CN110765937A - Coal yard spontaneous combustion detection method based on transfer learning - Google Patents
Coal yard spontaneous combustion detection method based on transfer learning Download PDFInfo
- Publication number
- CN110765937A CN110765937A CN201911007182.4A CN201911007182A CN110765937A CN 110765937 A CN110765937 A CN 110765937A CN 201911007182 A CN201911007182 A CN 201911007182A CN 110765937 A CN110765937 A CN 110765937A
- Authority
- CN
- China
- Prior art keywords
- neural network
- spontaneous combustion
- convolutional neural
- network model
- coal pile
- 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
- 239000003245 coal Substances 0.000 title claims abstract description 105
- 238000002485 combustion reaction Methods 0.000 title claims abstract description 99
- 230000002269 spontaneous effect Effects 0.000 title claims abstract description 98
- 238000001514 detection method Methods 0.000 title claims abstract description 31
- 238000013526 transfer learning Methods 0.000 title claims abstract description 13
- 238000013527 convolutional neural network Methods 0.000 claims abstract description 75
- 238000010586 diagram Methods 0.000 claims abstract description 48
- 238000013528 artificial neural network Methods 0.000 claims abstract description 36
- 238000012549 training Methods 0.000 claims abstract description 35
- 238000000034 method Methods 0.000 claims abstract description 22
- 238000011478 gradient descent method Methods 0.000 claims description 4
- 230000000644 propagated effect Effects 0.000 claims description 4
- 230000008569 process Effects 0.000 abstract description 8
- 238000003062 neural network model Methods 0.000 description 4
- 230000035945 sensitivity Effects 0.000 description 4
- 238000011161 development Methods 0.000 description 3
- 238000003331 infrared imaging Methods 0.000 description 3
- 238000004519 manufacturing process Methods 0.000 description 3
- 230000003042 antagnostic effect Effects 0.000 description 2
- 239000007789 gas Substances 0.000 description 2
- 239000000779 smoke Substances 0.000 description 2
- 241000282414 Homo sapiens Species 0.000 description 1
- 229910000831 Steel Inorganic materials 0.000 description 1
- 230000009471 action Effects 0.000 description 1
- QVGXLLKOCUKJST-UHFFFAOYSA-N atomic oxygen Chemical compound [O] QVGXLLKOCUKJST-UHFFFAOYSA-N 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000004880 explosion Methods 0.000 description 1
- 230000036541 health Effects 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000013508 migration Methods 0.000 description 1
- 230000005012 migration Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 230000003647 oxidation Effects 0.000 description 1
- 238000007254 oxidation reaction Methods 0.000 description 1
- 239000001301 oxygen Substances 0.000 description 1
- 229910052760 oxygen Inorganic materials 0.000 description 1
- 230000008092 positive effect Effects 0.000 description 1
- 230000002265 prevention Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 239000010959 steel Substances 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
- 230000004083 survival effect Effects 0.000 description 1
- 238000001931 thermography Methods 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/10—Terrestrial scenes
- G06V20/176—Urban or other man-made structures
-
- 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
- G08—SIGNALLING
- G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
- G08B17/00—Fire alarms; Alarms responsive to explosion
- G08B17/12—Actuation by presence of radiation or particles, e.g. of infrared radiation or of ions
- G08B17/125—Actuation by presence of radiation or particles, e.g. of infrared radiation or of ions by using a video camera to detect fire or smoke
Abstract
The invention discloses a coal pile spontaneous combustion automatic detection method based on transfer learning, which comprises the steps of firstly, acquiring a marked coal pile spontaneous combustion common flame image dataset and an unmarked coal pile infrared spontaneous combustion image, and inputting the coal pile spontaneous combustion common flame image dataset into a convolutional neural network model A for training; copying the parameters of the trained convolutional neural network model A to another convolutional neural network model B, designing an identification neural network D, and training the identification neural network D by taking a characteristic diagram generated by the convolutional neural network model A or B as a training set; and finally, updating the parameters of the convolutional neural network model B by using the identification result of the identification neural network D to obtain the convolutional neural network model B after the countermeasure training is finished. The invention can effectively improve the detection rate of spontaneous combustion of the coal pile; and can automatically detect the spontaneous combustion of the coal pile all weather under complex conditions; the method has the advantages of simple process, small calculated amount and high reliability.
Description
Technical Field
The invention relates to a coal yard spontaneous combustion detection method based on infrared imaging, in particular to a spontaneous combustion detection transfer learning algorithm utilizing a confrontation neural network.
Background
In recent years, with the continuous development of economy and society of China, the demand of coal-fired power plants, steel plants and chemical plants for coal is increased dramatically. When a large amount of coal is stored in the open air, the coal is subjected to the action of wind, rain, sunshine and oxygen in the air for a long time, the oxidation of the coal in a coal yard is intensified, the temperature is increased, a large amount of heat is generated, and further the spontaneous combustion of the coal is initiated. Coal piles are naturally a common major safety hazard. In open-pit coal mines and coal yards, a large amount of spare coal is stored, spontaneous combustion of the stored coal often occurs due to weather or artificial reasons, and in addition, the coal in a smoldering state can cause violent combustion and explosion accidents in the transportation process. Because of remote coal sites and high labor cost, an automatic natural coal site detection method is urgently needed in production practice. Among them, spontaneous combustion detection based on infrared imaging is a feasible method. However, the spontaneous combustion is various and complex in environment, and the marked samples are few, so that the existing method is not high in performance. Coal spontaneous combustion directly loses a large amount of coal resources, endangers the safety production of factories, particularly directly pollutes the atmosphere through the transmission of a large amount of harmful gases, deteriorates the regional ecological environment, reduces the quality of the living environment of local residents, and has great potential danger for the development of regional social economy and the survival and health of human beings. Therefore, the research on the automatic detection of the spontaneous combustion of the coal yard is a practical problem to be solved by urgent need of safety production.
Existing methods for coal spontaneous combustion detection can be broadly divided into traditional methods based on sensors and advanced methods based on computer vision. The former detects spontaneous combustion by mounting a temperature-sensitive, smoke-sensitive, light-sensitive or gas-measuring sensor. But the temperature-sensing sensor is not sensitive to smoldering fire in the coal pile and can generate false alarm when the ambient temperature is higher; the smoke-sensitive sensor has low sensitivity in an open environment; the photosensitive sensor is easily interfered by strong light and high-power light sources; the gas measurement type sensor has low detection sensitivity in an open place. Therefore, the conventional sensor method has not been suitable for coal pile spontaneous combustion detection. The thermal infrared imager is used for detecting the surface temperature distribution of the coal yard to achieve the purposes of real-time monitoring and coal pile spontaneous combustion prevention. The infrared thermal imaging has unique advantages, such as being used for detection at night and in severe environment, effectively finding out smoldering fire and accurately judging the fire place, thereby realizing the purpose of preventing fire. However, spontaneous combustion of coal is a slow process with no obvious signs, and it is difficult for people to observe infrared images continuously for 24 hours. Therefore, a method for automatically checking spontaneous combustion from infrared images based on computer vision is in need of development.
The spontaneous combustion detection method based on computer vision needs to have marked images for training an algorithm model, however, infrared images marked with spontaneous combustion are very few in reality, and manual marking of data is a time-consuming and expensive operation, and an effective mode for solving the problem is not available so far. Transfer learning refers to a learning process of applying a model learned in an old field to a new field by using data, tasks, or similarities between models. In order to solve the problem that infrared images have rare spontaneous combustion marks, the idea of transfer learning is adopted to find out the similarity between the infrared spontaneous combustion images and the common flame images and realize the knowledge transfer of spontaneous combustion detection.
Disclosure of Invention
The invention aims to design a coal yard spontaneous combustion detection method based on transfer learning, solve the problem of rare labeled infrared spontaneous combustion images and realize automatic detection of coal yard spontaneous combustion.
In order to solve the problem of rare labeled infrared spontaneous combustion images, the invention provides a migration learning-based automatic detection method for spontaneous combustion of a coal pile, which comprises the following steps:
s1: and acquiring a marked common flame image data set of the spontaneous combustion of the coal pile on the Internet as a training set of the convolutional neural network model A.
S2: and acquiring an unmarked infrared spontaneous combustion image of the coal pile.
S3: the convolutional neural network model A is divided into five layers, signals are transmitted between the layers by means of a feature map, common flame images of spontaneous combustion of the coal pile are input, the detection result of whether flames exist in the coal pile is output, and a common flame image data set of spontaneous combustion of the coal pile is input into the convolutional neural network model A for training to obtain the trained convolutional neural network model A.
S4: copying the parameters of the trained convolutional neural network model A into another convolutional neural network model B, wherein the convolutional neural network models A and B have the same structure; the input of the convolutional neural network model B is an unmarked coal pile infrared spontaneous combustion image, and the output of the convolutional neural network model B is expected to be information such as whether the coal pile infrared image has spontaneous combustion form and position.
S5: designing and identifying a neural network D, wherein the input of the neural network D is a characteristic diagram generated by a convolutional neural network model A or B, the network A generates a characteristic diagram of a common flame image of coal pile spontaneous combustion, the network B generates a characteristic diagram of a coal pile infrared spontaneous combustion image, the output of the neural network D is used for judging whether the characteristic diagram is the common flame image from the coal pile spontaneous combustion or the coal pile infrared spontaneous combustion image, and the characteristic diagram generated by the convolutional neural network model A or B is used as a training set to train the neural network D so as to identify the coal pile infrared spontaneous combustion image characteristic diagram and the common flame image characteristic diagram from the coal pile spontaneous combustion.
S6: updating parameters of the convolutional neural network model B by using the identification result of the identification neural network D, and further training the convolutional neural network model B, so that the characteristic diagram of the identification neural network D, which cannot identify the convolutional neural network model A and the convolutional neural network model B, is a coal pile infrared spontaneous combustion image characteristic diagram or a common flame image characteristic diagram of coal pile spontaneous combustion, and thus, the confrontation training is completed; and obtaining a convolutional neural network model B after the countermeasure training is finished, inputting the unmarked coal pile infrared image into the convolutional neural network model B, and outputting information such as the form, position and the like of spontaneous combustion in the coal pile infrared image by the convolutional neural network model B.
Further, when the convolutional neural network model A is trained, a BP algorithm is adopted, errors are reversely propagated to an input end from an output end, and network parameters are updated by a gradient descent method.
Due to the adoption of the technical scheme, compared with the prior art, the invention has the following advantages and positive effects:
the existing detection method aiming at the nature of the coal yard is mostly based on a sensor, and has the problems of insensitivity to smoldering fire, false alarm when the ambient temperature is higher, low sensitivity in an open environment and easy strong light interference on the sensor. The existing spontaneous combustion detection method based on infrared imaging has the advantages that the detection effect depends on the experience of inspectors, time and labor are consumed, and the detection quality is difficult to guarantee. The spontaneous combustion detection method based on transfer learning is high in sensitivity and does not depend on the experience of a detector, and the detection rate of spontaneous combustion of the coal pile can be effectively improved; and can automatically detect the spontaneous combustion of the coal pile all weather under complex conditions; the method has the advantages of simple process, small calculated amount and high reliability.
Drawings
FIG. 1 is a schematic diagram of a transfer learning process;
FIG. 2 is a schematic diagram of a convolutional neural network A or B;
FIG. 3 is a schematic diagram of a structure of an identified neural network D;
fig. 4 is a schematic diagram of a transitional confrontation learning process.
Concrete implementation method
The following describes embodiments of the present invention in further detail with reference to the accompanying drawings.
As shown in fig. 1, the method for automatically detecting spontaneous combustion of a coal pile based on transfer learning provided by the invention comprises the following steps:
s1: and acquiring more than 5000 marked common flame image data sets of the spontaneous combustion of the coal pile on the Internet as a training set of the convolutional neural network model A.
S2: the method comprises the steps of obtaining unmarked infrared spontaneous combustion images of the coal pile, wherein the infrared spontaneous combustion images of the coal pile are few in source and can be obtained by about 100.
S3: the convolutional neural network model A is divided into five layers, signals are transmitted between the layers by means of a feature map, common flame images of spontaneous combustion of the coal pile are input, the detection result of whether flames exist in the coal pile is output, and a common flame image data set of spontaneous combustion of the coal pile is input into the convolutional neural network model A for training to obtain a trained convolutional neural network model A; and during training of the convolutional neural network model A, the error is reversely propagated from the output end to the input end by adopting a BP algorithm, and the network parameters are updated by using a gradient descent method.
S4: copying the parameters of the trained convolutional neural network model A into another convolutional neural network model B, wherein the convolutional neural network models A and B have the same structure; the input of the convolutional neural network model B is an unmarked coal pile infrared spontaneous combustion image, and the output of the convolutional neural network model B is expected to be information such as whether the coal pile infrared image has spontaneous combustion form and position.
S5: designing and identifying a neural network D, wherein the input of the neural network D is a characteristic diagram generated by a convolutional neural network model A or B, the network A generates a characteristic diagram of a common flame image of coal pile spontaneous combustion, the network B generates a characteristic diagram of a coal pile infrared spontaneous combustion image, the output of the neural network D is used for judging whether the characteristic diagram is the common flame image from the coal pile spontaneous combustion or the coal pile infrared spontaneous combustion image, and the characteristic diagram generated by the convolutional neural network model A or B is used as a training set to train the neural network D so as to identify the coal pile infrared spontaneous combustion image characteristic diagram and the common flame image characteristic diagram from the coal pile spontaneous combustion.
S6: updating parameters of the convolutional neural network model B by using the identification result of the identification neural network D, and further training the convolutional neural network model B, so that the characteristic diagram of the identification neural network D, which cannot identify the convolutional neural network model A and the convolutional neural network model B, is a coal pile infrared spontaneous combustion image characteristic diagram or a common flame image characteristic diagram of coal pile spontaneous combustion, and thus, the confrontation training is completed; and obtaining a convolutional neural network model B after the countermeasure training is finished, inputting the unmarked coal pile infrared image into the convolutional neural network model B, and outputting information such as the form, position and the like of spontaneous combustion in the coal pile infrared image by the convolutional neural network model B.
The invention adopts a transfer learning method based on an antagonistic neural network. A common flame image data set of marked coal pile spontaneous combustion existing in a large amount on the Internet is collected firstly. And collecting the unmarked infrared spontaneous combustion images of the coal pile. And designing a neural network model A, and training the model A by using a common flame data set of spontaneous combustion of the coal pile so as to enable the model A to detect flame. And copying the parameters of the model A into another neural network model B. And designing an identification neural network D, and training the identification neural network D to distinguish the difference between the infrared spontaneous combustion image characteristic diagram of the coal pile and the common image characteristic diagram of the spontaneous combustion of the coal pile. And training the neural network model B by using feedback of the identification neural network D, so that the identification neural network D cannot identify whether the characteristic diagram of the convolutional neural network model A and the characteristic diagram of the convolutional neural network model B are the characteristic diagram of the coal pile infrared spontaneous combustion image or the characteristic diagram of the coal pile spontaneous combustion common flame image, thereby completing the countermeasure training, obtaining the neural network model B after the countermeasure training is completed, and automatically detecting the spontaneous combustion in the infrared image.
Fig. 2 is a schematic diagram of a network structure of the convolutional neural network a or B. The convolutional neural network A or B is divided into five layers, is designed according to the characteristics of an image detection task, the input of the convolutional neural network is an image, the output of the convolutional neural network is a detection result of whether flame or spontaneous combustion exists, and signals are transmitted between the layers by means of a feature map. It learns, automatically generates, corrects and highly summarizes the best network parameters from the training data. During training, the BP algorithm is adopted, errors are reversely propagated to the input end from the output end, and the network parameters are updated by a gradient descent method. Networks a and B share the same network structure and differ only in that the input images are one of the images of the general flames of spontaneous combustion of the coal pile and one of the images of the infrared of the coal pile.
Fig. 3 is a schematic structural diagram of the neural network D. The input of the identification neural network D is a characteristic diagram generated by the convolutional neural network A or B, and the output is the judgment of the characteristic diagram, namely the characteristic diagram is from a common flame image of spontaneous combustion of a coal pile or an infrared image of the coal pile. The discriminative neural network D provides the basis for the antagonistic training.
Fig. 4 is a schematic diagram of the transitional confrontation learning process. The convolutional neural network B and the identification neural network D have a confrontation game relationship: the confrontation training uses the recognition result of the recognition neural network D to update the parameters of the convolution neural network B until the recognition neural network D cannot judge whether the sample is from the infrared image or the common image, and then the confrontation training is completed. The result of the countertraining is that the convolutional neural network B continuously learns the characteristics of the infrared spontaneous combustion images of the coal pile, so that the identification neural network D cannot distinguish the characteristic diagram of the convolutional neural network B, and the convolutional neural network B has spontaneous combustion detection capability on the infrared images.
Claims (2)
1. A coal pile spontaneous combustion automatic detection method based on transfer learning is characterized by comprising the following steps:
s1: and acquiring a marked common flame image data set of the spontaneous combustion of the coal pile on the Internet as a training set of the convolutional neural network model A.
S2: and acquiring an unmarked infrared spontaneous combustion image of the coal pile.
S3: the convolutional neural network model A is divided into five layers, signals are transmitted between the layers by means of a feature map, common flame images of spontaneous combustion of the coal pile are input, the detection result of whether flames exist in the coal pile is output, and a common flame image data set of spontaneous combustion of the coal pile is input into the convolutional neural network model A for training to obtain the trained convolutional neural network model A.
S4: copying the parameters of the trained convolutional neural network model A into another convolutional neural network model B, wherein the convolutional neural network models A and B have the same structure; the input of the convolutional neural network model B is an unmarked coal pile infrared spontaneous combustion image, and the output of the convolutional neural network model B is expected to be information such as whether the coal pile infrared image has spontaneous combustion form and position.
S5: designing and identifying a neural network D, wherein the input of the neural network D is a characteristic diagram generated by a convolutional neural network model A or B, the network A generates a characteristic diagram of a common flame image of coal pile spontaneous combustion, the network B generates a characteristic diagram of a coal pile infrared spontaneous combustion image, the output of the neural network D is used for judging whether the characteristic diagram is the common flame image from the coal pile spontaneous combustion or the coal pile infrared spontaneous combustion image, and the characteristic diagram generated by the convolutional neural network model A or B is used as a training set to train the neural network D so as to identify the coal pile infrared spontaneous combustion image characteristic diagram and the common flame image characteristic diagram from the coal pile spontaneous combustion.
S6: updating parameters of the convolutional neural network model B by using the identification result of the identification neural network D, and further training the convolutional neural network model B, so that the characteristic diagram of the identification neural network D, which cannot identify the convolutional neural network model A and the convolutional neural network model B, is a coal pile infrared spontaneous combustion image characteristic diagram or a common flame image characteristic diagram of coal pile spontaneous combustion, and thus, the confrontation training is completed; and obtaining a convolutional neural network model B after the countermeasure training is finished, inputting the unmarked coal pile infrared image into the convolutional neural network model B, and outputting information such as the form, position and the like of spontaneous combustion in the coal pile infrared image by the convolutional neural network model B.
2. The method for automatically detecting the spontaneous combustion of the coal pile based on the transfer learning of claim 1, wherein during training of the convolutional neural network model A, a BP algorithm is adopted, errors are reversely propagated from an output end to an input end, and network parameters are updated by a gradient descent method.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911007182.4A CN110765937A (en) | 2019-10-22 | 2019-10-22 | Coal yard spontaneous combustion detection method based on transfer learning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911007182.4A CN110765937A (en) | 2019-10-22 | 2019-10-22 | Coal yard spontaneous combustion detection method based on transfer learning |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110765937A true CN110765937A (en) | 2020-02-07 |
Family
ID=69332470
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911007182.4A Pending CN110765937A (en) | 2019-10-22 | 2019-10-22 | Coal yard spontaneous combustion detection method based on transfer learning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110765937A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115083123A (en) * | 2022-05-17 | 2022-09-20 | 中国矿业大学 | Mine coal spontaneous combustion intelligent grading early warning method taking measured data as drive |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP0984413A2 (en) * | 1998-09-01 | 2000-03-08 | Deutsches Zentrum für Luft- und Raumfahrt e.V. | Method and system for automatic forest fire recognition |
| CN108257347A (en) * | 2018-01-10 | 2018-07-06 | 安徽大学 | A kind of flame image sequence sorting technique and device using convolutional neural networks |
| AU2018101317A4 (en) * | 2018-09-07 | 2018-10-11 | Chen, Guoyi Mr | A Deep Learning Based System for Animal Species Classification |
| CN108681752A (en) * | 2018-05-28 | 2018-10-19 | 电子科技大学 | A kind of image scene mask method based on deep learning |
| CN108985192A (en) * | 2018-06-29 | 2018-12-11 | 东南大学 | A kind of video smoke recognition methods based on multitask depth convolutional neural networks |
| CN109376747A (en) * | 2018-12-11 | 2019-02-22 | 北京工业大学 | A kind of video flame detecting method based on double-current convolutional neural networks |
| CN109376695A (en) * | 2018-11-26 | 2019-02-22 | 北京工业大学 | A kind of smog detection method based on depth hybrid neural networks |
| CN109460708A (en) * | 2018-10-09 | 2019-03-12 | 东南大学 | A kind of Forest fire image sample generating method based on generation confrontation network |
| CN109858516A (en) * | 2018-12-24 | 2019-06-07 | 武汉工程大学 | A kind of fire and smog prediction technique, system and medium based on transfer learning |
| CN109903507A (en) * | 2019-03-04 | 2019-06-18 | 上海海事大学 | A kind of fire disaster intelligent monitor system and method based on deep learning |
-
2019
- 2019-10-22 CN CN201911007182.4A patent/CN110765937A/en active Pending
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP0984413A2 (en) * | 1998-09-01 | 2000-03-08 | Deutsches Zentrum für Luft- und Raumfahrt e.V. | Method and system for automatic forest fire recognition |
| CN108257347A (en) * | 2018-01-10 | 2018-07-06 | 安徽大学 | A kind of flame image sequence sorting technique and device using convolutional neural networks |
| CN108681752A (en) * | 2018-05-28 | 2018-10-19 | 电子科技大学 | A kind of image scene mask method based on deep learning |
| CN108985192A (en) * | 2018-06-29 | 2018-12-11 | 东南大学 | A kind of video smoke recognition methods based on multitask depth convolutional neural networks |
| AU2018101317A4 (en) * | 2018-09-07 | 2018-10-11 | Chen, Guoyi Mr | A Deep Learning Based System for Animal Species Classification |
| CN109460708A (en) * | 2018-10-09 | 2019-03-12 | 东南大学 | A kind of Forest fire image sample generating method based on generation confrontation network |
| CN109376695A (en) * | 2018-11-26 | 2019-02-22 | 北京工业大学 | A kind of smog detection method based on depth hybrid neural networks |
| CN109376747A (en) * | 2018-12-11 | 2019-02-22 | 北京工业大学 | A kind of video flame detecting method based on double-current convolutional neural networks |
| CN109858516A (en) * | 2018-12-24 | 2019-06-07 | 武汉工程大学 | A kind of fire and smog prediction technique, system and medium based on transfer learning |
| CN109903507A (en) * | 2019-03-04 | 2019-06-18 | 上海海事大学 | A kind of fire disaster intelligent monitor system and method based on deep learning |
Non-Patent Citations (1)
| Title |
|---|
| 唐贤伦: "基于条件深度卷积生成对抗网络的图像识别方法" * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115083123A (en) * | 2022-05-17 | 2022-09-20 | 中国矿业大学 | Mine coal spontaneous combustion intelligent grading early warning method taking measured data as drive |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| JP6961806B2 (en) | Thermographic image processing with neural networks to identify under-insulation corrosion (CUI) | |
| CN101251942B (en) | Underground space fire intelligent detection early alarming and forecasting method and apparatus | |
| CN106864739A (en) | A kind of six rotor flying robots for underground pipe gallery detection | |
| CN102737247A (en) | Identification system of smoke intensity image of tail gas of diesel vehicle | |
| CN108288055A (en) | Block of bow collector of electric locomotive based on depth network and placement test and arc method for measuring | |
| CN102208018A (en) | Method for recognizing fire disaster of power transmission line based on video variance analysis | |
| CN112002095A (en) | Fire early warning method in mine tunnel | |
| CN112668873B (en) | Mine safety situation analysis and prediction early warning method | |
| CN110765937A (en) | Coal yard spontaneous combustion detection method based on transfer learning | |
| CN111783945B (en) | Long oil pipeline leakage real-time detection system based on transfer learning LSTM | |
| CN116882758A (en) | Dam safety monitoring physical examination list generation method | |
| CN108446838A (en) | A kind of bridge safety supervision system based on big data | |
| CN113283276A (en) | Linkage thermal imaging self-learning fire point detection method and system | |
| JP6954242B2 (en) | How to investigate the installation location of the stationary gas detector | |
| CN104200599B (en) | Bus multisensor fire detection method | |
| CN108520200B (en) | A kind of effective coal-mine fire detection system | |
| Xie et al. | A new anomaly detector of flare pilot based on siamese network | |
| CN112784706B (en) | Oil testing test operation area safety control method based on image intelligent identification | |
| Sun et al. | Environmental risk assessment system for phosphogypsum tailing dams | |
| KR20180080579A (en) | Structure visual inspection device for condition assessment of sewer pipe and sewer culvert | |
| CN113284103A (en) | Substation equipment defect online detection system based on space transformation fast R-CNN model | |
| Yang et al. | Prediction of pollutant diffusion trend using CNN-GRU hybrid network in IoT air monitoring system | |
| CN116778192B (en) | Fire safety early warning system based on air-ground equipment cooperation | |
| CN106643885A (en) | Building safety monitoring system and monitoring method | |
| CN117310093A (en) | Atmospheric environment monitoring system based on Internet of things |
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 | ||
| CB03 | Change of inventor or designer information |
Inventor after: Song Xiaoling Inventor after: Ma Longhua Inventor after: Wen Gang Inventor after: Liu Cong Inventor after: Yao Jiaqing Inventor after: Xu Ming Inventor after: Geng Runhua Inventor before: Song Xiaoling Inventor before: Ma Longhua Inventor before: Wen Gang Inventor before: Liu Cong Inventor before: Yao Jiaqing Inventor before: Xu Ming Inventor before: Geng Runhua |
|
| CB03 | Change of inventor or designer information | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20200207 |
|
| RJ01 | Rejection of invention patent application after publication |