method for discriminating personnel on electric power operation site and related device | |
| CN113034826A (en) | Video-based abnormal event warning method and system, equipment and storage medium thereof | |
| CN111414812A (en) | Human body attribute identification method, system, computer device and storage medium | |
| CN111931652A (en) | Dressing detection method and device and monitoring terminal | |
| CN110781735A (en) | Alarm method and system for identifying on-duty state of personnel | |
| CN112364740A (en) | Unmanned machine room monitoring method and system based on computer vision | |
| KR20220051609A (en) | A system for recognizing workers' wearing safety protection equipment through real-time image analysis | |
| CN113313186B (en) | Method and system for identifying irregular wearing work clothes | |
| CN108765454A (en) | A kind of smog detection method, device and device end based on video | |
| CN113673495B (en) | Intelligent security method and system based on neural network and readable storage medium | |
| CN113485277B (en) | Intelligent power plant video identification monitoring management system and method | |
| CN113743725B (en) | Ascending qualification detection method and system for ascending operators based on edge calculation | |
| CN113963162A (en) | Helmet wearing identification method and device, computer equipment and storage medium | |
| US20220036568A1 (en) | Systems and methods for detecting and tracking humans in captured images | |
| CN113822197A (en) | Work dressing identification method and device, electronic equipment and storage medium | |
| CN111460895B (en) | Method and device for detecting protective equipment in visual data and storage medium | |
| CN115830641B (en) | Employee identification method and device, electronic equipment and storage medium | |
| CN111582183A (en) | Mask identification method and system in public place | |
| CN113536842A (en) | Electric power operator safety dressing identification method and device | |
| WO2022121886A1 (en) | Method and apparatus for identifying dress code for electric power operations | |
| CN110796068A (en) | Drowning detection method and system for community swimming pool | |
| CN115953815A (en) | Monitoring method and device for infrastructure site | |
| CN113869245A (en) | Method and device for identifying safety region | |
| CN114140819A (en) | Device and method for assisting safety management in power grid operation |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20211221 |