WO2022127683A1 - 一种基于流态视频监测的过水建筑物结构损伤诊断方法及系统 - Google Patents

一种基于流态视频监测的过水建筑物结构损伤诊断方法及系统 Download PDF

Info

Publication number
WO2022127683A1
WO2022127683A1 PCT/CN2021/136749 CN2021136749W WO2022127683A1 WO 2022127683 A1 WO2022127683 A1 WO 2022127683A1 CN 2021136749 W CN2021136749 W CN 2021136749W WO 2022127683 A1 WO2022127683 A1 WO 2022127683A1
Authority
WO
WIPO (PCT)
Prior art keywords
water
flow state
water surface
image
database
Prior art date
Application number
PCT/CN2021/136749
Other languages
English (en)
French (fr)
Inventor
庞博慧
张陆陈
迟福东
肖海斌
王忠军
Original Assignee
华能澜沧江水电股份有限公司
水利部交通运输部国家能源局南京水利科学研究院
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by 华能澜沧江水电股份有限公司, 水利部交通运输部国家能源局南京水利科学研究院 filed Critical 华能澜沧江水电股份有限公司
Priority to NO20230279A priority Critical patent/NO20230279A1/en
Publication of WO2022127683A1 publication Critical patent/WO2022127683A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
    • G01N21/84—Systems specially adapted for particular applications
    • G01N21/88—Investigating the presence of flaws or contamination
    • G01N21/95—Investigating the presence of flaws or contamination characterised by the material or shape of the object to be examined
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
    • G06V10/74—Image or video pattern matching; Proximity measures in feature spaces
    • G06V10/75—Organisation of the matching processes, e.g. simultaneous or sequential comparisons of image or video features; Coarse-fine approaches, e.g. multi-scale approaches; using context analysis; Selection of dictionaries
    • G06V10/751—Comparing pixel values or logical combinations thereof, or feature values having positional relevance, e.g. template matching
    • G—PHYSICS
    • G08—SIGNALLING
    • G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
    • G08B21/00—Alarms responsive to a single specified undesired or abnormal condition and not otherwise provided for
    • G08B21/18—Status alarms
    • G08B21/182—Level alarms, e.g. alarms responsive to variables exceeding a threshold
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N23/00—Cameras or camera modules comprising electronic image sensors; Control thereof

Definitions

  • the invention belongs to the technical field of water conservancy and hydropower engineering and video monitoring, in particular to a method and system for diagnosing structural damage of water-passing buildings based on flow state video monitoring.
  • the water-passing structures are responsible for the important task of flood discharge and over-current, especially when the flood flow is large during the flood season, the water-passing structures are in a long-term discharge state, and the water flow velocity of the water-passing structures in large-scale hydropower projects is often as high as 40-50m/m. s. Due to the long-term scouring of high-speed water flow, the probability of damage to the water-passing structures is very high. According to incomplete statistics, 1/3 of the water-passing structures have been damaged to varying degrees, and some are quite serious.
  • the water-passing structure is damaged during the discharge, due to various reasons such as the untimely inspection of personnel, the inability to observe clearly due to the influence of atomization, and the obstruction of the water flow by structures, it will not be discovered in time and measures will be taken in time, resulting in losses.
  • high-definition camera technology long-distance, fog-penetrating, and high-definition camera technology can replace manual real-time monitoring of the discharge process. Equipped with high-speed, high-reliability image processing and analysis technology, the discharge flow pattern and water surface line can be found in time.
  • the abnormal condition of the water-passing building can be realized, and the function of diagnosing the structural damage of the water-passing building is realized.
  • the structural damage diagnosis method of water-passing buildings basically adopts sensor technology, which mainly has the following shortcomings: (1) The layout of sensors generally needs to be constructed on the structure of water-passing buildings, which requires a lot of preparation work, complicated process, many coordination parties, and energy consumption. (2) The sensor equipment is exposed to high-speed water flow for a long time, the durability is poor, and the failure rate is high; (3) The sensor monitoring is all "point" observation , it is difficult to truly and comprehensively reflect the overall working behavior changes of the discharge structure. Therefore, how to overcome the deficiencies of the prior art is an urgent problem to be solved in the present technical field.
  • the purpose of the present invention is to solve the deficiencies of the prior art, and to provide a method and system for diagnosing structural damage of water-passing buildings based on flow state video monitoring.
  • a structural damage diagnosis system for water-passing buildings based on flow state video monitoring comprising a high-definition fog-penetrating camera, a base, a dedicated workstation and a computer monitor;
  • the high-definition fog-penetrating camera is installed on the base; the base is provided with a cavity; the network interface and power interface of the high-definition fog-penetrating camera are installed in the cavity;
  • a smart wiper module is installed on the HD fog-penetrating camera; the smart wiper module is used to clean the water droplets on the lens of the HD fog-penetrating camera in time to keep the lens clean;
  • the power interface, computer monitor, and special workstation are connected to the power supply through a power cord respectively;
  • the network interface is connected with the dedicated workstation through the data transmission line;
  • the dedicated workstation includes an image analysis module, a background database and an alarm module;
  • the image analysis module is respectively connected with the background database, the alarm module and the computer display;
  • the image analysis module is used to process the image captured by the high-definition fog-penetrating camera, analyze the contour of the water surface line, and then compare the feature point coordinates of the obtained water surface line contour with the image feature point coordinates corresponding to normal working conditions in the background database. If it is abnormal, send an instruction to the alarm module to alarm;
  • the background database is also used to store the images before processing by the image analysis module and the processed data;
  • the computer monitor is used to display the image before processing by the image analysis module and the processed data.
  • a protective cover is also included; the protective cover is arranged outside the base and the high-definition fog-penetrating camera, and is used to protect the base and the high-definition fog-penetrating camera.
  • the background database also includes a normal flow state database and an abnormal flow state database; when the image analysis module compares the image water surface line profile captured by the high-definition fog-penetrating camera with the normal flow state data in the background database, the comparison result is abnormal.
  • the image analysis module compares the image water surface line profile captured by the high-definition fog-penetrating camera with the normal flow state data in the background database, the comparison result is abnormal.
  • the image analysis module is used, the image before processing and the processed data are stored in the abnormal flow state database; otherwise, they are stored in the normal flow state database.
  • step g Calculate line by line according to step c to step f, respectively connect x i , left and x i , and right as the left contour and right contour of the water surface line.
  • the present invention also provides a method for diagnosing structural damage of water-passing buildings based on flow state video monitoring, using the above-mentioned system for diagnosing structural damage of water-passing buildings based on flow state video monitoring, including the following steps:
  • Step (1) according to the characteristics of the discharge flow, lay out several feature points on the site to cover the entire shooting range as much as possible, place anti-scour and easy-to-recognize signs at the feature points, measure their coordinates, and input them into the background database. ;
  • Step (2) store the discharge flow pattern, water surface line and characteristic point data of the water-passing structure under normal working conditions, as well as the prototype observation data obtained in real time in the field into the normal flow pattern database;
  • the data of the building discharge flow state, water surface line and characteristic point, as well as the abnormal original view data that actually occurred on site, are stored in the abnormal flow state database;
  • Step (3) turn on the high-definition fog-penetrating camera, aim at the discharge flow state of the water-passing building, adjust the shooting height so that the captured image covers the entire water surface and the side walls on both sides of the water-passing building; start the smart wiper module, and adjust the aperture , focal length, exposure parameters and shutter speed to ensure clear images;
  • Step (4) formally shooting after obtaining an image whose resolution meets the requirements, shooting a high-definition image every 10 minutes, and performing image processing and analysis to determine the outline of the water surface line;
  • Step (5) compare the coordinate H i (X, Y) of the feature point of the water surface line profile obtained by the image calculation with the coordinate H' i ( X, Y) of the image feature point corresponding to the normal working condition in the background database, when ⁇ H i (X,Y)-H′ i (X,Y) ⁇ /H′ i (X,Y)>10%, it is regarded as an abnormal flow state, an alarm is activated and the data is stored in the abnormal flow state database; when ⁇ H i When (X,Y)-H′ i (X,Y) ⁇ /H′ i (X,Y) ⁇ 10%, it is regarded as a normal flow state and stored in the normal flow state database;
  • step (6) after the leakage is completed, the high-definition fog-penetrating camera is turned off, and all images are stored in the background database.
  • the outline of the water surface line is compared with the situation of manual observation. If the situation is not satisfactory, a 0 , n 0 , and m 0 can be modified, and steps c to f are repeated until the analysis accuracy requirements are met.
  • the principle of feature point selection is to cover the entire contour of the water surface as uniformly as possible, and to include a certain number of fixed points with known coordinates.
  • the abnormal flow state database mainly accumulates data for analyzing the correlation between damage characteristics and discharge flow state in the future, and can also be used to review and compare data of frequently occurring abnormal working conditions when the amount of data is large.
  • the structural damage diagnosed in the present invention means that the damage degree of the structure, including erosion pits, abrasion pits, cavitation pits, etc., has an impact on the water flow that exceeds the tolerance range, that is, 10%.
  • the high-definition fog-penetrating camera has the functions of preventing rain, fog-penetrating, and anti-vibration, and adopts a high shutter speed to meet the shooting effect of 40-50m/s high-speed dynamic water flow, and realize the clear shooting function in the high-speed water flow atomization area. .
  • the base should be installed according to the size of the image, the focal length of the lens, the size of the target and the actual conditions of the site.
  • the base has the characteristics of stability and anti-scour.
  • the network interface and power interface are arranged in the base. Through the power interface, the high-definition fog camera can be charged, and the captured images can be transmitted to the dedicated workstation in real time.
  • the dedicated workstation has powerful image processing capability, large capacity, high speed and high reliability.
  • the feature points refer to the feature points extracted from the image with clear targets, easy identification and no change in position in order to improve the image registration accuracy.
  • the alarm function will be activated when the water surface line difference is greater than the specified threshold.
  • the back-end database stores the pictures, water surface line data and characteristic point data of the water-passing structures, which are obtained by prototype observation, numerical simulation, physical model test and other methods, and whose measurement accuracy meets the requirements, including the normal water-passing structures.
  • Working status database and abnormal working status database are obtained by prototype observation, numerical simulation, physical model test and other methods, and whose measurement accuracy meets the requirements, including the normal water-passing structures.
  • HD fog cameras and bases should be equipped with protective covers to protect against harsh weather conditions or man-made damage when working in the field.
  • the image processing of the image analysis module includes water surface line analysis, feature point extraction, and feature point abnormal judgment; the background database stores the flow diagram, water surface line and feature point data under various working conditions of the water-passing building.
  • the installation operation of the system of the invention is as follows: on the side perpendicular to the water flow direction of the water-passing building, the shooting position is determined according to the image size, the focal length of the lens, the size of the target object and the actual conditions of the scene; the base is installed at the shooting position; Fix it on the base, and connect the power interface and network interface; according to the transmission distance requirements and on-site network conditions, install a special workstation in the corresponding position, connect it with the network interface of the high-definition fog-penetrating camera, and receive the data transmitted by the high-definition fog-penetrating camera in real time. image data.
  • the high-definition required resolution of the high-definition fog-penetrating camera of the present invention is not less than 2560*1440.
  • the present invention has the following beneficial effects:
  • the invention Realize non-destructive and non-contact observation of structural health status of water-passing buildings.
  • the invention adopts long-distance, fog-penetrating, high-definition camera technology to photograph the discharge flow state, and uses the causal relationship that the structural damage will have a certain impact on the water flow state to carry out structural health diagnosis, and does not need to carry out the structural health diagnosis on the structure of the water-passing building itself. Damaged construction, avoiding the disadvantages of new defect incentives caused by the installation of sensors on the structure.
  • the present invention does not observe some "points”, but observes the overall leakage flow state, because any defect at any point or a defect with a certain threat will inevitably lead to the change of the leakage flow state, so the leakage flow state can be observed by observing the leakage flow state. Determine whether the structure itself is damaged.
  • the invention does not need to coordinate all parties for approval and scheduling, etc., the preparation workload is small, the process is simple, the required personnel is small, the equipment can be recycled, and the labor cost and equipment cost are low.
  • the present invention does not directly contact the high-speed water flow, which avoids the failure of the sensor caused by the high-speed water flow.
  • the method for diagnosing structural damage of water-passing structures based on flow state video monitoring provided by the present invention has the advantages of economy, simplicity, speed, high reliability, and strong durability, and can comprehensively and truly reflect the health state of the discharge structure.
  • the advantages of non-destructive and non-contact real-time observation of the structural health status of water-passing buildings are widely used in the fields of hydropower engineering and other industries, such as safety monitoring of water-passing buildings, high-speed water flow monitoring, scientific research, and education.
  • Fig. 1 is the structural representation of the structure damage diagnosis system of water-passing buildings based on flow state video monitoring of the present invention
  • Fig. 2 is the software and hardware system architecture diagram of the present invention
  • Fig. 3 is a schematic diagram of the identification and diagnosis flow chart of structural damage of water-passing buildings based on flow state video monitoring;
  • Fig. 4 is the schematic flow chart of water surface line profile analysis
  • Fig. 5 is a real shot of the discharge flow state of the water-passing building
  • Fig. 6 is the water surface line profile analysis processing result figure
  • Figure 7 is a diagram of the water surface and its contour.
  • 1-HD fog-penetrating camera 2-Smart wiper module; 3-Base; 4-Network interface; 5-Power interface; 6-Power cable; 7-Data transmission cable; 8-Computer monitor; 9-Special purpose Workstation; 10-protective cover; 11-power supply; 12-image analysis module; 13-backend database; 14-alarm module.
  • plural means two or more.
  • the orientation or state relationship indicated by the terms “inside”, “upper”, “lower”, etc. is based on the orientation or state relationship shown in the accompanying drawings, and is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying the indicated A device or element must have a particular orientation, be constructed and operate in a particular orientation, and therefore should not be construed as limiting the invention.
  • the terms “installed”, “connected” and “provided with” should be understood in a broad sense, for example, it may be a fixed connection or a Removal connection, or integral connection; it can be a mechanical connection or an electrical connection; it can be directly connected or indirectly connected through an intermediate medium.
  • installed e.g., it may be a fixed connection or a Removal connection, or integral connection; it can be a mechanical connection or an electrical connection; it can be directly connected or indirectly connected through an intermediate medium.
  • a system for diagnosing structural damage of water-passing buildings based on flow-state video monitoring includes a high-definition fog-penetrating camera 1 , a base 3 , a dedicated workstation 9 and a computer monitor 8 ;
  • the high-definition fog-penetrating camera 1 is installed on the base 3; the base 3 is provided with a cavity; the network interface 4 and the power interface 5 of the high-definition fog-penetrating camera 1 are installed in the cavity;
  • a smart wiper module 2 is installed on the high-definition fog-penetrating camera 1; the smart wiper module 2 is used to clean the water droplets on the lens of the high-definition fog-penetrating camera 1 in time to keep the lens clean;
  • the power interface 5, the computer monitor 8, and the dedicated workstation 9 are respectively connected to the power source 11 through a power cord 6;
  • the network interface 4 is connected with the dedicated workstation 9 through the data transmission line 7;
  • the dedicated workstation 9 includes an image analysis module 12, a background database 13 and an alarm module 14;
  • the image analysis module 12 is respectively connected with the background database 13, the alarm module 14 and the computer display 8;
  • the image analysis module 12 is used to process the image captured by the high-definition fog-penetrating camera 1, analyze the outline of the water surface line, and then compare the feature point coordinates of the obtained water surface line outline with the image feature point coordinates corresponding to normal operating conditions in the background database 13, If the comparison result is abnormal, send an instruction to the alarm module 14 for alarming;
  • the background database 13 is also used to store the images before the processing by the image analysis module 12 and the data obtained by processing;
  • the computer monitor 8 is used for displaying the image before the image analysis module 12 processes the image and the data obtained by the process.
  • a protective cover 10 is also included; the protective cover 10 is arranged outside the base 3 and the high-definition fog-penetrating camera 1 , and is used to protect the base 3 and the high-definition fog-penetrating camera 1 .
  • the background database 13 also includes a normal flow state database and an abnormal flow state database; when the image analysis module 12 compares the image water surface line profile captured by the high-definition fog-penetrating camera 1 with the normal flow state data in the background database, the comparison result is abnormal. , the image before the image analysis module 12 and the processed data are stored in the abnormal flow state database; otherwise, they are stored in the normal flow state database.
  • step g Calculate line by line according to step c to step f, respectively connect x i , left and x i , and right as the left contour and right contour of the water surface line.
  • a 0 is 0.85-0.95; n 0 is 1-5; m 0 is 2-5.
  • the method for diagnosing structural damage of water-passing buildings based on flow state video monitoring using the above-mentioned system for diagnosing structural damage of water-passing buildings based on flow state video monitoring, includes the following steps:
  • Step (1) according to the characteristics of the discharge flow, lay out several feature points on the site to cover the entire shooting range as much as possible, place anti-scour and easy-to-recognize signs at the feature points, measure their coordinates, and input them into the background database. ;
  • Step (2) store the discharge flow pattern, water surface line and characteristic point data of the water-passing structure under normal working conditions, as well as the prototype observation data obtained in real time in the field into the normal flow pattern database;
  • the data of the building discharge flow state, water surface line and characteristic point, as well as the abnormal original view data that actually occurred on site, are stored in the abnormal flow state database;
  • Step (3) turn on the high-definition fog-penetrating camera, aim at the discharge flow state of the water-passing building, adjust the shooting height so that the captured image covers the entire water surface and the side walls on both sides of the water-passing building; start the smart wiper module, and adjust the aperture , focal length, exposure parameters and shutter speed to ensure clear images;
  • Step (4) formally shooting after obtaining an image whose resolution meets the requirements, shooting a high-definition image every 10 minutes, and performing image processing and analysis to determine the outline of the water surface line;
  • Step (5) compare the coordinate H i (X, Y) of the feature point of the water surface line profile obtained by the image calculation with the coordinate H' i ( X, Y) of the image feature point corresponding to the normal working condition in the background database, when ⁇ H i (X,Y)-H′ i (X,Y) ⁇ /H′ i (X,Y)>10%, it is regarded as an abnormal flow state, an alarm is activated and the data is stored in the abnormal flow state database; when ⁇ H i When (X,Y)-H′ i (X,Y) ⁇ /H′ i (X,Y) ⁇ 10%, it is regarded as a normal flow state and stored in the normal flow state database;
  • step (6) after the leakage is completed, the high-definition fog-penetrating camera is turned off, and all images are stored in the background database.
  • the device for diagnosing structural damage of water-passing buildings based on flow-state video monitoring includes a high-definition fog-penetrating camera 1 embedded with a smart wiper module 2, a base 3 on which the high-definition fog-penetrating camera 1 is installed, and is installed on the base 3.
  • the network interface 4 and the power interface 5 in the cavity are connected to the data transmission line 7 of the high-definition fog-proof camera 1 and the dedicated workstation 9, and the power line 6 of the high-definition fog-proof camera 1 and the power supply 11 is connected.
  • the workstation 9 is connected to the computer monitor 8 used for displaying the analysis results in the special workstation 9 , and is used to cover the high-definition fog-penetrating camera 1 and the protective cover 10 of the base 3 .
  • the installation operation is as follows: Embed the smart wiper module 2 into the high-definition fog-penetrating camera 1, and determine the high-definition fog-penetrating camera 1 according to the image size, lens focal length, target size and actual conditions on the side perpendicular to the water flow of the water-passing building.
  • a dedicated workstation 9 is installed at the corresponding position, which is connected to the network interface 4 of the high-definition fog-penetrating camera 1 to receive the image data transmitted by the high-definition fog-penetrating camera 1 in real time.
  • the image analysis and processing software system 12 is installed on the special workstation 9, which is used for water surface line analysis, feature point extraction, and abnormal judgment of feature points; on the special workstation 9, a background database 13 is constructed to store various working conditions of water-passing buildings. The flow chart, water surface line and feature point data below; after the discharge is completed, remove the high-definition fog-penetrating camera 1 and place it in a special box, and then cover the protective cover 10 on the base 3.
  • the protective cover 10 Do not cover the protective cover 10 during the shooting process. If the camera is not used for a long time after the shooting, remove the camera and put it in a special box. If the shooting is only temporarily interrupted, such as waiting for the adjustment of the working conditions, the protective cover should be placed on the base 3 and the high-definition fog. on camera 1.
  • a method for identifying and diagnosing structural damage of a water-passing building based on flow state video monitoring, using the above device, and the steps are as follows:

Landscapes

  • Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Physics & Mathematics (AREA)
  • General Health & Medical Sciences (AREA)
  • Multimedia (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Health & Medical Sciences (AREA)
  • Theoretical Computer Science (AREA)
  • Computing Systems (AREA)
  • Analytical Chemistry (AREA)
  • Medical Informatics (AREA)
  • Software Systems (AREA)
  • Databases & Information Systems (AREA)
  • Artificial Intelligence (AREA)
  • Signal Processing (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Chemical & Material Sciences (AREA)
  • Evolutionary Computation (AREA)
  • Biochemistry (AREA)
  • Immunology (AREA)
  • Pathology (AREA)
  • Business, Economics & Management (AREA)
  • Emergency Management (AREA)
  • Image Processing (AREA)
  • Alarm Systems (AREA)
  • Emergency Alarm Devices (AREA)

Abstract

本发明涉及一种基于流态视频监测的过水建筑物结构损伤诊断方法及系统，该系统包括高清透雾摄像机、基座、专用工作站和电脑显示器；高清透雾摄像机安装在基座上；基座内设有空腔；高清透雾摄像机的网络接口、电源接口安装在空腔内；高清透雾摄像机上安装有智能雨刮模块；电源接口、电脑显示器、专用工作站分别通过一根电源线与电源相连；网络接口通过数据传输线与专用工作站相连；专用工作站包括图像分析模块、后台数据库和报警模块。本发明创新了采用视频监测泄流流态和水面线诊断过水建筑物结构是否发生损伤的方法，能够及时发现泄流过程中的异常流态，及时采取措施止损，可在泄流设施安全监测、教研等领域广泛应用。

Description

一种基于流态视频监测的过水建筑物结构损伤诊断方法及系统 技术领域
本发明属于水利水电工程和视频监测技术领域，具体涉及一种基于流态视频监测的过水建筑物结构损伤诊断方法及系统。
背景技术
过水建筑物承担着泄洪、过流的重要任务，尤其汛期洪水流量较大时，过水建筑物处于长时间泄流状态，且大型水电工程的过水建筑物水流流速往往高达40～50m/s。受高速水流长期冲刷作用，过水建筑物发生破坏的几率很高，据不完全统计，有1/3过水建筑物发生了不同程度的破坏，有的相当严重。过水建筑物在泄流时若发生破坏，由于人员巡视不及时、受雾化影响无法临近清晰观察、水流受到结构物遮挡等各种原因导致不能及时发现、及时采取措施而造成损失。随着高清摄像技术的发展，远距离、透雾、高清摄像技术可替代人工实现泄流过程的实时监测，配置高速、高可靠性的图像处理分析技术，能够及时发现泄流流态和水面线的异常状况，进而实现过水建筑物结构损伤诊断的功能。因此，研究基于流态视频监测的过水建筑物结构损伤诊断方法，对于实现过水建筑物运行状态实时监测，确保结构安全，降低经济损失和人工成本，提高监测水平具有重要的理论意义与实用价值。
目前过水建筑物结构损伤诊断方法基本采用传感器技术，主要有以下不足：(1)传感器的布设一般需要在过水建筑物结构体上进行施工，准备工作多，流程复杂，协调方多，耗时长，且施工质量缺陷会成为过水建筑物破坏的新增诱因；(2)传感器设备长期接触高速水流，耐久性较差，失效率较高；(3)传感器监测都是“点”的观测，难以真实、全面地反映泄流结构整体的工作性态变化。因此如何克服现有技术的不足是目前本技术领域亟需解决的问题。
发明内容
本发明的目的是为了解决现有技术的不足，提供一种基于流态视频监测的过水建筑物结构损伤诊断方法及系统。
为实现上述目的，本发明采用的技术方案如下：
一种基于流态视频监测的过水建筑物结构损伤诊断系统，包括高清透雾摄像机、基座、专用工作站和电脑显示器；
高清透雾摄像机安装在基座上；基座内设有空腔；高清透雾摄像机的网络接口、电源接口安装在空腔内；
高清透雾摄像机上安装有智能雨刮模块；智能雨刮模块用于及时清理高清透雾摄像机镜头上的水滴，保持镜头干净；
电源接口、电脑显示器、专用工作站分别通过一根电源线与电源相连；
网络接口通过数据传输线与专用工作站相连；
专用工作站包括图像分析模块、后台数据库和报警模块；
图像分析模块分别与后台数据库、报警模块、电脑显示器相连；
图像分析模块用于处理高清透雾摄像机拍摄到的图像，分析水面线轮廓，然后将得到的水面线轮廓的特征点坐标与后台数据库中对应正常工况的图像特征点坐标进行对比，若对比结果为异常，则发送指令至报警模块进行报警；
后台数据库还用于存储图像分析模块处理前的图像及处理得到的数据；
电脑显示器用于显示图像分析模块处理前的图像及处理得到的数据。
进一步，优选的是，还包括保护罩；保护罩设于基座及高清透雾摄像机外，用于保护基座及高清透雾摄像机。
进一步，优选的是，后台数据库中还包括正常流态数据库和异常流态数据库；当图像分析模块对比高清透雾摄像机拍摄到的图像水面线轮廓与后台数据库中正常流态数据，对比结果为异常时，将图像分析模块处理前的图像及处理得到的数据存储于异常流态数据库；反之，则存储在正常流态数据库。
水面线轮廓分析的具体步骤如下：
a、逐行逐列扫描图像，获取各像素点的RGB颜色值，分别保存为R i,j、G i,j、B i,j，其中i为像素行编号，j为像素的列编号；
b、设定每一行相邻像素点RGB相似阈值a 0，设定不连续对比数阈值n 0，设定连续对比数阈值m 0；
c、从左向右计算相邻像素点相似值：
Figure PCTCN2021136749-appb-000001
若a i,j≥a 0，边墙区域不连续对比数n 边＝0；若a i,j<a 0，则n 边＝n 边+1；当n 边＝n 0时，记录第i行左边墙清晰边界的像素坐标值x i, 左墙；
d、从x i, 左墙继续向右对比，若a i,j≥a 0，水面区域连续对比数m 水＝m 水+1；若a i,j<a 0，则m 水＝0；当m 水＝m 0时，记录第i行水面左清晰边界的像素坐标值x i, 左水，取x i, 左墙和x i, 左水的均值作为水面线左边界坐标x i, 左；
e、从x i, 左水继续向右对比，若a i,j≥a 0，水面区域不连续对比数n 水＝0；若a i,j<a 0，则n 水＝n 水+1；当n 水＝n 0时，记录第i行水面右清晰边界的像素坐标值x i, 右水；
f、从x i, 右水继续向右对比，若a i,j≥a 0，边墙区域连续对比数m 边＝m 边+1；若a i,j<a 0，则m 边＝0；当m 边＝m 0时，记录第i行水面右边墙清晰边界的像素坐标值x i, 右墙，取x i, 右水和x i, 右墙的均值作为水面线左边界坐标x i, 右；
g、按步骤c至步骤f逐行计算，分别连接x i, 左和x i, 右作为水面线左轮廓与右轮廓。
进一步，优选的是，a 0为0.85～0.95；n 0为1～5；m 0为2～5。
进一步，优选的是，将水面线轮廓特征点坐标H i(X,Y)与后台数据库中对应正常工况的图像特征点坐标H′ i(X,Y)对比，当∣H i(X,Y)-H′ i(X,Y)∣/H′ i(X,Y)>10％时视为异常流态，启动报警并将数据存储于异常流态数据库；当∣H i(X,Y)-H′ i(X,Y)∣/H′ i(X,Y)<10％时视为正常流态，存储于正常流态数据库。
本发明同时提供基于流态视频监测的过水建筑物结构损伤诊断方法，使用上述基于流态视频监测的过水建筑物结构损伤诊断系统，包括如下步骤：
步骤(1)，针对泄流流态特征，在现场布设数个特征点，使其尽可能覆盖整个拍摄范围，在特征点处放置抗冲刷、易识别的标识，并测量其坐标，输入后台数据库；
步骤(2)，将正常工况下，过水建筑物的泄流流态、水面线及特征点数据，以及现场实时获得的原型观测数据存储到正常流态数据库；将异常工况下过水建筑物泄流流态、水面线及特征点数据，以及现场实际发生的 异常原观数据，存储到异常流态数据库；
步骤(3)，开启高清透雾摄像机，对准过水建筑物泄流流态，调整拍摄高度至所摄图像覆盖整个水面、过水建筑物两侧边墙；启动智能雨刮模块，调节光圈、焦距、曝光参数及快门速度，确保拍摄图像清晰；
步骤(4)，获得分辨率满足要求的图像后正式拍摄，每隔10分钟拍摄一张高清图像，并进行图像处理分析，确定水面线轮廓；
步骤(5)，将图像计算得到的水面线轮廓特征点坐标H i(X,Y)与后台数据库中对应正常工况的图像特征点坐标H′ i(X,Y)对比，当∣H i(X,Y)-H′ i(X,Y)∣/H′ i(X,Y)>10％时视为异常流态，启动报警并将数据存储于异常流态数据库；当∣H i(X,Y)-H′ i(X,Y)∣/H′ i(X,Y)≤10％时视为正常流态，存储于正常流态数据库；
步骤(6)，泄流结束后，关闭高清透雾摄像机，将所有图像存储于后台数据库。
本发明系统在使用时，水面线轮廓与人工观测的情况对比，若符合情况不佳，可以修改a 0、n 0、m 0，重复步骤c至步骤f，直至满足分析精度要求。
本发明中，特征点选择原则是，尽可能均匀覆盖整个水面线轮廓，且包含一定数量的已知坐标的固定点。
本发明中，异常流态数据库主要是为以后分析损伤特性与泄流流态之间的相关关系积累数据，同时数据量多时也可用于复核对比频繁出现的异常工况的数据。
本发明所诊断的结构损伤是指结构发生的破坏程度包括冲蚀坑、磨蚀坑、气蚀坑等对水流流态已造成的影响超出了容忍范围即10％。
本发明中，高清透雾摄像机具有防雨、透雾、抗振功能，采用较高的快门速度，满足40～50m/s高速动态水流的拍摄效果，实现在高速水流雾化区的清晰拍摄功能。
基座应根据图像尺寸、镜头焦距、目标物尺寸及现场实际条件确定安装位置。基座具有稳定性和防冲刷特性，基座内布设网络接口和电源接口，通过电源接口可为高清透雾摄像机充电，将拍摄的图像实时传输到专用工作站。
专用工作站具有强大的图像处理能力、大容量、高速度、高可靠性。
本发明中，所述特征点是指为提高图像配准精度而提取图像中目标清晰、易于识别、位置不发生变动的特征点。经图像分析后水面线差异大于规定阈值后启动报警功能。
后台数据库中存储有原型观测、数值模拟、物理模型试验等方法获得的、经验证测量精度满足要求的过水建筑物泄流流态图片、水面线数据及特征点数据，包括过水建筑物正常工作状态数据库和异常工作状态数据库。
高清透雾摄像机和基座应配备保护罩，用于防止野外工作时出现恶劣天气条件或人为损坏情况。
图像分析模块的图像处理包括水面线分析、特征点提取、特征点异常判断；所述后台数据库存储过水建筑物各种工况下的流态图、水面线及特征点数据。
本发明系统的安装操作为：在垂直于过水建筑物水流流向一侧，根据图像尺寸、镜头焦距、目标物尺寸及现场实际条件确定拍摄位置；在拍摄位置安装基座；高清透雾摄像机置于基座上固定好，连接好电源接口和网络接口；根据传输距离要求和现场网络情况，在相应位置安装专用工作站，与高清透雾摄像机网络接口连接，用于实时接收高清透雾摄像机传输的图像数据。
本发明高清透雾摄像机的高清要求分辨率不低于2560*1440。
本发明与现有技术相比，其有益效果为：
(1)实现过水建筑物结构健康状态的无损非接触式观测。本发明采用远距离、透雾、高清摄像技术对泄流流态进行拍摄，利用结构损伤对水流流态会造成一定影响的因果关系进行结构健康诊断，不需对过水建筑物结构本身进行有损施工，避免了在结构上因安装传感器而形成新的缺陷诱因的弊端。
(2)可全面、真实反映泄流结构整体的健康状态。本发明不对某些“点”进行观测，而是观测整体的泄流流态，因为任何点的缺陷、具有一定威胁的缺陷必然会导致泄流流态的改变，因此通过观测泄流流态可判断出结构本身是否发生了破坏。
(3)经济、方便、快捷。本发明不需协调各方进行审批和调度等，准 备工作量小，流程简单，所需人员少，设备可循环使用，人工成本和设备成本低。
(4)可靠性高，耐久性强。本发明不直接接触高速水流，避免了传感器受高速水流冲刷而造成的失效等问题，高清透雾摄像机在拍摄时开启，不使用时予以保护，可靠性高，耐久性强。
综上所述，本发明提供的基于流态视频监测的过水建筑物结构损伤诊断方法具有经济、简便、快捷、可靠性高，耐久性强，可全面、真实反映泄流结构健康状态，能对过水建筑物结构健康状态进行无损非接触式实时观测的优点，在水电工程及其他行业过水建筑物安全监测、高速水流流态监测、科研、教育等领域广泛应用。
附图说明
为了更清楚地说明本发明实施例或现有技术中的技术方案，下面将对实施例或现有技术描述中所需要使用的附图作简单地介绍，显而易见地，下面描述中的附图仅仅是本发明的一些实施例，对于本领域普通技术人员来讲，在不付出创造性劳动性的前提下，还可以根据这些附图获得其他的附图。
图1是本发明基于流态视频监测的过水建筑物结构损伤诊断系统的结构示意图；
图2是本发明软硬件系统架构图；
图3是基于流态视频监测的过水建筑物结构损伤识别诊断流程示意图；
图4是水面线轮廓分析流程示意图；
图5是过水建筑物泄流流态现场实拍图；
图6是水面线轮廓分析处理结果图；
图7为水面及其轮廓线图。
图中，1-高清透雾摄像机；2-智能雨刮模块；3-基座；4-网络接口；5-电源接口；6-电源线；7-数据传输线；8-电脑显示器；9-专用工作站；10-保护罩；11-电源；12-图像分析模块；13-后台数据库；14-报警模块。
具体实施方式
下面结合实施例对本发明作进一步的详细描述。
本领域技术人员将会理解，下列实施例仅用于说明本发明，而不应视 为限定本发明的范围。实施例中未注明具体技术或条件者，按照本领域内的文献所描述的技术或条件或者按照产品说明书进行。所用材料或设备未注明生产厂商者，均为可以通过购买获得的常规产品。
本技术领域技术人员可以理解，除非特意声明，这里使用的单数形式“一”、“一个”、“所述”和“该”也可包括复数形式。应该进一步理解的是，本发明的说明书中使用的措辞“包括”是指存在所述特征、整数、步骤、操作、元件和/或组件，但是并不排除存在或添加一个或多个其他特征、整数、步骤、操作、元件、组件和/或它们的组。应该理解，当我们称元件被“连接”到另一元件时，它可以直接连接到其他元件，或者也可以存在中间元件。此外，这里使用的“连接”可以包括无线连接。
在本发明的描述中，除非另有说明，“多个”的含义是两个或两个以上。术语“内”、“上”、“下”等指示的方位或状态关系为基于附图所示的方位或状态关系，仅是为了便于描述本发明和简化描述，而不是指示或暗示所指的装置或元件必须具有特定的方位、以特定的方位构造和操作，因此不能理解为对本发明的限制。
在本发明的描述中，需要说明的是，除非另有明确的规定和限定，术语“安装”、“连接”、“设有”应做广义理解，例如，可以是固定连接，也可以是可拆卸连接，或一体地连接；可以是机械连接，也可以是电连接；可以是直接相连，也可以通过中间媒介间接相连。对于本领域的普通技术人员而言，根据具体情况理解上述术语在本发明中的具体含义。
本技术领域技术人员可以理解，除非另外定义，这里使用的所有术语(包括技术术语和科学术语)具有与本发明所属领域中的普通技术人员的一般理解相同的意义。还应该理解的是，诸如通用字典中定义的那些术语应该被理解为具有与现有技术的上下文中的意义一致的意义，并且除非像这里一样定义，不会用理想化或过于正式的含义来解释。
如图1所示，一种基于流态视频监测的过水建筑物结构损伤诊断系统，包括高清透雾摄像机1、基座3、专用工作站9和电脑显示器8；
高清透雾摄像机1安装在基座3上；基座3内设有空腔；高清透雾摄像机1的网络接口4、电源接口5安装在空腔内；
高清透雾摄像机1上安装有智能雨刮模块2；智能雨刮模块2用于及时 清理高清透雾摄像机1镜头上的水滴，保持镜头干净；
电源接口5、电脑显示器8、专用工作站9分别通过一根电源线6与电源11相连；
网络接口4通过数据传输线7与专用工作站9相连；
专用工作站9包括图像分析模块12、后台数据库13和报警模块14；
图像分析模块12分别与后台数据库13、报警模块14、电脑显示器8相连；
图像分析模块12用于处理高清透雾摄像机1拍摄到的图像，分析水面线轮廓，然后将得到的水面线轮廓的特征点坐标与后台数据库13中对应正常工况的图像特征点坐标进行对比，若对比结果为异常，则发送指令至报警模块14进行报警；
后台数据库13还用于存储图像分析模块12处理前的图像及处理得到的数据；
电脑显示器8用于显示图像分析模块12处理前的图像及处理得到的数据。
优先，还包括保护罩10；保护罩10设于基座3及高清透雾摄像机1外，用于保护基座3及高清透雾摄像机1。
优先，后台数据库13中还包括正常流态数据库和异常流态数据库；当图像分析模块12对比高清透雾摄像机1拍摄到的图像水面线轮廓与后台数据库中正常流态数据，对比结果为异常时，将图像分析模块12处理前的图像及处理得到的数据存储于异常流态数据库；反之，则存储在正常流态数据库。
水面线轮廓分析的具体步骤如下：
a、逐行逐列扫描图像，获取各像素点的RGB颜色值，分别保存为R i,j、G i,j、B i,j，其中i为像素行编号，j为像素的列编号；
b、设定每一行相邻像素点RGB相似阈值a 0，设定不连续对比数阈值n 0，设定连续对比数阈值m 0；
c、从左向右计算相邻像素点相似值：
Figure PCTCN2021136749-appb-000002
若a i,j≥a 0，边墙区域不连续对比数n 边＝0；若a i,j<a 0，则n 边＝n 边+1；当n 边＝n 0时，记录第i行左边墙清晰边界的像素坐标值x i, 左墙；
d、从x i, 左墙继续向右对比，若a i,j≥a 0，水面区域连续对比数m 水＝m 水+1；若a i,j<a 0，则m 水＝0；当m 水＝m 0时，记录第i行水面左清晰边界的像素坐标值x i, 左水，取x i, 左墙和x i, 左水的均值作为水面线左边界坐标x i, 左；
e、从x i, 左水继续向右对比，若a i,j≥a 0，水面区域不连续对比数n 水＝0；若a i,j<a 0，则n 水＝n 水+1；当n 水＝n 0时，记录第i行水面右清晰边界的像素坐标值x i, 右水；
f、从x i, 右水继续向右对比，若a i,j≥a 0，边墙区域连续对比数m 边＝m 边+1；若a i,j<a 0，则m 边＝0；当m 边＝m 0时，记录第i行水面右边墙清晰边界的像素坐标值x i, 右墙，取x i, 右水和x i, 右墙的均值作为水面线左边界坐标x i, 右；
g、按步骤c至步骤f逐行计算，分别连接x i, 左和x i, 右作为水面线左轮廓与右轮廓。
优先，a 0为0.85～0.95；n 0为1～5；m 0为2～5。
优先，将水面线轮廓特征点坐标H i(X,Y)与后台数据库13中对应正常工况的图像特征点坐标H′ i(X,Y)对比，当∣H i(X,Y)-H′ i(X,Y)∣/H′ i(X,Y)>10％时视为异常流态，启动报警并将数据存储于异常流态数据库；当∣H i(X,Y)-H′ i(X,Y)∣/H′ i(X,Y)<10％时视为正常流态，存储于正常流态数据库。
基于流态视频监测的过水建筑物结构损伤诊断方法，使用上述基于流态视频监测的过水建筑物结构损伤诊断系统，包括如下步骤：
步骤(1)，针对泄流流态特征，在现场布设数个特征点，使其尽可能覆盖整个拍摄范围，在特征点处放置抗冲刷、易识别的标识，并测量其坐标，输入后台数据库；
步骤(2)，将正常工况下，过水建筑物的泄流流态、水面线及特征点数据，以及现场实时获得的原型观测数据存储到正常流态数据库；将异常工况下过水建筑物泄流流态、水面线及特征点数据，以及现场实际发生的异常原观数据，存储到异常流态数据库；
步骤(3)，开启高清透雾摄像机，对准过水建筑物泄流流态，调整拍摄高度至所摄图像覆盖整个水面、过水建筑物两侧边墙；启动智能雨刮模 块，调节光圈、焦距、曝光参数及快门速度，确保拍摄图像清晰；
步骤(4)，获得分辨率满足要求的图像后正式拍摄，每隔10分钟拍摄一张高清图像，并进行图像处理分析，确定水面线轮廓；
步骤(5)，将图像计算得到的水面线轮廓特征点坐标H i(X,Y)与后台数据库中对应正常工况的图像特征点坐标H′ i(X,Y)对比，当∣H i(X,Y)-H′ i(X,Y)∣/H′ i(X,Y)>10％时视为异常流态，启动报警并将数据存储于异常流态数据库；当∣H i(X,Y)-H′ i(X,Y)∣/H′ i(X,Y)≤10％时视为正常流态，存储于正常流态数据库；
步骤(6)，泄流结束后，关闭高清透雾摄像机，将所有图像存储于后台数据库。
应用实例1
如图1所示，基于流态视频监测的过水建筑物结构损伤诊断装置包括嵌入有智能雨刮模块2的高清透雾摄像机1，安装高清透雾摄像机1的基座3，安装在基座3空腔内的网络接口4和电源接口5，连接高清透雾摄像机1和专用工作站9的数据传输线7，连接高清透雾摄像机1和电源11的电源线6，安装有图像分析软件12的专用工作站9，连接专用工作站9用于显示分析结果的电脑显示器8，用于罩住高清透雾摄像机1和基座3的保护罩10。
安装操作为：将智能雨刮模块2嵌入到高清透雾摄像机1，在垂直于过水建筑物水流流向一侧，根据图像尺寸、镜头焦距、目标物尺寸及现场实际条件确定高清透雾摄像机1的安装位置；在拍摄位置安装基座3，基座3下面设置为空腔，空腔内设置有网络接口4和电源接口5；高清透雾摄像机1置于基座3上固定好，连接好电源接口5和网络接口4；根据传输距离要求和现场网络情况，在相应位置安装专用工作站9，与高清透雾摄像机1的网络接口4连接，用于实时接收高清透雾摄像机1传输的图像数据；专用工作站9上安装图像分析处理软件系统12，用于进行水面线分析、特征点提取、特征点异常判断；在专用工作站9上构建后台数据库13，用于存储过水建筑物各种工况下的流态图、水面线及特征点数据；泄流结束后，将高清透雾摄像机1拆下放置于专用箱后，将保护罩10盖于基座3上。
拍摄过程不罩保护罩10，拍摄结束后如长期不用摄像机，要将摄像机 拆下放在专用箱，如果拍摄只是暂时中断比如等待工况调整等，要将保护罩罩在基座3和高清透雾摄像机1上。
一种基于流态视频监测的过水建筑物结构损伤识别诊断方法，使用上述装置，其步骤如下：
(1)接通电源线6和数据传输线7，开启高清透雾摄像机1，对准过水建筑物泄流流态，调整拍摄高度、拍摄角度，试拍数张图像，拍摄图像应覆盖整个水面、过水建筑物两侧边墙；
(2)启动智能雨刮模块2，调节光圈、焦距、曝光参数及快门速度，试拍数张照片，确保拍摄图像清晰；
(3)按照调整好的拍摄高度、角度及拍摄参数，每隔10分钟拍摄一张高清图像，传输至专用工作站9，运行图像处理分析软件12进行图像处理分析；
(4)在电脑显示器8上实时观察图像处理分析软件12的分析结果，及时处理预警异常信息；
(5)图像拍摄完成后，关闭智能雨刮模块2、高清透雾摄像机1，将高清透雾摄像机1拆下放置于专用箱；将保护罩10盖于基座3上；
(6)确定水面线轮廓分析参数，其步骤如下，过水建筑物泄流流态现场实拍如图5所示，以此照片为例进行水面线分析。
a、逐行逐列扫描图像，获取各像素点的RGB颜色值，分别保存为R i,j、G i,j、B i,j，其中i为像素行编号(i＝1～2101)，j为像素的列编号(j＝1～1931)；
b、实施例对象图5水面线为上下型，故需逐列对比图像像素值，设定每一列相邻像素点RGB相似阈值a 0＝0.95，设定不连续对比数阈值n 0＝1，设定连续对比数阈值m 0＝3；
c、从上向下计算相邻像素点相似值，以第1列第1行和第2行的像素点对比为例，R 1,1＝88、G 1,1＝116、B 1,1＝156，R 2,1＝86、G 2,1＝114、B 2,1＝154，则
Figure PCTCN2021136749-appb-000003
a 1,1≥a 0，边墙区域不连续对比数n 边＝0，继续向下对比，当j＝713时，a 713,1＝0.94<a 0，此时n 边＝1＝n 0，记录第1列上边界清晰边界的像素坐标值x 上边界, 1＝713；
d、从i＝x 1, 上边界继续向下对比，a 714,1＝0.91<a 0，则水面区域连续对比数 m 水＝0，a 715,1＝0.86<a 0，则m 水＝0，a 716,1＝0.93<a 0，则m 水＝0，a 717,1＝0.98≥a 0，则m 水＝1，a 718,1＝0.99≥a 0，则m 水＝2，a 719,1＝0.99≥a 0，此时m 水＝3＝m 0，记录第1列水面上清晰边界的像素坐标值x 上水, 1＝719，取x 上边界, 1和x 上水, 1的均值x 上, 1＝716作为水面线上端坐标；
e、从i＝x 上水, 1继续向下对比，一直到j＝1019时，a 1019,1＝0.94<a 0，此时n 水＝1＝n 0时，记录第1列水面下清晰边界的像素坐标值x 下水, 1＝1019；
f、从i＝x 下水, 1继续向下对比，a 1020,1＝0.94<a 0，则水面区域连续对比数m 水＝0，a 1021,1＝0.95≥a 0，则m 水＝1，a 1022,1＝0.96≥a 0，则m 水＝2，a 1023,1＝0.98≥a 0，则m 水＝3，此时m 水＝m 0，记录第1列下清晰边界的像素坐标值x 下边 界, 1＝1023，取x 下边界, 1和x 下水, 1的均值x 下, 1＝1021作为水面线下端坐标；
g、按步骤c至步骤f逐列计算，连接x 上, j(j＝1～1931)作为水面线上轮廓，连接x 下, j(j＝1～1931)作为水面线下轮廓，如图6上下两条黑线所示，通过人工观测对比发现，识别出的水面线轮廓较好包络了图中白色高掺气水体，识别精度满足分析要求。
(7)以图6所示的水面线上轮廓线为例提取特征点，沿x方向选取10个特征点，均匀覆盖整条轮廓线，提取到的特征点坐标依次为(1，716)、(211，622)、(421，549)、(631，556)、(841，557)、(1051，588)、(1261，647)、(1471，740)、(1681，811)、(1891，951)。
(8)将提取到的特征点坐标与后台数据库中正常状态下的坐标进行对比，对比时选择X坐标一致，结果如表1所示，可看出提取到的特征点Y坐标与后台数据库中正常状态下坐标Y’相差均在10％以内，故此时属于正常状态，不启动报警，将坐标数据存储于正常流态数据库。
表1
Figure PCTCN2021136749-appb-000004
Figure PCTCN2021136749-appb-000005
应用实例2
应用实例2与应用实例1的区别在于最后两步不同，其余皆相同。
(7)以图7所示的水面线右轮廓线为例提取特征点，沿y方向选取10个特征点，均匀覆盖整条轮廓线，提取到的特征点坐标依次为(207，5)、(209，45)、(210，85)、(212，125)、(221，165)、(260，205)、(314，245)、(294，285)、(299，325)、(366，365)、(392，405)。
(8)将提取到的特征点坐标与后台数据库中正常状态下的坐标进行对比，对比时选择Y坐标一致，结果如表2所示，可看出从Y＝205开始，提取到的特征点X坐标与后台数据库中正常状态下坐标X’相差超过10％，甚至达到了20～44％，说明此时结构发生了较严重破坏，启动报警并将数据存储于异常流态数据库。
表2
Figure PCTCN2021136749-appb-000006
现有方法基本是靠肉眼识别，肉眼识别存在不及时、不准确的缺点， 现场巡视人员观察会受雾化影响、结构物遮挡等各种原因无法临近清晰观察，且受身体耐受性影响不能24小时不间断观察，往往是泄流结束后再去现场观察，导致不能及时发现异常现象。而本发明的高清透雾摄像机，配置智能雨刮模块，可解决雾化水滴对清晰度的影响，并可安置在不受结构物遮挡的位置，获得最佳观测视野，且可24小时不间断观测。
以上显示和描述了本发明的基本原理、主要特征和本发明的优点。本行业的技术人员应该了解，本发明不受上述实施例的限制，上述实施例和说明书中描述的只是说明本发明的原理，在不脱离本发明精神和范围的前提下，本发明还会有各种变化和改进，这些变化和改进都落入要求保护的本发明范围内。本发明要求保护范围由所附的权利要求书及其等效物界定。

Claims (7)

  1. 一种基于流态视频监测的过水建筑物结构损伤诊断系统，其特征在于，包括高清透雾摄像机(1)、基座(3)、专用工作站(9)和电脑显示器(8)；
    高清透雾摄像机(1)安装在基座(3)上；基座(3)内设有空腔；高清透雾摄像机(1)的网络接口(4)、电源接口(5)安装在空腔内；
    高清透雾摄像机(1)上安装有智能雨刮模块(2)；智能雨刮模块(2)用于及时清理高清透雾摄像机(1)镜头上的水滴，保持镜头干净；
    电源接口(5)、电脑显示器(8)、专用工作站(9)分别通过一根电源线(6)与电源(11)相连；
    网络接口(4)通过数据传输线(7)与专用工作站(9)相连；
    专用工作站(9)包括图像分析模块(12)、后台数据库(13)和报警模块(14)；
    图像分析模块(12)分别与后台数据库(13)、报警模块(14)、电脑显示器(8)相连；
    图像分析模块(12)用于处理高清透雾摄像机(1)拍摄到的图像，分析水面线轮廓，然后将得到的水面线轮廓的特征点坐标与后台数据库(13)中对应正常工况的图像特征点坐标进行对比，若对比结果为异常，则发送指令至报警模块(14)进行报警；
    后台数据库(13)还用于存储图像分析模块(12)处理前的图像及处理得到的数据；
    电脑显示器(8)用于显示图像分析模块(12)处理前的图像及处理得到的数据。
  2. 根据权利要求1所述的基于流态视频监测的过水建筑物结构损伤诊断系统，其特征在于，还包括保护罩(10)；保护罩(10)设于基座(3)及高清透雾摄像机(1)外，用于保护基座(3)及高清透雾摄像机(1)。
  3. 根据权利要求1所述的基于流态视频监测的过水建筑物结构损伤诊断系统，其特征在于，后台数据库(13)中还包括正常流态数据库和异常流态数据库；当图像分析模块(12)对比高清透雾摄像机(1)拍摄到的图 像水面线轮廓与后台数据库中正常流态数据，对比结果为异常时，将图像分析模块(12)处理前的图像及处理得到的数据存储于异常流态数据库；反之，则存储在正常流态数据库。
  4. 根据权利要求1所述的基于流态视频监测的过水建筑物结构损伤诊断系统，其特征在于，水面线轮廓分析的具体步骤如下：
    a、逐行逐列扫描图像，获取各像素点的RGB颜色值，分别保存为R i,j、G i,j、B i,j，其中i为像素行编号，j为像素的列编号；
    b、设定每一行相邻像素点RGB相似阈值a 0，设定不连续对比数阈值n 0，设定连续对比数阈值m 0；
    c、从左向右计算相邻像素点相似值：
    Figure PCTCN2021136749-appb-100001
    若a i,j≥a 0，边墙区域不连续对比数n 边＝0；若a i,j<a 0，则n 边＝n 边+1；当n 边＝n 0时，记录第i行左边墙清晰边界的像素坐标值x i,左墙；
    d、从x i,左墙继续向右对比，若a i,j≥a 0，水面区域连续对比数m 水＝m 水+1；若a i,j<a 0，则m 水＝0；当m 水＝m 0时，记录第i行水面左清晰边界的像素坐标值x i,左水，取x i,左墙和x i,左水的均值作为水面线左边界坐标x i,左；
    e、从x i,左水继续向右对比，若a i,j≥a 0，水面区域不连续对比数n 水＝0；若a i,j<a 0，则n 水＝n 水+1；当n 水＝n 0时，记录第i行水面右清晰边界的像素坐标值x i,右水；
    f、从x i,右水继续向右对比，若a i,j≥a 0，边墙区域连续对比数m 边＝m 边+1；若a i,j<a 0，则m 边＝0；当m 边＝m 0时，记录第i行水面右边墙清晰边界的像素坐标值x i,右墙，取x i,右水和x i,右墙的均值作为水面线左边界坐标x i,右；
    g、按步骤c至步骤f逐行计算，分别连接x i,左和x i,右作为水面线左轮廓与右轮廓。
  5. 根据权利要求4所述的基于流态视频监测的过水建筑物结构损伤诊断系统，其特征在于，a 0为0.85～0.95；n 0为1～5；m 0为2～5。
  6. 根据权利要求1所述的基于流态视频监测的过水建筑物结构损伤诊断系统，其特征在于，将水面线轮廓特征点坐标H i(X,Y)与后台数据库(13)中对应正常工况的图像特征点坐标H′ i(X,Y)对比，当∣H i(X,Y)-H′ i(X,Y) ∣/H′ i(X,Y)>10％时视为异常流态，启动报警并将数据存储于异常流态数据库；当∣H i(X,Y)-H′ i(X,Y)∣/H′ i(X,Y)<10％时视为正常流态，存储于正常流态数据库。
  7. 基于流态视频监测的过水建筑物结构损伤诊断方法，使用权利要求1至6中任意一项所述的基于流态视频监测的过水建筑物结构损伤诊断系统，其特征在于，包括如下步骤：
    步骤(1)，针对泄流流态特征，在现场布设数个特征点，使其尽可能覆盖整个拍摄范围，在特征点处放置抗冲刷、易识别的标识，并测量其坐标，输入后台数据库；
    步骤(2)，将正常工况下，过水建筑物的泄流流态、水面线及特征点数据，以及现场实时获得的原型观测数据存储到正常流态数据库；将异常工况下过水建筑物泄流流态、水面线及特征点数据，以及现场实际发生的异常原观数据，存储到异常流态数据库；
    步骤(3)，开启高清透雾摄像机，对准过水建筑物泄流流态，调整拍摄高度至所摄图像覆盖整个水面、过水建筑物两侧边墙；启动智能雨刮模块，调节光圈、焦距、曝光参数及快门速度，确保拍摄图像清晰；
    步骤(4)，获得分辨率满足要求的图像后正式拍摄，每隔10分钟拍摄一张高清图像，并进行图像处理分析，确定水面线轮廓；
    步骤(5)，将图像计算得到的水面线轮廓特征点坐标H i(X,Y)与后台数据库中对应正常工况的图像特征点坐标H′ i(X,Y)对比，当∣H i(X,Y)-H′ i(X,Y)∣/H′ i(X,Y)>10％时视为异常流态，启动报警并将数据存储于异常流态数据库；当∣H i(X,Y)-H′ i(X,Y)∣/H′ i(X,Y)≤10％时视为正常流态，存储于正常流态数据库；
    步骤(6)，泄流结束后，关闭高清透雾摄像机，将所有图像存储于后台数据库。
PCT/CN2021/136749 2020-12-17 2021-12-09 一种基于流态视频监测的过水建筑物结构损伤诊断方法及系统 WO2022127683A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
NO20230279A NO20230279A1 (en) 2020-12-17 2021-12-09 Method and system for diagnosing structural damage of overwater building on basis of flow-state video monitoring

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202011503647.8 2020-12-17
CN202011503647.8A CN113155860A (zh) 2020-12-17 2020-12-17 一种基于流态视频监测的过水建筑物结构损伤诊断方法及系统

Publications (1)

Publication Number Publication Date
WO2022127683A1 true WO2022127683A1 (zh) 2022-06-23

Family

ID=76882691

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2021/136749 WO2022127683A1 (zh) 2020-12-17 2021-12-09 一种基于流态视频监测的过水建筑物结构损伤诊断方法及系统

Country Status (3)

Country Link
CN (1) CN113155860A (zh)
NO (1) NO20230279A1 (zh)
WO (1) WO2022127683A1 (zh)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115879084A (zh) * 2023-02-28 2023-03-31 天津辰航安全技术服务有限公司 基于图像分析的安全隐患模拟咨询平台
CN116528036A (zh) * 2023-07-04 2023-08-01 北京亚琪夜游科技有限公司 一种基于智慧文旅的景区安全实时监控系统及方法
CN117057778A (zh) * 2023-08-15 2023-11-14 上海韦地科技集团有限公司 基于核工业物联网的多功能摄像机数据监管系统及方法

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113155860A (zh) * 2020-12-17 2021-07-23 华能澜沧江水电股份有限公司 一种基于流态视频监测的过水建筑物结构损伤诊断方法及系统
CN114202894B (zh) * 2021-11-04 2024-01-30 浙江省水利河口研究院(浙江省海洋规划设计研究院) 一种用于桥墩的多目标联合监测预警装置及其方法、桩基

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN201993108U (zh) * 2010-12-24 2011-09-28 成都飞辉虹科技有限责任公司 尾矿库在线智能监测系统
CN103325216A (zh) * 2012-03-23 2013-09-25 罗普特（厦门）科技集团有限公司 基于视频监控的水利防汛监控预警方法与系统
CN106123785A (zh) * 2016-07-25 2016-11-16 肖锐 一种用于水利水电工程的拱坝监控系统
CN109302692A (zh) * 2018-10-26 2019-02-01 广州益牛科技有限公司 大型拦水坝结构损伤智能实时监控系统
CN109612527A (zh) * 2018-12-20 2019-04-12 联誉信息股份有限公司 水电站生态泄流监控系统
CN110646434A (zh) * 2019-08-21 2020-01-03 南京理工大学 视频自动实时监控水坝破损变形的装置和方法
CN113155860A (zh) * 2020-12-17 2021-07-23 华能澜沧江水电股份有限公司 一种基于流态视频监测的过水建筑物结构损伤诊断方法及系统

Family Cites Families (20)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2000293696A (ja) * 1999-04-07 2000-10-20 Matsushita Electric Ind Co Ltd 画像認識装置
JP4703504B2 (ja) * 2006-07-21 2011-06-15 ソニー株式会社 画像処理装置、画像処理方法、及びプログラム
CN102419196B (zh) * 2011-09-02 2013-02-06 广西远长公路桥梁工程有限公司 利用多个水位遥控检测装置自动绘制河流水面线的方法
CN102758416A (zh) * 2012-07-25 2012-10-31 长江水利委员会长江科学院 一种研究泄洪雾化雾源分布规律的装置
KR101570665B1 (ko) * 2012-08-30 2015-11-30 이우성 수해조절장치시스템
KR101426378B1 (ko) * 2013-03-19 2014-08-07 권순각 깊이 정보를 이용한 프레젠테이션 이벤트 처리 장치 및 방법
KR101500267B1 (ko) * 2014-04-04 2015-03-06 경성대학교 산학협력단 시공간 영상의 화소 강도의 시간적 변화 특성 분석 결과를 이용한 하천 수표면 검출 방법
JP2017139637A (ja) * 2016-02-04 2017-08-10 三井不動産株式会社 監視システム及び監視方法
CN205421200U (zh) * 2016-03-21 2016-08-03 中国电建集团成都勘测设计研究院有限公司 泄水建筑物出口挑流消能结构
US11107025B2 (en) * 2016-12-13 2021-08-31 STREAM METHODS, Inc. System and method for producing and distributing information relevant to water events
CN107622113B (zh) * 2017-09-18 2020-01-10 广西宏智科技有限公司 水库安全智能化管理系统
CN108535269A (zh) * 2018-04-24 2018-09-14 中国水利水电科学研究院 基于薄层水流水动力特性的溢洪道安全检测系统和方法
CN108805023B (zh) * 2018-04-28 2023-12-19 平安科技（深圳）有限公司 一种图像检测方法、装置、计算机设备及存储介质
CN109359533B (zh) * 2018-09-12 2021-06-18 浙江海洋大学 一种基于多波段遥感影像的海岸线提取方法
CN109443476B (zh) * 2018-10-17 2021-07-09 水利部交通运输部国家能源局南京水利科学研究院 一种水位波动过程非接触量测装置及方法
CN210104664U (zh) * 2019-04-04 2020-02-21 天津大学 一种测定低气压环境模型水舌外缘挑距空中移位量的装置
CN110414333A (zh) * 2019-06-20 2019-11-05 平安科技（深圳）有限公司 一种图像边界的检测方法及装置
CN110503807A (zh) * 2019-08-25 2019-11-26 武汉和时利自动化系统工程有限公司 一种洪水水灾后水体变化监控系统及其方法
CN111896950B (zh) * 2020-08-06 2021-06-15 华能澜沧江水电股份有限公司 一种基于地基雷达的滑坡灾害应急监测方法
CN112040223B (zh) * 2020-08-25 2022-08-12 RealMe重庆移动通信有限公司 一种图像处理方法、终端设备及存储介质

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN201993108U (zh) * 2010-12-24 2011-09-28 成都飞辉虹科技有限责任公司 尾矿库在线智能监测系统
CN103325216A (zh) * 2012-03-23 2013-09-25 罗普特（厦门）科技集团有限公司 基于视频监控的水利防汛监控预警方法与系统
CN106123785A (zh) * 2016-07-25 2016-11-16 肖锐 一种用于水利水电工程的拱坝监控系统
CN109302692A (zh) * 2018-10-26 2019-02-01 广州益牛科技有限公司 大型拦水坝结构损伤智能实时监控系统
CN109612527A (zh) * 2018-12-20 2019-04-12 联誉信息股份有限公司 水电站生态泄流监控系统
CN110646434A (zh) * 2019-08-21 2020-01-03 南京理工大学 视频自动实时监控水坝破损变形的装置和方法
CN113155860A (zh) * 2020-12-17 2021-07-23 华能澜沧江水电股份有限公司 一种基于流态视频监测的过水建筑物结构损伤诊断方法及系统

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115879084A (zh) * 2023-02-28 2023-03-31 天津辰航安全技术服务有限公司 基于图像分析的安全隐患模拟咨询平台
CN116528036A (zh) * 2023-07-04 2023-08-01 北京亚琪夜游科技有限公司 一种基于智慧文旅的景区安全实时监控系统及方法
CN116528036B (zh) * 2023-07-04 2023-09-08 李丽 一种基于智慧文旅的景区安全实时监控系统及方法
CN117057778A (zh) * 2023-08-15 2023-11-14 上海韦地科技集团有限公司 基于核工业物联网的多功能摄像机数据监管系统及方法

Also Published As

Publication number Publication date
NO20230279A1 (en) 2023-03-15
CN113155860A (zh) 2021-07-23

Similar Documents

Publication Publication Date Title
WO2022127683A1 (zh) 一种基于流态视频监测的过水建筑物结构损伤诊断方法及系统
CN105784710B (zh) 一种基于数字图像处理的混凝土桥梁裂缝检测装置
WO2019051961A1 (zh) 一种管道检测方法、装置以及存储介质
Schmugge et al. Crack segmentation by leveraging multiple frames of varying illumination
KR101086142B1 (ko) 카메라 영상신호를 이용한 누설판별 방법 및 시스템
CN110646434A (zh) 视频自动实时监控水坝破损变形的装置和方法
CN104700395A (zh) 一种构造物外观裂缝检测方法及系统
CN204758083U (zh) 一种基于图像处理的河道水位报警装置
CN204405556U (zh) 一种基于数字图像处理的混凝土桥梁裂缝检测装置
CN115060343B (zh) 一种基于点云的河流水位检测系统、检测方法
Liang et al. Research on concrete cracks recognition based on dual convolutional neural network
CN113688817A (zh) 一种自动巡检的仪表识别方法及识别系统
TWI396832B (zh) Liquid level identification method
CN202024878U (zh) 风挡雨刷刮净能力的测试装置
CN115082849A (zh) 基于深度学习的模板支架安全智能监测方法
CN103578114A (zh) 基于图像处理的喷煤支管测堵防堵方法
CN110378866A (zh) 一种基于无人机巡检的渠道衬砌破损图像识别方法
CN114022435A (zh) 基于机器视觉的管路泄漏识别系统建立方法
CN111573200B (zh) 一种井下皮带转接处堵塞的视觉检测系统及检测方法
JP3500425B2 (ja) 可視画像式路面状況把握装置における路面状況判定方法
Vijayan et al. A survey on surface crack detection in concretes using traditional, image processing, machine learning, and deep learning techniques
CN114283126B (zh) 一种输电线路监控设备偏移检测方法
CN114943691A (zh) 一种基于机器视觉和循环卷积网络的古城墙损伤实时检测方法
CN110969103B (zh) 一种基于ptz相机的高速公路路面病害长度测量方法
CN114723691A (zh) 基于人工智能的液压系统漏油故障程度检测方法

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 21905612

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 21905612

Country of ref document: EP

Kind code of ref document: A1

122 Ep: pct application non-entry in european phase

Ref document number: 21905612

Country of ref document: EP

Kind code of ref document: A1