CN112689078A - Rainwater identification management system based on artificial intelligence video analysis - Google Patents
Rainwater identification management system based on artificial intelligence video analysis Download PDFInfo
- Publication number
- CN112689078A CN112689078A CN202110086997.7A CN202110086997A CN112689078A CN 112689078 A CN112689078 A CN 112689078A CN 202110086997 A CN202110086997 A CN 202110086997A CN 112689078 A CN112689078 A CN 112689078A
- Authority
- CN
- China
- Prior art keywords
- rainwater
- water level
- model
- identification model
- level monitoring
- 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
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A50/00—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE in human health protection, e.g. against extreme weather
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Remote Monitoring And Control Of Power-Distribution Networks (AREA)
Abstract
The invention discloses a rainwater identification management system based on artificial intelligence video analysis, which comprises a basic data access layer, an intelligent identification model layer and an intelligent analysis application layer, wherein the intelligent identification model layer comprises a rainwater identification model and a water level monitoring model, the basic data access layer respectively transmits a cutting picture and the accumulated water amount of each transformer substation in a jurisdiction to the rainwater identification model and the water level monitoring model of the intelligent identification model layer, the rainwater identification model and the water level monitoring model in the intelligent identification model layer respectively transmit a rainwater identification result and a water level monitoring result to the intelligent analysis application layer, and the intelligent analysis application layer carries out flood prevention monitoring and early warning on each transformer substation in the jurisdiction according to received data And (5) problems are solved.
Description
Technical Field
The invention relates to the technical field of machine vision, in particular to a rainwater identification management system based on artificial intelligence video analysis.
Background
With the continuous fusion development of the digital revolution and the energy revolution, the artificial intelligence technology driven by mass data can play a role difficult to replace, the state positions the artificial intelligence as one of the novel infrastructures at the science and technology end, the power industry is an important field of the infrastructures, and the application of the artificial intelligence in the aspects of equipment operation and maintenance, power grid dispatching, intelligent customer service and the like is actively expanded.
At present, management is still performed by means of excessive manual intervention in flood prevention monitoring management of each transformer substation, on one hand, certain manual management cost is needed, on the other hand, various management loopholes exist in manual management, and a set of automatic flood prevention monitoring management system is not formed in flood prevention monitoring management of each transformer substation in the prior art.
The present invention provides a new solution to this problem.
Disclosure of Invention
Aiming at the defects in the prior art, the invention aims to provide a rainwater identification management system based on artificial intelligence video analysis, which effectively solves the problems that the transformer substations are still managed by means of excessive manual intervention in the prior art, and a set of automatic flood prevention monitoring management system is not formed in flood prevention monitoring management of each transformer substation.
The technical scheme for solving the problem is that the rainwater identification management system based on artificial intelligence video analysis comprises a basic data access layer, an intelligent identification model layer and an intelligent analysis application layer; the intelligent identification model layer comprises a rainwater identification model and a water level monitoring model, the basic data access layer acquires video streams on a unified video platform, acquires video streams on monitoring cameras of substations in the jurisdiction, and transmits a cutting picture obtained by cutting the video streams to the rainwater identification model of the intelligent identification model layer, the basic data access layer also acquires accumulated water quantity acquired by water level monitoring equipment of each substation in the jurisdiction and transmits the accumulated water quantity to the water level monitoring model of the intelligent identification model layer, the rainwater identification model and the water level monitoring model in the intelligent identification model layer respectively transmit a rainwater identification result and a water level monitoring result to the intelligent analysis application layer, and the intelligent analysis application layer carries out flood prevention monitoring and early warning on each substation in the jurisdiction according to received data;
the method comprises the steps that a basic data access layer obtains video streams on a unified video platform, the unified video platform collects the video streams on monitoring cameras which have the capacity of sharing the video streams to the outside in substations in the jurisdiction, the basic data access layer also obtains the video streams on the monitoring cameras of the substations in the jurisdiction, the monitoring cameras do not have the monitoring cameras which share the video streams to the unified video platform, each monitoring camera has a serial number, the serial number of each monitoring camera serves as a unique identifier of the source of each section of the video streams, each water level monitoring device has a serial number, and the serial number of each water level monitoring device serves as a unique identifier of the source of each accumulated water quantity.
The invention has the following beneficial effects:
the invention discloses a rainwater identification management system based on artificial intelligence video analysis, which utilizes a basic data access layer to obtain video streams on a unified video platform and also obtain video streams on monitoring cameras of substations in a jurisdiction, and transmits a cutting picture obtained by cutting the video streams to a rainwater identification model of an intelligent identification model layer, the basic data access layer also obtains accumulated water quantity collected by water level monitoring equipment of each substation in the jurisdiction and transmits the accumulated water quantity to the water level monitoring model of the intelligent identification model layer, the rainwater identification model and the water level monitoring model in the intelligent identification model layer respectively transmit a rainwater identification result and a water level monitoring result to an intelligent analysis application layer, the intelligent analysis application layer carries out flood prevention monitoring and early warning reminding on each substation in the jurisdiction according to received data, and the problem that the substation is still managed by excessive manual intervention in the prior art is effectively solved, the problem that a set of automatic flood prevention monitoring and management system is not formed in flood prevention monitoring and management of each transformer substation is solved.
Drawings
FIG. 1 is a system architecture diagram of the present invention.
Fig. 2 is a flow chart of an application of the present invention.
Detailed Description
In order to achieve the foregoing and other objects, features and advantages of the invention, the following description should be taken in conjunction with the accompanying drawings
The detailed description of the embodiments, with reference to fig. 1-2, will be apparent. The structural contents mentioned in the following embodiments are all referred to the attached drawings of the specification.
The following describes in detail a rain recognition management system based on artificial intelligence video analysis according to an embodiment of the present invention with reference to the accompanying drawings.
A rainwater identification management system based on artificial intelligence video analysis comprises a basic data access layer, an intelligent identification model layer and an intelligent analysis application layer; the intelligent identification model layer comprises a rainwater identification model and a water level monitoring model, the basic data access layer acquires video streams on a unified video platform and video streams on monitoring cameras of substations in the jurisdiction, cutout pictures obtained after the video streams are cut are transmitted to the rainwater identification model of the intelligent identification model layer, the basic data access layer also acquires water accumulation quantity acquired by water level monitoring equipment of the substations in the jurisdiction and transmits the water accumulation quantity to the water level monitoring model of the intelligent identification model layer, the rainwater identification model and the water level monitoring model in the intelligent identification model layer respectively transmit rainwater identification results and water level monitoring results to the intelligent analysis application layer, and the intelligent analysis application layer carries out flood prevention monitoring and early warning on the substations in the jurisdiction according to received data.
The basic data access layer is responsible for acquiring data, the acquired data comprise unstructured data and structured data, the unstructured data refer to video streams on monitoring cameras of substations in the jurisdiction through an RTSP protocol, video streams on a unified video platform are acquired, and cut pictures obtained after the video streams are cut are transmitted to a rainwater identification model in the intelligent identification model layer in a real-time transmission mode;
the structured data refer to the ponding quantity collected by the water level monitoring equipment of each transformer substation in the jurisdiction, the ponding quantity is transmitted to the water level monitoring model of the intelligent identification model layer, the ponding quantity collected by the water level monitoring equipment of each transformer substation in the jurisdiction is obtained at regular time, and the water level monitoring equipment is laid at the appointed position of each transformer substation in the jurisdiction;
the working process of the basic data access layer comprises the following steps:
s1, acquiring video streams on monitoring cameras of substations in the jurisdiction in real time through an RTSP (real time streaming protocol) and acquiring video streams on a unified video platform;
acquiring the water accumulation quantity collected by water level monitoring equipment of each transformer substation in the jurisdiction at regular time through a standard interface;
s2: performing high-frequency picture cutting on the video stream in the step S1 to obtain a picture cutting picture;
s3: transmitting the picture cutting picture in the step S2 to a rainwater identification model of the intelligent identification model layer;
the accumulated water amount in step S1 is transmitted to the water level monitoring model of the smart recognition model layer.
The intelligent identification model layer comprises a rainwater identification model and a water level monitoring model, the rainwater identification model receives the cutting picture transmitted by the basic data access layer and outputs a rainwater identification result to the intelligent analysis application layer, and the water level monitoring model receives the accumulated water amount transmitted by the basic data access layer and transmits the water level monitoring result to the intelligent analysis application layer;
the rainwater identification model is constructed by the following steps:
a1: copying a historical monitoring video stream as a historical exercise sample, wherein the historical monitoring video stream refers to a video stream stored by a monitoring camera of a unified video platform and a transformer substation;
a2, selecting a historical exercise sample to cut pictures, wherein the obtained cut pictures need to be under different time, different illumination and different weather conditions, and specifically comprise pictures under different weather conditions such as cloudy days, sunny days, rainy days, direct light, reverse light and strong wind;
a3, labeling the recognition areas of the cut pictures, wherein the recognition areas refer to a certain relatively fixed and motionless area in each cut picture, and the recognition areas can be a certain fixed ground, because the ground is motionless under any condition, the recognition areas are selected and labeled in the cut pictures by using a labelme image labeling tool, so as to eliminate the interference of other areas in the cut pictures on rainwater recognition and reduce the number of samples;
a4: extracting raindrop characteristics in the cut picture identification area, wherein the raindrop characteristics refer to the size and the number of raindrops;
a5: creating a rain sample library according to the rain drop characteristics in the step A4;
a6, training a rainwater identification model by using a rainwater sample library, wherein the rainwater identification model is a machine learning model and obtains the rainwater identification capability through the provided learning material, namely the rainwater sample library;
the training process of the rain recognition model comprises the following steps:
a6.1: preparing a training sample from a rainwater sample library, and dividing the training sample into three parts, namely a training sample, a test sample and a verification sample;
a6.2: inputting the practice sample into a rainwater identification model for training the rainwater identification model;
a6.3: adjusting and optimizing the rainwater identification model through the test sample;
a6.4: verifying the stability of the rainwater identification model by using a verification sample;
a7, deploying a rainwater identification model to enter practical application, receiving the cutting picture transmitted by the basic data access layer by the rainwater identification model, and outputting a rainwater identification result, wherein the rainwater identification result comprises: surveillance camera head serial number, discernment time, rainfall size, the rainfall size specifically divide into: no rain, light rain, medium rain, heavy rain and extra heavy rain;
the water level monitoring model receives the water accumulation quantity collected by the water level monitoring equipment in the basic data access layer, the water level monitoring model calculates the increasing speed according to the received water accumulation quantity and the monitoring time for receiving the water accumulation quantity, for example, the water accumulation quantity at t1 is h1, the water accumulation quantity at t2 is h2, the increasing speed is (h2-h1)/(t2-t1), wherein the frequency of water accumulation quantity collection can be collected once according to 5 minutes, the collection time interval can also be adjusted according to the actual situation, the water level monitoring model transmits the water level monitoring result to the intelligent analysis application layer, and the water level monitoring result comprises the water accumulation quantity collected by the water level monitoring equipment, the serial number of the water level monitoring equipment, the monitoring time and the increasing speed.
The intelligent analysis application layer carries out flood prevention monitoring management and early warning reminding on each transformer substation in the district according to the received rainwater identification result and water level monitoring result transmitted by the intelligent identification model layer; the intelligent analysis application layer comprises a monitoring center, a data management center, an early warning management center, an equipment management center and a panoramic analysis center; specifically, the monitoring center monitors the real-time states of rain and water accumulation of all substations in the jurisdiction, the data management center comprehensively manages the rain identification results and water level monitoring results output by the intelligent identification model layer, the early warning management center timely sends out warning prompts under the condition of meeting early warning conditions, the equipment management center manages the running states and the access conditions of all equipment accessed into the management system, the equipment accessed into the management system comprises a monitoring camera, a unified video platform and water level monitoring equipment, the panoramic analysis center analyzes and predicts the rain and water accumulation from the angles of the area, the history period and the like of the equipment accessed into the management system, and the intelligent analysis application layer provides all-around and all-weather flood prevention monitoring and early warning management.
When the intelligent identification model layer is used specifically, the intelligent identification model layer comprises a rainwater identification model and a water level monitoring model, the basic data access layer acquires video streams on monitoring cameras of substations in the jurisdiction through an RTSP (real time streaming protocol) protocol, acquires video streams on a unified video platform, transmits a cutting picture obtained by cutting the video streams to the rainwater identification model of the intelligent identification model layer, the basic data access layer also acquires accumulated water quantity acquired by water level monitoring equipment of all the substations in the jurisdiction and transmits the accumulated water quantity to the water level monitoring model of the intelligent identification model layer, the rainwater identification model and the water level monitoring model in the intelligent identification model layer respectively transmit a rainwater identification result and a water level monitoring result to the intelligent analysis application layer, and the intelligent analysis application layer carries out flood prevention monitoring on all the substations in the jurisdiction according to the received data, The early warning reminding effectively solves the problems that the transformer substations are still managed by means of excessive manual intervention in the prior art, and a set of automatic flood prevention monitoring management system is not formed in the flood prevention monitoring management of each transformer substation.
Claims (6)
1. A rainwater identification management system based on artificial intelligence video analysis is characterized by comprising a basic data access layer, an intelligent identification model layer and an intelligent analysis application layer, wherein the intelligent identification model layer comprises a rainwater identification model and a water level monitoring model; the method comprises the steps that a basic data access layer is used for obtaining video streams on a unified video platform and video streams on monitoring cameras of substations in a district, cut pictures obtained after the video streams are cut are transmitted to a rainwater identification model of an intelligent identification model layer, the basic data access layer also obtains water accumulation quantity collected by water level monitoring equipment of the substations in the district and transmits the water accumulation quantity to a water level monitoring model of the intelligent identification model layer, the rainwater identification model and the water level monitoring model in the intelligent identification model layer respectively transmit a rainwater identification result and a water level monitoring result to an intelligent analysis application layer, and the intelligent analysis application layer carries out flood prevention monitoring and early warning on the substations in the district according to received data.
2. The rainwater identification management system based on artificial intelligence video analysis as claimed in claim 1, wherein the basic data access layer obtains video streams on a unified video platform, the unified video platform collects video streams on monitoring cameras of substations in the jurisdiction, the monitoring cameras share the video streams to the unified video platform, each monitoring camera has a number, the number of the monitoring camera is used as a unique identifier of a source of each section of video stream, each water level monitoring device has a number, and the number of the water level monitoring device is used as a unique identifier of a source of a water retention amount each time.
3. The artificial intelligence video analysis-based rain recognition management system according to claim 1, wherein the working process of the basic data access layer comprises the following steps:
s1, acquiring video streams on monitoring cameras of substations in the jurisdiction in real time through an RTSP (real time streaming protocol) and acquiring video streams on a unified video platform;
acquiring the water accumulation quantity collected by water level monitoring equipment of each transformer substation in the jurisdiction at regular time;
s2: cutting the video stream in the step S1 to obtain a cut picture;
s3: transmitting the picture cutting picture in the step S2 to a rainwater identification model of the intelligent identification model layer;
the accumulated water amount in step S1 is transmitted to the water level monitoring model of the smart recognition model layer.
4. The artificial intelligence video analysis-based rainwater identification management system according to claim 1, wherein the construction steps of the rainwater identification model in the intelligent identification model layer are as follows:
a1: copying a historical monitoring video stream as a historical exercise sample, wherein the historical monitoring video stream refers to a video stream stored by a monitoring camera of a unified video platform and a transformer substation;
a2, selecting a historical exercise sample to cut the picture, wherein the obtained cut picture needs to be the picture under different time, different illumination and different meteorological conditions;
a3, marking an identification area of the picture cutting picture, wherein the identification area refers to a relatively fixed and immobile area in each picture cutting picture;
a4: extracting raindrop characteristics in the cut picture identification area, wherein the raindrop characteristics refer to the size and the number of raindrops;
a5: creating a rain sample library according to the rain drop characteristics in the step A4;
a6, training a rainwater identification model by using a rainwater sample library, wherein the rainwater identification model is a machine learning model and obtains the rainwater identification capability through the provided learning material, namely the rainwater sample library;
a7, the rain water identification model outputs a rain water identification result according to the cutting picture transmitted by the basic data access layer, and the rain water identification result comprises: monitoring camera serial number, discernment time, rainfall size.
5. The artificial intelligence video analysis-based rain recognition management system according to claim 4, wherein in the rain recognition model building step A6, the step of training the rain recognition model using the rain sample library includes:
a6.1: preparing a training sample from a rainwater sample library, and dividing the training sample into three parts, namely a training sample, a test sample and a verification sample;
a6.2: inputting the practice sample into a rainwater identification model to train the rainwater identification model;
a6.3: adjusting and optimizing the rainwater identification model through the test sample;
a6.4: and verifying the stability of the rainwater identification model by using the verification sample.
6. The rainwater identification management system based on artificial intelligence video analysis of claim 1, wherein the water level monitoring model in the intelligent identification model layer receives the accumulated water amount collected by the water level monitoring equipment in the basic data access layer, the water level monitoring model calculates the growth rate according to the received accumulated water amount and the monitoring time for receiving the accumulated water amount, the water level monitoring model transmits the water level monitoring result to the intelligent analysis application layer, and the water level monitoring result comprises the accumulated water amount collected by the water level monitoring equipment, the water level monitoring equipment number, the monitoring time and the growth rate.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110086997.7A CN112689078A (en) | 2021-01-22 | 2021-01-22 | Rainwater identification management system based on artificial intelligence video analysis |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110086997.7A CN112689078A (en) | 2021-01-22 | 2021-01-22 | Rainwater identification management system based on artificial intelligence video analysis |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112689078A true CN112689078A (en) | 2021-04-20 |
Family
ID=75458905
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110086997.7A Pending CN112689078A (en) | 2021-01-22 | 2021-01-22 | Rainwater identification management system based on artificial intelligence video analysis |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112689078A (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107195164A (en) * | 2017-07-03 | 2017-09-22 | 长江水利委员会长江科学院 | Mountain flood on-line monitoring identification method for early warning and its early warning system |
| CN109191358A (en) * | 2018-10-19 | 2019-01-11 | 中电科新型智慧城市研究院有限公司 | Urban Flood control prevention waterlogging monitoring commander's operating system |
| CN110610190A (en) * | 2019-07-31 | 2019-12-24 | 浙江大学 | Convolutional neural network rainfall intensity classification method for rainy pictures |
| CN111353380A (en) * | 2020-01-08 | 2020-06-30 | 广州华工中云信息技术有限公司 | Urban road ponding image recognition system based on machine image recognition technology |
| KR20200087347A (en) * | 2018-12-31 | 2020-07-21 | 부산대학교 산학협력단 | Method and Apparatus for Rainfall Recognition from Road Surveillance Videos Using TSN |
-
2021
- 2021-01-22 CN CN202110086997.7A patent/CN112689078A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107195164A (en) * | 2017-07-03 | 2017-09-22 | 长江水利委员会长江科学院 | Mountain flood on-line monitoring identification method for early warning and its early warning system |
| CN109191358A (en) * | 2018-10-19 | 2019-01-11 | 中电科新型智慧城市研究院有限公司 | Urban Flood control prevention waterlogging monitoring commander's operating system |
| KR20200087347A (en) * | 2018-12-31 | 2020-07-21 | 부산대학교 산학협력단 | Method and Apparatus for Rainfall Recognition from Road Surveillance Videos Using TSN |
| CN110610190A (en) * | 2019-07-31 | 2019-12-24 | 浙江大学 | Convolutional neural network rainfall intensity classification method for rainy pictures |
| CN111353380A (en) * | 2020-01-08 | 2020-06-30 | 广州华工中云信息技术有限公司 | Urban road ponding image recognition system based on machine image recognition technology |
Non-Patent Citations (1)
| Title |
|---|
| 褚泽帆等: "基于视频的水位识别监测***设计", 《电子设计工程》 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108109385B (en) | System and method for identifying and judging dangerous behaviors of power transmission line anti-external damage vehicle | |
| CN110784511B (en) | Intelligent street lamp system based on edge internet of things agent | |
| CN113780312B (en) | Highway road surface condition detecting system | |
| CN101709954B (en) | Overhead wire icing on-line monitoring device based on machine vision | |
| CN111667230A (en) | Unmanned aerial vehicle autonomous inspection operation monitoring and analyzing system and method | |
| CN110188586A (en) | System and application method based on meteorological observation, road camera shooting visibility identification | |
| CN110955288B (en) | Environment supervision system based on big data | |
| CN113110200A (en) | Urban waterlogging early warning system based on weather and rainfall flood model | |
| CN109208468B (en) | Automatic bridge inspection system based on video analysis | |
| CN202331779U (en) | Collector, collection device and collection system for crop multisource record information | |
| CN111985352B (en) | AI front-end substation inspection video real-time identification method and system | |
| CN113888871B (en) | Automatic handling linkage system and method for expressway traffic incidents | |
| CN206649006U (en) | A kind of monitoring device for air pollution | |
| CN114333313B (en) | Intelligent inspection method based on highway monitoring | |
| CN113115248A (en) | High-voltage power transmission equipment on-line monitoring system based on Lora | |
| CN115271113A (en) | Intelligent inspection method for construction process of pumped storage power station | |
| CN115390504A (en) | Wisdom farming system based on 5G thing networking | |
| CN201742505U (en) | Vehicle-mounted monitoring system based on 3G network transmission | |
| CN102523432A (en) | Intelligent vehicle-mounted monitoring system based on third generation (3G) | |
| CN109543588A (en) | Method, apparatus, system, service platform and the medium that traffic accident responsibility determines | |
| CN114690693A (en) | Remote gate monitoring system and method based on remote measuring terminal | |
| CN112689078A (en) | Rainwater identification management system based on artificial intelligence video analysis | |
| CN116886869A (en) | Video monitoring system and video tracing method based on AI | |
| CN210865032U (en) | Urban waterlogging monitoring and early warning system | |
| CN115421503B (en) | Unmanned aerial vehicle inspection system for bridge |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20210420 |