CN115905807B - Coarse grid optimization operation method based on deep learning - Google Patents
Coarse grid optimization operation method based on deep learning Download PDFInfo
- Publication number
- CN115905807B CN115905807B CN202211446523.XA CN202211446523A CN115905807B CN 115905807 B CN115905807 B CN 115905807B CN 202211446523 A CN202211446523 A CN 202211446523A CN 115905807 B CN115905807 B CN 115905807B
- Authority
- CN
- China
- Prior art keywords
- grid
- training
- slag
- transducer
- judging whether
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Landscapes
- Image Processing (AREA)
- Image Analysis (AREA)
Abstract
The invention provides a coarse grid optimizing operation method based on deep learning, which belongs to the technical field of communication, and automatically judges impurity quantity and abnormal impurities through a transducer neural network model, and adjusts the coarse grid operation rate by combining the impurity quantity and ensures the safety of an operation site by combining a harmful gas concentration sensor; the invention solves the problems of automatic impurity quantity identification and abnormal impurity identification, and provides a data basis for automatically processing impurities.
Description
Technical Field
The invention belongs to the technical field of coarse grid optimization, and particularly relates to a sewage coarse grid optimization operation method based on deep learning.
Background
As the first structure on the sewage plant process flow, the coarse grille can effectively intercept large-particle impurities in sewage, which is generally more than 15mm, so as to ensure that the large-particle impurities do not enter a pump house and do not influence a water pump or subsequent structure treatment. At present, the automation degree of the operation of the coarse grille is not high enough, the operation of the grille slag is controlled by the front and back water levels of the grille, feedback is often delayed, the operation and maintenance environment of the coarse grille is poor, and the odor and the mosquitoes and flies are not easy to get restless. The anaerobic environment of the coarse grille is easy to generate toxic gases such as methane, hydrogen sulfide and the like, and a plurality of coarse grille ponds are overhauled in China to poison and coma, so that the event of worker drowning is caused. On the other hand, the transportation of the coarse grid slag is always subject to the problem that people are often required to push the grid slag, the working environment is poor, and the automation and the intelligent degree are urgently required to be improved.
Disclosure of Invention
Aiming at the defects in the prior art, the coarse grid optimizing operation method based on deep learning provided by the invention solves the problems of automatic impurity identification and abnormal impurities, provides a data basis for automatically processing the impurities, simultaneously detects the safety degree of an impurity area in real time through a sensor, solves the detection and alarm of harmful gas, and effectively solves the safety problem of people.
In order to achieve the above purpose, the invention adopts the following technical scheme:
the scheme provides a coarse grid optimization operation method based on deep learning, which comprises the following steps:
s1, collecting video images above grid channels and positions where the grid slag barrels are placed, marking the collected images by a camera above the grid channels to obtain classification training positive and negative samples for judging whether abnormal impurities exist or not and training samples for judging the impurity quantity, and marking the video images above the positions where the grid slag barrels are placed to obtain classification training positive and negative samples for judging whether the grid slag barrels exist or not and positive and negative samples for judging whether the grid slag in the grid slag barrels is full or not;
s2, identifying a slag barrel in the video image above the position where the slag barrel is placed, judging whether the slag barrel exists according to the identification result, if so, performing a step S4, otherwise, entering a step S3;
s3, acquiring a concentration value of toxic gas, performing exception handling by using a toxic gas sensor, and ending the flow;
s4, training an improved transducer neural network model by utilizing classification training positive and negative samples for judging whether a grating slag barrel exists or not;
s5, judging the condition of the grid slag in the grid slag barrel by using a trained transducer neural network model according to positive and negative samples of whether the grid slag in the grid slag barrel is full, and ending the flow according to a judging result;
s6, training the improved transducer neural network model by using a training sample for judging impurity quantity to obtain a first classification model, identifying impurities by using the first classification model, adjusting the coarse grid rotation rate by combining the liquid level difference, and ending the flow;
and S7, training the improved transducer neural network model by utilizing the positive and negative samples for classification training for judging whether the abnormal impurities exist, obtaining a second classification model, judging whether the abnormal impurities exist by utilizing the second classification model, and ending the flow according to the judging result.
The beneficial effects of the invention are as follows: according to the invention, the impurity amount and abnormal impurities are automatically judged through a transducer neural network model, and the operation rate of the coarse grid is regulated by combining the impurity amount; and the safety of the operation site is ensured by combining a harmful gas concentration sensor; through the design, the automatic impurity treatment device has higher automation treatment level, can greatly reduce the workload of operators, improves the impurity treatment efficiency, solves the problems of automatic impurity identification and abnormal impurity identification, and provides a data basis for automatically treating the impurities.
Further, the step S3 includes the steps of:
s301, acquiring a concentration value of toxic gas;
s302, judging whether the concentration value of the toxic gas is larger than a set safety value by utilizing a toxic gas sensor, if so, giving an alarm and ending the flow, otherwise, judging that the concentration value of the toxic gas is lower than the set safety value and ending the flow.
The beneficial effects of the above-mentioned further scheme are: the safety degree of the impurity region is detected in real time through the sensor, so that the detection and alarm of harmful gas are solved, and the safety problem of life is effectively solved.
Still further, the step S5 includes the steps of:
s501, judging whether the grid slag in the grid slag barrel is full or not by utilizing a trained transducer neural network model according to positive and negative samples of whether the grid slag in the grid slag barrel is full, if yes, entering a step S502, otherwise, ending the flow;
s502, sending a lifting and replacing grid slag barrel command, and ending the flow.
The beneficial effects of the above-mentioned further scheme are: the invention utilizes the improved transducer neural network model to judge whether the grid slag in the grid slag barrel is full, can stably and accurately judge whether the grid slag barrel is full, and provides a judging basis for automatically replacing the grid slag barrel by a system.
Still further, the transducer neural network model comprises a full connection layer, an encoder, a multi-layer perception classifier and an output layer which are sequentially connected;
the full-connection layer is used for obtaining an image characteristic block with the same size as the video image block according to an exception processing result, and paving and compressing the image characteristic block;
the encoder is used for extracting abstract image features of the image feature blocks and positions of the encoded image blocks according to the processing result;
the multi-layer perception classifier is used for classifying the coding result of the transducer neural network model according to the extracted result to obtain image classification;
the output layer is used for outputting the grid slag condition in the grid slag barrel according to the image classification.
The beneficial effects of the above-mentioned further scheme are: after the sample training is performed by using the transducer neural network model, whether the grid slag barrel is full of slag or not can be judged stably and accurately, and a judgment basis is provided for automatically replacing the grid slag barrel by the system.
Still further, the step S7 includes the steps of:
s701, training an improved transducer neural network model by utilizing a positive and negative sample of classification training for judging whether abnormal impurities exist or not to obtain a second classification model;
s702, judging whether abnormal impurities exist or not by using the second classification model, if yes, executing the step S3, otherwise, ending the flow.
Drawings
FIG. 1 is a flow chart of the method of the present invention.
FIG. 2 is a schematic diagram of a transducer model structure.
Detailed Description
The following description of the embodiments of the present invention is provided to facilitate understanding of the present invention by those skilled in the art, but it should be understood that the present invention is not limited to the scope of the embodiments, and all the inventions which make use of the inventive concept are protected by the spirit and scope of the present invention as defined and defined in the appended claims to those skilled in the art.
Examples
As shown in fig. 1, the invention provides a coarse grid optimization operation method based on deep learning, which comprises the following steps:
s1, collecting video images above grid channels and positions where the grid slag barrels are placed, marking the collected images by a camera above the grid channels to obtain classification training positive and negative samples for judging whether abnormal impurities exist or not and training samples for judging the impurity quantity, and marking the video images above the positions where the grid slag barrels are placed to obtain classification training positive and negative samples for judging whether the grid slag barrels exist or not and positive and negative samples for judging whether the grid slag in the grid slag barrels is full or not;
in the embodiment, videos above a grid channel and a grid slag barrel are collected, two kinds of training samples (normal impurities and abnormal impurities) for judging whether abnormal impurities exist or not and training samples for judging the impurity quantity (five kinds of training samples are in total: trace, small, medium, large and large) are marked, about 2 ten thousand images of image data of the two kinds of samples are taken as positive samples, a certain amount of negative sample quantity is collected at the same time, and the samples are marked manually. Then step S2 and step S5 are performed simultaneously in parallel.
S2, identifying a slag barrel in the video image above the position where the slag barrel is placed, judging whether the slag barrel exists according to the identification result, if so, performing a step S4, otherwise, entering a step S3;
in this embodiment, it is determined whether there is a slag ladle: and identifying the grid slag barrel in the video image above the position where the grid slag barrel is placed, judging whether the grid slag barrel exists according to the identification result, executing the step S4 if the grid slag barrel does not exist, and executing the step S3 if the grid slag barrel exists.
S3, acquiring a concentration value of toxic gas, performing exception handling by using a toxic gas sensor, and ending the flow, wherein the implementation method comprises the following steps:
s301, acquiring a concentration value of toxic gas;
s302, judging whether the concentration value of the toxic gas is larger than a set safety value by utilizing a toxic gas sensor, if so, giving an alarm and ending the flow, otherwise, judging that the concentration value of the toxic gas is lower than the set safety value and ending the flow.
In this embodiment, firstly, toxic gas sensors, such as a methane concentration sensor and a hydrogen sulfide concentration sensor, are called, if one or more toxic gas concentrations are greater than a set safety value, an alarm completion process is performed, and if all the monitored toxic gas concentrations are lower than the safety value, a manual processing completion process is performed.
S4, training an improved transducer neural network model by utilizing classification training positive and negative samples for judging whether a grating slag barrel exists or not;
s5, judging the condition of the grid slag in the grid slag barrel by using a trained transducer neural network model according to positive and negative samples of whether the grid slag in the grid slag barrel is full, and ending the flow according to a judging result, wherein the implementation method comprises the following steps:
s501, judging whether the grid slag in the grid slag barrel is full or not by utilizing a trained transducer neural network model according to positive and negative samples of whether the grid slag in the grid slag barrel is full, if yes, entering a step S502, otherwise, ending the flow;
s502, sending a lifting and replacing grid slag barrel command, and ending the flow.
The transducer neural network model comprises a full-connection layer, an encoder, a multi-layer perception classifier and an output layer which are sequentially connected;
the full-connection layer is used for obtaining an image characteristic block with the same size as the video image block according to an exception processing result, and paving and compressing the image characteristic block;
the encoder is used for extracting abstract image features of the image feature blocks and positions of the encoded image blocks according to the processing result;
the multi-layer perception classifier is used for classifying the coding result of the transducer neural network model according to the extracted result to obtain image classification;
the output layer is used for outputting the grid slag condition in the grid slag barrel according to the image classification.
In this embodiment, a transducer model is called to determine whether the slag ladle is full: and calling an improved transducer model to judge whether the grid slag barrel is full, sending a lifting and replacing grid slag barrel instruction to finish the process if the grid slag barrel is full, and ending the process if the grid slag barrel is not full.
In order to improve the information of each image block, the transducer model is shown in fig. 2, and the full connection is integrated in the transducer, the image block first passes through the full connection layer to obtain a feature block with the same size as the image block, and the image feature block performs the operations of image block flattening and compression (Linear Projection of Flattened Patches) and the subsequent encoding flow of the subsequent transducer. Extracting abstract image features of the image feature blocks and positions of the encoded image blocks by using an encoder; classifying the coding result of the transducer neural network model by using a multi-layer perception classifier to obtain image classification; and finally, outputting the grid slag condition in the grid slag barrel through the output layer.
S6, training the improved transducer neural network model by using a training sample for judging impurity quantity to obtain a first classification model, identifying impurities by using the first classification model, adjusting the coarse grid rotation rate by combining the liquid level difference, and ending the flow;
in this embodiment, the improved transducer model for training is invoked to identify the impurity condition types (micro, small, medium, large and abnormal impurities), and the coarse grid running rate is adjusted by combining the liquid level difference to complete the flow. Wherein the trained modified transducer model is the first classification model obtained by training the transducer model in step S4 with the impurity amount training sample data of step S1.
S7, training the improved transducer neural network model by utilizing positive and negative samples for classifying and training for judging whether abnormal impurities exist, obtaining a second classification model, judging whether abnormal impurities exist by utilizing the second classification model, ending the flow according to a judging result, and realizing the method as follows:
s701, training an improved transducer neural network model by utilizing a positive and negative sample of classification training for judging whether abnormal impurities exist or not to obtain a second classification model;
s702, judging whether abnormal impurities exist or not by using the second classification model, if yes, executing the step S3, otherwise, ending the flow.
In this embodiment, the modified transducer model for training is invoked to determine whether there is an abnormal impurity, if so, step S3 is executed, and if not, the flow is ended. The improved transducer model is a second classification model obtained by training the transducer model in the step S4 by using the two-classification sample data of judging whether the abnormal impurities exist in the step S1.
According to the invention, the impurity amount and abnormal impurities are automatically judged through a transducer neural network model, and the operation rate of the coarse grid is regulated by combining the impurity amount; and the safety of the operation site is ensured by combining a harmful gas concentration sensor; through the design, the automatic impurity treatment device has higher automation treatment level, can greatly reduce the workload of operators, improves the impurity treatment efficiency, solves the problems of automatic impurity identification and abnormal impurity identification, and provides a data basis for automatically treating the impurities.
Claims (3)
1. The coarse grid optimizing operation method based on deep learning is characterized by comprising the following steps of:
s1, collecting video images above grid channels and positions where the grid slag barrels are placed, marking the collected images by a camera above the grid channels to obtain classification training positive and negative samples for judging whether abnormal impurities exist or not and training samples for judging the impurity quantity, and marking the video images above the positions where the grid slag barrels are placed to obtain classification training positive and negative samples for judging whether the grid slag barrels exist or not and positive and negative samples for judging whether the grid slag in the grid slag barrels is full or not;
s2, identifying a slag barrel in the video image above the position where the slag barrel is placed, judging whether the slag barrel exists according to the identification result, if so, performing a step S4, otherwise, entering a step S3;
s3, acquiring a concentration value of toxic gas, performing exception handling by using a toxic gas sensor, and ending the flow;
s4, training an improved transducer neural network model by utilizing classification training positive and negative samples for judging whether a grating slag barrel exists or not;
the transducer neural network model comprises a full-connection layer, an encoder, a multi-layer perception classifier and an output layer which are sequentially connected;
the full-connection layer is used for obtaining an image characteristic block with the same size as the video image block according to an exception processing result, and paving and compressing the image characteristic block;
the encoder is used for extracting abstract image features of the image feature blocks and positions of the encoded image blocks according to the processing result;
the multi-layer perception classifier is used for classifying the coding result of the transducer neural network model according to the extracted result to obtain image classification;
the output layer is used for outputting the grid slag condition in the grid slag barrel according to the image classification;
s5, judging the condition of the grid slag in the grid slag barrel by using a trained transducer neural network model according to positive and negative samples of whether the grid slag in the grid slag barrel is full, and ending the flow according to a judging result;
s6, training the improved transducer neural network model by using a training sample for judging impurity quantity to obtain a first classification model, identifying impurities by using the first classification model, adjusting the coarse grid rotation rate by combining the liquid level difference, and ending the flow;
invoking an improved transducer model to identify the impurity condition types, and adjusting the coarse grid running speed by combining the liquid level difference to finish the process, wherein the trained improved transducer model is a first classification model obtained by training the transducer model in the step S4 by using the impurity amount training sample data of the step S1;
s7, training the improved transducer neural network model by utilizing positive and negative samples for classifying and training for judging whether abnormal impurities exist, obtaining a second classification model, judging whether the abnormal impurities exist by utilizing the second classification model, and ending the flow according to a judgment result;
the step S7 includes the steps of:
s701, training an improved transducer neural network model by utilizing a positive and negative sample of classification training for judging whether abnormal impurities exist or not to obtain a second classification model;
s702, judging whether abnormal impurities exist or not by using the second classification model, if yes, executing the step S3, otherwise, ending the flow.
2. The deep learning-based coarse grid optimization operation method according to claim 1, wherein the step S3 comprises the steps of:
s301, acquiring a concentration value of toxic gas;
s302, judging whether the concentration value of the toxic gas is larger than a set safety value by utilizing a toxic gas sensor, if so, giving an alarm and ending the flow, otherwise, judging that the concentration value of the toxic gas is lower than the set safety value and ending the flow.
3. The deep learning-based coarse grid optimization operation method according to claim 2, wherein the step S5 comprises the steps of:
s501, judging whether the grid slag in the grid slag barrel is full or not by utilizing a trained transducer neural network model according to positive and negative samples of whether the grid slag in the grid slag barrel is full, if yes, entering a step S502, otherwise, ending the flow;
s502, sending a lifting and replacing grid slag barrel command, and ending the flow.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211446523.XA CN115905807B (en) | 2022-11-18 | 2022-11-18 | Coarse grid optimization operation method based on deep learning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211446523.XA CN115905807B (en) | 2022-11-18 | 2022-11-18 | Coarse grid optimization operation method based on deep learning |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115905807A CN115905807A (en) | 2023-04-04 |
| CN115905807B true CN115905807B (en) | 2023-10-20 |
Family
ID=86476227
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211446523.XA Active CN115905807B (en) | 2022-11-18 | 2022-11-18 | Coarse grid optimization operation method based on deep learning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115905807B (en) |
Citations (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| DE19608826A1 (en) * | 1996-03-07 | 1997-09-11 | Walter Kuntschar | Gasification reactor |
| CN107162315A (en) * | 2017-05-18 | 2017-09-15 | 重庆市托尔阿诗环保有限公司 | Biological cycle method sewage disposal system |
| CN107563431A (en) * | 2017-08-28 | 2018-01-09 | 西南交通大学 | A kind of image abnormity detection method of combination CNN transfer learnings and SVDD |
| CN109389180A (en) * | 2018-10-30 | 2019-02-26 | 国网四川省电力公司广元供电公司 | A power equipment image-recognizing method and inspection robot based on deep learning |
| CN110175247A (en) * | 2019-03-13 | 2019-08-27 | 北京邮电大学 | A method of abnormality detection model of the optimization based on deep learning |
| CN110516583A (en) * | 2019-08-21 | 2019-11-29 | 中科视语(北京)科技有限公司 | A kind of vehicle recognition methods, system, equipment and medium again |
| CN211312804U (en) * | 2019-10-24 | 2020-08-21 | 江苏中兴水务有限公司 | Integrated pump station with rotary type grating dirt remover |
| CN111882057A (en) * | 2020-06-19 | 2020-11-03 | 苏州浪潮智能科技有限公司 | Multi-stage deep learning method oriented to space-time sequence data modeling and application |
| CN112364918A (en) * | 2020-11-10 | 2021-02-12 | 深圳力维智联技术有限公司 | Abnormality recognition method, terminal, and computer-readable storage medium |
| CN112418334A (en) * | 2020-11-26 | 2021-02-26 | 哈尔滨市科佳通用机电股份有限公司 | Method for identifying deformation fault of skirtboard grating of railway bullet train |
| CN113605490A (en) * | 2021-08-09 | 2021-11-05 | 中铁工程设计咨询集团有限公司 | Riverbed type underground water taking structure |
| CN114332007A (en) * | 2021-12-28 | 2022-04-12 | 福州大学 | Transformer-based industrial defect detection and identification method |
| WO2022160413A1 (en) * | 2021-01-29 | 2022-08-04 | 南方电网调峰调频发电有限公司 | Electric power production anomaly monitoring method and apparatus, and computer device and storage medium |
-
2022
- 2022-11-18 CN CN202211446523.XA patent/CN115905807B/en active Active
Patent Citations (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| DE19608826A1 (en) * | 1996-03-07 | 1997-09-11 | Walter Kuntschar | Gasification reactor |
| CN107162315A (en) * | 2017-05-18 | 2017-09-15 | 重庆市托尔阿诗环保有限公司 | Biological cycle method sewage disposal system |
| CN107563431A (en) * | 2017-08-28 | 2018-01-09 | 西南交通大学 | A kind of image abnormity detection method of combination CNN transfer learnings and SVDD |
| CN109389180A (en) * | 2018-10-30 | 2019-02-26 | 国网四川省电力公司广元供电公司 | A power equipment image-recognizing method and inspection robot based on deep learning |
| CN110175247A (en) * | 2019-03-13 | 2019-08-27 | 北京邮电大学 | A method of abnormality detection model of the optimization based on deep learning |
| CN110516583A (en) * | 2019-08-21 | 2019-11-29 | 中科视语(北京)科技有限公司 | A kind of vehicle recognition methods, system, equipment and medium again |
| CN211312804U (en) * | 2019-10-24 | 2020-08-21 | 江苏中兴水务有限公司 | Integrated pump station with rotary type grating dirt remover |
| CN111882057A (en) * | 2020-06-19 | 2020-11-03 | 苏州浪潮智能科技有限公司 | Multi-stage deep learning method oriented to space-time sequence data modeling and application |
| CN112364918A (en) * | 2020-11-10 | 2021-02-12 | 深圳力维智联技术有限公司 | Abnormality recognition method, terminal, and computer-readable storage medium |
| CN112418334A (en) * | 2020-11-26 | 2021-02-26 | 哈尔滨市科佳通用机电股份有限公司 | Method for identifying deformation fault of skirtboard grating of railway bullet train |
| WO2022160413A1 (en) * | 2021-01-29 | 2022-08-04 | 南方电网调峰调频发电有限公司 | Electric power production anomaly monitoring method and apparatus, and computer device and storage medium |
| CN113605490A (en) * | 2021-08-09 | 2021-11-05 | 中铁工程设计咨询集团有限公司 | Riverbed type underground water taking structure |
| CN114332007A (en) * | 2021-12-28 | 2022-04-12 | 福州大学 | Transformer-based industrial defect detection and identification method |
Non-Patent Citations (2)
| Title |
|---|
| Sitalakshmi Venkataraman.A Grid-based Neural Network Framework for Multimodal Biometrics.《World Academy of Science,Engineering and Technology》.2010,第298-303页. * |
| 城市生活污水处理过程中职业病危害探讨;孔国栋;《中国优秀硕士学位论文全文数据库医药卫生科技辑》(第3期);第E055-12页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115905807A (en) | 2023-04-04 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111930049A (en) | Forestry forest fire prevention safety intelligent monitoring management system based on thing networking | |
| CN103993649B (en) | Rainwater collection pool Controlling System and method | |
| CN110335294B (en) | Mine water pump house water leakage detection method based on frame difference method and 3D convolutional neural network | |
| CN113155102A (en) | River comprehensive treatment system | |
| CN116152748B (en) | River and lake supervision method and system based on blue algae identification | |
| CN103096121A (en) | Camera moving detecting method and device | |
| CN114943917A (en) | Algorithm for visually identifying aeration rate of aerobic tank of sewage plant | |
| CN100469138C (en) | Power transformer draught fan state recognizing method based on video monitoring and image recognition | |
| CN110599458A (en) | Underground pipe network detection and evaluation cloud system based on convolutional neural network | |
| CN115905807B (en) | Coarse grid optimization operation method based on deep learning | |
| CN117314709A (en) | Intelligent monitoring system for sewage treatment progress | |
| CN108664886A (en) | A kind of fast face recognition method adapting to substation's disengaging monitoring demand | |
| CN117649123A (en) | Wisdom garden management system | |
| CN116721550B (en) | Traffic management system based on grid computing | |
| CN113255519A (en) | Crane lifting arm identification system and multi-target tracking method for power transmission line dangerous vehicle | |
| CN112906593A (en) | Sluice image identification method based on fast RCNN | |
| CN116977909A (en) | Deep learning fire intensity recognition method and system based on multi-modal data | |
| CN106846312A (en) | A kind of method of the monitoring sludge system hydraulics change based on color space | |
| CN1975392A (en) | Flue gas discharge continuous monitoring system based on image processing | |
| CN110807174A (en) | Effluent analysis and abnormity identification method for sewage plant group based on statistical distribution | |
| CN105187821A (en) | Video quality intelligent diagnostic analysis system | |
| CN115496716A (en) | Single and double micronucleus cell image detection method based on deep learning and related equipment | |
| CN111675257B (en) | Remote centralized control method and system for sewage treatment plant | |
| CN113781388A (en) | Image enhancement-based power transmission line channel hidden danger image identification method and device | |
| CN117557166B (en) | Multi-mode and real-time data user data environment intelligent monitoring system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |