CN112115876A - Water-soluble method experimental process identification method based on 3D convolutional neural network - Google Patents
Water-soluble method experimental process identification method based on 3D convolutional neural network Download PDFInfo
- Publication number
- CN112115876A CN112115876A CN202010995196.8A CN202010995196A CN112115876A CN 112115876 A CN112115876 A CN 112115876A CN 202010995196 A CN202010995196 A CN 202010995196A CN 112115876 A CN112115876 A CN 112115876A
- Authority
- CN
- China
- Prior art keywords
- neural network
- convolutional neural
- water
- video clip
- soluble
- 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
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/40—Scenes; Scene-specific elements in video content
-
- 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
- G06N3/045—Combinations of networks
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Biomedical Technology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Multimedia (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a water-soluble method experiment process identification method based on a 3D convolutional neural network, which comprises the following steps: s1: acquiring a video image of an oil discharge area of a gas station by using a monitoring system to obtain a real-time video; s2: dividing the real-time video into a plurality of video segments; processing the video clip to obtain a processed video clip; s3: constructing a 3D convolutional neural network, and training the processed video clip by using the 3D convolutional neural network to obtain a trained 3D convolutional neural network; s4: and (4) carrying out recognition analysis on the processed video clip by using the trained 3D convolutional neural network, and judging whether the processed video clip contains a water-soluble experimental process clip. According to the invention, the original monitoring system of the gas station is upgraded and energized by combining the mode of acquiring the image of the oil unloading area by the camera and the intelligent analysis algorithm, so that the original manual inspection is replaced, and the labor cost is reduced; meanwhile, an oil enterprise manager can effectively supervise the oil unloading operation of the oil enterprise end in real time.
Description
Technical Field
The invention belongs to the field of artificial intelligence and graphic image technology in the field of computers, and particularly relates to a water-soluble method experimental process identification method based on a 3D convolutional neural network.
Background
And (3) installing a camera in the oil unloading area at the beginning of the establishment of the gas station according to the security protection requirement, and inspecting the oil unloading safety operation, the oil unloading quality detection operation and the safety operation of the oil unloading area in a camera monitoring mode. The mode has great defects, and safety early warning information and various risk information cannot be timely, quickly and effectively transmitted to a manager.
The water-soluble method experiment process in the oil unloading process is an important link of oil quality detection, and whether the current oil quality is qualified or not is analyzed through an experiment method. The existing processing mode only monitors each oil unloading process through monitoring, and safety events cannot be transmitted to oil enterprise managers at the first time; on the other hand, personnel management of the gas station site cannot be effectively implemented.
In the prior art, a method of human intervention is completely adopted, whether a water-soluble method experimental process exists in the oil unloading process is monitored through a camera, and an objective, non-manual and accurate method for intelligently monitoring whether the water-soluble method experimental process exists in the oil unloading process is not provided. This totally manual method presents the following problems in 3: 1. the human cost is high, needs the staff to carry out real time monitoring. 2. The risk of error is high and manual inspection always leads to errors due to occasional fatigue or inadvertence. 3. The superior leader basically cannot perform supervision and management.
Disclosure of Invention
The invention aims to solve the technical problem that the prior art completely adopts a human intervention method, monitors whether a water-soluble method experimental process exists in an oil unloading process through a camera, and does not have a method for objectively, non-manually and accurately and intelligently monitoring whether the water-soluble method experimental process exists in the oil unloading process, so that the water-soluble method experimental process identification method based on the 3D convolutional neural network is provided, and the problems in the background technology are solved.
The invention is realized by the following technical scheme:
a water-soluble method experiment process identification method based on a 3D convolutional neural network comprises the following steps:
s1: acquiring a video image of an oil discharge area of a gas station by using a monitoring system to obtain a real-time video;
s2: dividing a real-time video into a plurality of video segments, wherein each video segment consists of continuous L-frame pictures; processing the video clip to obtain a processed video clip;
s3: constructing a 3D convolutional neural network, and training the processed video clip by using the 3D convolutional neural network to obtain a trained 3D convolutional neural network;
s4: carrying out recognition analysis on the processed video clip by using the trained 3D convolutional neural network, and judging whether the processed video clip contains a water-soluble experimental process clip or not; if yes, ending; if not, return is made to step S1.
Further, a water-soluble method experiment process identification method based on a 3D convolutional neural network, step S2 specifically includes the following steps:
s21: taking one frame per second from the real-time video obtained in the step S1, and cumulatively taking 9 frames to obtain 9 pictures, namely a video clip;
s22: converting the 9 pictures into 9 gray-scale images, and obtaining a video clip containing the 9 gray-scale images according to the 9 gray-scale images; and the video clip containing the 9 gray level images is the processed video clip.
Further, a water-soluble method experiment process identification method based on a 3D convolutional neural network, where the step S3 specifically includes:
s31: collecting 4000 sections of processed video clips as a training set, and collecting 400 sections of processed video clips as a verification set;
s32: defining a 3D convolution kernel, the convolution kernel size being 3 x 9 x n;
s33: building a 3D convolutional neural network, and inputting 256 × 9 from the input end of the 3D convolutional neural network; outputting 1 x 2 from the output end of the 3D convolutional neural network, wherein the output is an instant output;
s34: defining a loss function, and calculating the formula as follows:
where y represents the network prediction value,
represents the actual tag value;
s35: training the training set by using a gradient descent method through a loss function to optimize a 3D convolutional neural network;
s36: and (3) verifying the verification set by using the 3D convolutional neural network, and finishing the training of the 3D convolutional neural network when the verification precision is more than 95% and the 3D convolutional neural network is not lifted, so that the trained 3D convolutional neural network is obtained.
Further, in step S31, in the training set and the verification set, the ratio of the data amount of the water soluble experiment process fragments to the data amount of the water soluble experiment process fragments not included is 1: 3.
Further, a water-soluble method experiment process identification method based on a 3D convolutional neural network, step S3 further includes: and calculating and analyzing the probability that the processed video clip contains the water-soluble method experimental process and the probability that the processed video clip does not contain the water-soluble method experimental process through the trained 3D convolutional neural network.
Further, the water-soluble method experiment process identification method based on the 3D convolutional neural network is characterized in that the monitoring system is a high-level camera arranged in a gas station oil unloading area.
The invention provides a water-soluble method experimental process identification method based on a 3D convolutional neural network, which upgrades and enables an original monitoring system of a gas station by combining a mode of collecting images of an oil discharge area by a camera and an intelligent analysis algorithm, and replaces the original manual inspection. The method enables the monitoring of the oil unloading operation to be intelligent, can reduce the labor cost to a great extent, and can ensure timely, objective and accurate analysis and timely early warning because of the operation of the machine, thereby reducing the risk of accidents.
Compared with the prior art, the invention has the following advantages and beneficial effects:
according to the invention, the original monitoring system of the gas station is upgraded and energized by combining the mode of acquiring the image of the oil unloading area by the camera and the intelligent analysis algorithm, so that the original manual inspection is replaced, and the labor cost is reduced; meanwhile, an oil enterprise manager can effectively supervise the oil unloading operation of the oil enterprise end in real time.
Drawings
The accompanying drawings, which are included to provide a further understanding of the embodiments of the invention and are incorporated in and constitute a part of this application, illustrate embodiment(s) of the invention and together with the description serve to explain the principles of the invention. In the drawings:
FIG. 1 is a flow chart of identification of an experimental process of a water dissolving method.
FIG. 2 is a schematic diagram of a high-phase camera.
Fig. 3 is a convolution kernel.
Fig. 4 is a 3D convolutional neural network.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is further described in detail below with reference to examples and accompanying drawings, and the exemplary embodiments and descriptions thereof are only used for explaining the present invention and are not meant to limit the present invention.
Examples
As shown in fig. 1, the method for identifying the water-soluble experimental process based on the 3D convolutional neural network of the present invention comprises the following steps:
s1: acquiring a video image of an oil discharge area of a gas station by using a monitoring system to obtain a real-time video;
s2: dividing a real-time video into a plurality of video segments, wherein each video segment consists of continuous L-frame pictures; processing the video clip to obtain a processed video clip;
s3: constructing a 3D convolutional neural network, and training the processed video clip by using the 3D convolutional neural network to obtain a trained 3D convolutional neural network;
s4: carrying out recognition analysis on the processed video clip by using the trained 3D convolutional neural network, and judging whether the processed video clip contains a water-soluble experimental process clip or not; if yes, ending; if not, return is made to step S1.
The step S2 specifically includes the following steps:
s21: taking one frame per second from the real-time video obtained in the step S1, and cumulatively taking 9 frames to obtain 9 pictures, namely a video clip;
s22: converting the 9 pictures into 9 gray-scale images, and obtaining a video clip containing the 9 gray-scale images according to the 9 gray-scale images; and the video clip containing the 9 gray level images is the processed video clip.
The step S3 specifically includes:
s31: collecting 4000 sections of processed video clips as a training set, and collecting 400 sections of processed video clips as a verification set;
s32: as shown in fig. 3, a 3D convolution kernel is defined, the convolution kernel size being 3 x 9 x n;
s33: as shown in fig. 4, a 3D convolutional neural network is constructed, and 256 × 9 is input from the input end of the 3D convolutional neural network; outputting 1 x 2 from the output end of the 3D convolutional neural network, wherein the output is an instant output;
s34: defining a loss function, and calculating the formula as follows:
where y represents the network prediction value,
represents the actual tag value;
s35: training the training set by using a gradient descent method through a loss function to optimize a 3D convolutional neural network;
s36: and (3) verifying the verification set by using the 3D convolutional neural network, and finishing the training of the 3D convolutional neural network when the verification precision is more than 95% and the 3D convolutional neural network is not lifted, so that the trained 3D convolutional neural network is obtained.
In step S31, in the training set and the verification set, the ratio of the data of the test process fragments containing the water-soluble method to the data of the test process fragments not containing the water-soluble method is 1: 3.
Step S3 further includes: and calculating and analyzing the probability that the processed video clip contains the water-soluble method experimental process and the probability that the processed video clip does not contain the water-soluble method experimental process through the trained 3D convolutional neural network.
As shown in FIG. 2, the monitoring system is a high-level camera arranged in the oil unloading area of the oil filling station, and the height of the high-level camera from the oil unloading area is 7 meters.
The above-mentioned embodiments are intended to illustrate the objects, technical solutions and advantages of the present invention in further detail, and it should be understood that the above-mentioned embodiments are merely exemplary embodiments of the present invention, and are not intended to limit the scope of the present invention, and any modifications, equivalent substitutions, improvements and the like made within the spirit and principle of the present invention should be included in the scope of the present invention.
Claims (6)
1. A water-soluble method experiment process identification method based on a 3D convolutional neural network is characterized by comprising the following steps:
s1: acquiring a video image of an oil discharge area of a gas station by using a monitoring system to obtain a real-time video;
s2: dividing a real-time video into a plurality of video segments, wherein each video segment consists of continuous L-frame pictures; processing the video clip to obtain a processed video clip;
s3: constructing a 3D convolutional neural network, and training the processed video clip by using the 3D convolutional neural network to obtain a trained 3D convolutional neural network;
s4: carrying out recognition analysis on the processed video clip by using the trained 3D convolutional neural network, and judging whether the processed video clip contains a water-soluble experimental process clip or not; if yes, ending; if not, return is made to step S1.
2. The identification method for the water-soluble experimental process based on the 3D convolutional neural network as claimed in claim 1, wherein the step S2 specifically comprises the following steps:
s21: taking one frame per second from the real-time video obtained in the step S1, and cumulatively taking 9 frames to obtain 9 pictures, namely a video clip;
s22: converting the 9 pictures into 9 gray-scale images, and obtaining a video clip containing the 9 gray-scale images according to the 9 gray-scale images; and the video clip containing the 9 gray level images is the processed video clip.
3. The method for identifying the water-soluble experimental process based on the 3D convolutional neural network as claimed in claim 2, wherein the step S3 specifically comprises:
s31: collecting 4000 sections of processed video clips as a training set, and collecting 400 sections of processed video clips as a verification set;
s32: defining a 3D convolution kernel, the convolution kernel size being 3 x 9 x n;
s33: building a 3D convolutional neural network, and inputting 256 × 9 from the input end of the 3D convolutional neural network; outputting 1 x 2 from the output end of the 3D convolutional neural network, wherein the output is an instant output;
s34: defining a loss function, and calculating the formula as follows:
where y represents the network prediction value,
represents the actual tag value;
s35: training the training set by using a gradient descent method through a loss function to optimize a 3D convolutional neural network;
s36: and (3) verifying the verification set by using the 3D convolutional neural network, and finishing the training of the 3D convolutional neural network when the verification precision is more than 95% and the 3D convolutional neural network is not lifted, so that the trained 3D convolutional neural network is obtained.
4. The method for identifying the water soluble experimental process based on the 3D convolutional neural network as claimed in claim 3, wherein in the step S31, the ratio of the data of the water soluble experimental process fragments to the data of the water soluble experimental process fragments is 1: 3.
5. The identification method for the water-soluble experimental process based on the 3D convolutional neural network as claimed in claim 3, wherein the step S3 further comprises: and calculating and analyzing the probability that the processed video clip contains the water-soluble method experimental process and the probability that the processed video clip does not contain the water-soluble method experimental process through the trained 3D convolutional neural network.
6. The method for identifying the water-soluble experimental process based on the 3D convolutional neural network as claimed in claim 1, wherein the monitoring system is an overhead camera arranged in a fuel unloading area of a fuel station.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010995196.8A CN112115876A (en) | 2020-09-21 | 2020-09-21 | Water-soluble method experimental process identification method based on 3D convolutional neural network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010995196.8A CN112115876A (en) | 2020-09-21 | 2020-09-21 | Water-soluble method experimental process identification method based on 3D convolutional neural network |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112115876A true CN112115876A (en) | 2020-12-22 |
Family
ID=73799964
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010995196.8A Pending CN112115876A (en) | 2020-09-21 | 2020-09-21 | Water-soluble method experimental process identification method based on 3D convolutional neural network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112115876A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2023123926A1 (en) * | 2021-12-28 | 2023-07-06 | 苏州浪潮智能科技有限公司 | Artificial intelligence task processing method and apparatus, electronic device, and readable storage medium |
Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104966104A (en) * | 2015-06-30 | 2015-10-07 | 孙建德 | Three-dimensional convolutional neural network based video classifying method |
| US10089556B1 (en) * | 2017-06-12 | 2018-10-02 | Konica Minolta Laboratory U.S.A., Inc. | Self-attention deep neural network for action recognition in surveillance videos |
| CN108734095A (en) * | 2018-04-10 | 2018-11-02 | 南京航空航天大学 | A kind of motion detection method based on 3D convolutional neural networks |
| CN109255284A (en) * | 2018-07-10 | 2019-01-22 | 西安理工大学 | A kind of Activity recognition method of the 3D convolutional neural networks based on motion profile |
| CN109271938A (en) * | 2018-09-19 | 2019-01-25 | 上海鸢安智能科技有限公司 | A kind of gas station's emptying Safety Monitoring Control method based on intelligent video analysis technology |
| CN109829443A (en) * | 2019-02-23 | 2019-05-31 | 重庆邮电大学 | Video behavior recognition methods based on image enhancement Yu 3D convolutional neural networks |
| CN109947990A (en) * | 2019-03-29 | 2019-06-28 | 北京奇艺世纪科技有限公司 | A kind of wonderful detection method and system |
| CN110070002A (en) * | 2019-03-29 | 2019-07-30 | 上海理工大学 | A kind of Activity recognition method based on 3D convolutional neural networks |
| US20190244366A1 (en) * | 2017-09-07 | 2019-08-08 | Comcast Cable Communications, Llc | Relevant Motion Detection in Video |
| CN110456723A (en) * | 2019-08-15 | 2019-11-15 | 成都睿晓科技有限公司 | A kind of emptying area, gas station security management and control system based on deep learning |
| CN110555368A (en) * | 2019-06-28 | 2019-12-10 | 西安理工大学 | Fall-down behavior identification method based on three-dimensional convolutional neural network |
| CN111126115A (en) * | 2018-11-01 | 2020-05-08 | 顺丰科技有限公司 | Violence sorting behavior identification method and device |
-
2020
- 2020-09-21 CN CN202010995196.8A patent/CN112115876A/en active Pending
Patent Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104966104A (en) * | 2015-06-30 | 2015-10-07 | 孙建德 | Three-dimensional convolutional neural network based video classifying method |
| US10089556B1 (en) * | 2017-06-12 | 2018-10-02 | Konica Minolta Laboratory U.S.A., Inc. | Self-attention deep neural network for action recognition in surveillance videos |
| US20190244366A1 (en) * | 2017-09-07 | 2019-08-08 | Comcast Cable Communications, Llc | Relevant Motion Detection in Video |
| CN108734095A (en) * | 2018-04-10 | 2018-11-02 | 南京航空航天大学 | A kind of motion detection method based on 3D convolutional neural networks |
| CN109255284A (en) * | 2018-07-10 | 2019-01-22 | 西安理工大学 | A kind of Activity recognition method of the 3D convolutional neural networks based on motion profile |
| CN109271938A (en) * | 2018-09-19 | 2019-01-25 | 上海鸢安智能科技有限公司 | A kind of gas station's emptying Safety Monitoring Control method based on intelligent video analysis technology |
| CN111126115A (en) * | 2018-11-01 | 2020-05-08 | 顺丰科技有限公司 | Violence sorting behavior identification method and device |
| CN109829443A (en) * | 2019-02-23 | 2019-05-31 | 重庆邮电大学 | Video behavior recognition methods based on image enhancement Yu 3D convolutional neural networks |
| CN109947990A (en) * | 2019-03-29 | 2019-06-28 | 北京奇艺世纪科技有限公司 | A kind of wonderful detection method and system |
| CN110070002A (en) * | 2019-03-29 | 2019-07-30 | 上海理工大学 | A kind of Activity recognition method based on 3D convolutional neural networks |
| CN110555368A (en) * | 2019-06-28 | 2019-12-10 | 西安理工大学 | Fall-down behavior identification method based on three-dimensional convolutional neural network |
| CN110456723A (en) * | 2019-08-15 | 2019-11-15 | 成都睿晓科技有限公司 | A kind of emptying area, gas station security management and control system based on deep learning |
Non-Patent Citations (5)
| Title |
|---|
| SIMONE ACCATTOLI 等: "Violence detection in videos by combining 3D convolutional neural networks and support vector machines", 《APPLIED ARTIFICIAL INTELLIGENCE》, vol. 34, no. 04, pages 329 - 344 * |
| ZHANG WENJIN 等: "Dynamic Hand Gesture Recognition Based on 3D Convolutional Neural Network Models", 《2019 IEEE 16TH INTERNATIONAL CONFERENCE ON NETWORKING, SENSING AND CONTROL (ICNSC)》, pages 224 - 229 * |
| 何坤鹏: "基于视频流的人体行为识别研究与设计", 《中国优秀硕士学位论文全文数据库 (信息科技辑)》, no. 07, pages 138 - 1000 * |
| 张颖 等: "基于3D卷积神经网络的人体行为识别方法", 《软件导刊》, vol. 16, no. 11, pages 9 - 11 * |
| 林云玫: "深度学习与HMM在基本行为识别上的应用", 《中国优秀硕士学位论文全文数据库 (信息科技辑)》, no. 05, pages 138 - 1049 * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2023123926A1 (en) * | 2021-12-28 | 2023-07-06 | 苏州浪潮智能科技有限公司 | Artificial intelligence task processing method and apparatus, electronic device, and readable storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111383429A (en) | Method, system, device and storage medium for detecting dress of workers in construction site | |
| CN112966772A (en) | Multi-person online image semi-automatic labeling method and system | |
| CN111896549B (en) | Building crack monitoring system and method based on machine learning | |
| CN112613569B (en) | Image recognition method, training method and device for image classification model | |
| CN111900694B (en) | Relay protection equipment information acquisition method and system based on automatic identification | |
| CN112115927A (en) | Intelligent machine room equipment identification method and system based on deep learning | |
| CN114913619A (en) | Intelligent mobile inspection method and system | |
| CN113869162A (en) | Violation identification method and system based on artificial intelligence | |
| CN113888514A (en) | Method and device for detecting defects of ground wire, edge computing equipment and storage medium | |
| CN111144749A (en) | Intelligent labeling crowdsourcing platform for power grid images and working method | |
| CN112906488A (en) | Security protection video quality evaluation system based on artificial intelligence | |
| CN113988573A (en) | Risk judgment method, system and medium for routing inspection unmanned aerial vehicle based on power system | |
| CN115937101A (en) | Quality detection method, device, equipment and storage medium | |
| CN112115876A (en) | Water-soluble method experimental process identification method based on 3D convolutional neural network | |
| CN117237925B (en) | Intelligent road disease inspection method and system based on computer vision | |
| KR20220105689A (en) | System for determining defect of display panel based on machine learning model | |
| CN117093822A (en) | Industrial brain data analysis platform based on industrial knowledge graph | |
| CN116934674A (en) | Large-granularity silicon material detection and removal system based on YOLOv8 network | |
| CN116958035A (en) | Product part defect detection method, device, medium and electronic equipment | |
| CN115689206A (en) | Intelligent monitoring method for transformer substation infrastructure progress based on deep learning | |
| KR20220161601A (en) | System for determining defect of image inspection target using deep learning model | |
| CN115170783A (en) | Expressway pavement crack detection method using street view image | |
| CN113850773A (en) | Detection method, device, equipment and computer readable storage medium | |
| CN113837178A (en) | Deep learning-based automatic positioning and unified segmentation method for meter of transformer substation | |
| KR102660626B1 (en) | System and method for vision inspection based on AI that can be verified on site |
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 |