JP2008057994A - Water level observation system by image processing - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To solve the problems wherein the number of observation points is insufficient, and future water level prediction is impossible only by observation at the present time because the measurement is of a relative water level at each point. <P>SOLUTION: This system is equipped with a photographing means 1 such as a CCTV camera, a static image acquisition means 2 for acquiring image data from the photographing means 1 as a static image, a water level measuring means 3 for measuring the water level by image processing, a water level display means 4 for displaying the water level longitudinally, and an alarm/delivery means 5 for issuing an alarm to a river manager. The photographing means 1 mainly utilizes an existing CCTV camera installed for observing a prescribed water area of a river or the like, and the static image acquisition means 2 acquires as the static image, the image data from the photographing means 1 acquired by photographing the water surface and a measuring object such as a structure existing in the river. The water level measuring means 3 is equipped with a water level measuring device 6 for detecting the water surface by image processing of CCTV or the like, and measuring the water level by using the position of the water surface and three-dimensional space information measured beforehand or a reference image. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

The present invention calculates the cross-sectional water volume of the river and, in addition, grasps the river flow rate and flow velocity longitudinally along the basin based on the absolute water level information based on the altitude, thereby achieving higher accuracy. The present invention relates to a useful water level observation system that can provide disaster information.

Conventionally, in rivers, seas, lakes, etc., the water level is constantly measured and the measurement results are stored over a long period of time to prevent disasters caused by rising water levels and to control the flow of dams. . In addition, the current state of water level observation in rivers is that mechanical observations are performed regularly at fixed water level observation stations.

For example, such mechanical water level observation includes an automatic water level recorder that floats on the surface of the water and measures the movement of the rope that suspends the position of the float with the amount of rotation of the pulley. The float of this water level automatic recorder is to dig a deep well on land close to the river, and connect the well and the river with underground water so that the water level of the river and the water level of the well match. In some cases, a strong iron pipe is set up in water and the float is moved up and down in the iron pipe (for example, Patent Document 1).
reference). In addition, a method of measuring the water level using ultrasonic waves (ultrasonic method) and a method of measuring water pressure via a precise instrument (water pressure method) have already been devised.
JP-A-7-114416

However, in the conventional mechanical water level observation described above, the number of observation points is insufficient and the relative water level is measured at each point. There is a problem that the water level cannot be predicted, and as a result, the flow rate and flow velocity of the river cannot be observed longitudinally along the basin.

In addition, in the conventional mechanical water level observation described above, there is a problem in the reliability of the measurement water level, such as malfunctions such as floats often occurring due to water stains, etc. In addition, there is a problem that the construction cost of the measurement system becomes high because it is arranged at each measurement site.

Moreover, in any of the above-described conventional methods (mechanical method, ultrasonic method, water pressure method),
In order to install a precise instrument in an environment that is subject to the effects of wind and water damage, maintenance and management due to the occurrence of a failure requires a great deal of effort, and during a failure, no data at the time of a disaster can be obtained. There is a problem. Moreover, correction (calibration) of the measured value with a quantity water plate is essential, and if the correction deviation occurs, there is a problem that the reliability of the measured water level is lowered.

The present invention has been made in view of such conventional problems, and without using any mechanical water level observation, it calculates the amount of water across the river, and in addition it is absolute based on elevation. The purpose is to provide a water level observation system with useful image processing that can predict more accurate disaster information by longitudinally grasping the river flow rate and flow velocity along the basin based on the water level information. is there.

In order to solve the conventional problems as described above and achieve the intended purpose, the configuration as the gist of the present invention is a CCTV camera or digital camera installed for monitoring a predetermined water area such as a river, a dam or a lake. In a water level observation system that observes the water level using a photographing means such as a camera,
A still image acquisition means for acquiring, as a still image, image data from the imaging means, which is obtained by copying a water surface transmitted via a network and a measurement object such as a structure existing in a river;
Water level measurement means that detects the water level using image processing such as TV and measures the water level using the 3D spatial information or reference image measured in advance and the water level, and simultaneously measures the water level information at multiple points. Summarize and select all or any of the water level display means for displaying the level and / or vertical water level and the warning / distribution means for warning the river manager, disaster prevention personnel, etc. It exists in the water level observation system by the combined image processing.

Further, the water level measurement means includes an image information composition unit that synthesizes the acquired image data and three-dimensional spatial information, and each of the images displaying at least one attribute information including a group of latitude, longitude, and altitude of the subject. An attribute information adding unit to be applied to the pixel, a luminance information acquiring unit for acquiring luminance information of the subject surface from the image data, a physical property determining unit for determining the type and state of the object based on the luminance information, the latitude of the subject surface, A three-dimensional surface information extraction unit that extracts three-dimensional surface information composed of a group of longitude and elevation, an attribute information drawing unit that draws attribute information of the subject surface obtained by the three-dimensional surface information extraction unit on a two-dimensional screen, A semantic information creation unit that creates spatial data given meaning to the three-dimensional spatial information of the subject, and a pixel information classification unit that classifies each pixel of the two-dimensional image by the same spatial data or the same three-dimensional information ,etc A change curve creation unit that creates a luminance information change curve based on the luminance information acquired for each line, and a liquid level boundary information extraction unit that extracts elevation information of the liquid level boundary line of the subject based on the luminance information change curve It is good to provide the water level measuring device which selects or combines all or any.

The present invention is configured as described above, and since river information can be grasped longitudinally along the basin, the flow rate / velocity can be grasped longitudinally, and thus there is an effect that real-time disaster prediction information can be provided to the disaster prevention system. is there.

In other words, by displaying time-series fluctuations in the water level, it is possible to infer future changes in the water level, so river managers and disaster prevention personnel, etc. will be accurate from river information such as rainfall and water level that change from moment to moment. Since the system information necessary for flood control activities can be distributed in real time, it is extremely efficient and effective for disaster countermeasures.

In particular, CCTV is constantly transmitted on LANs constructed by national and local governments.
Camera still images used for water level observation are extracted from the camera video data at regular time intervals and automatically loaded into the system. The water level can be measured in real time during normal times and during floods, and displayed on a personal computer. It is possible to select an arbitrary CCTV camera from the distribution map of the CCTV cameras being used, measure the water level in real time at an arbitrary time, and display the result.

  Moreover, by acquiring a plurality of CCTV camera still images in a short time and averaging the measured water level, it is possible to remove the frequency due to waves and the like and to improve the water level measurement accuracy.

In addition, the water level observation system of the present invention can be set to continuous automatic observation mode or any CCTV.
With simple operations just by selecting a camera, observation results such as the water level can be obtained immediately, making it easy to use maps and lists on the PC screen, and the water level for which image processing was requested for each water level observation target point. Not only can it be stored and utilized as a database, but can be stored and used as a database for each river, and by adding the water level data of a conventional water level observatory, longitudinal water level fluctuations from the upstream of the river to the estuary are displayed on a map, time series graph, and list. It also has the effect that it can be displayed easily in a table.

Furthermore, the measured water level is compared with the dangerous water level, warning water level, etc., and if it exceeds these levels, a warning mark indicating that the water level has reached a certain level or higher is displayed on the plan view of the river basin. A warning sound can be used to clearly indicate a point at which the river manager is at risk of flooding, and by displaying time-series fluctuations in the water level, it can also be provided as auxiliary data for estimating future water level changes.

A still image acquisition means for acquiring, as a still image, image data from the photographing means, which is obtained by copying a water surface transmitted through a network and a measurement object such as a structure existing in a river;
Water level measurement means for detecting the water level by image processing such as CTV and measuring the water level using the three-dimensional spatial information or reference image measured in advance and the water level, and simultaneously measuring the water level information at a plurality of points. In summary, there are provided a water level display means for performing a planar and / or longitudinal water level display, and a warning / distribution means for giving a warning to a related person such as a river manager or a disaster prevention person.

Next, an embodiment of the present invention will be described with reference to the drawings. In the figure, A is a water level observation system by image processing according to the present invention (hereinafter, simply referred to as “main water level observation system”). This main water level observation system A is photographed by a CCTV camera or the like as shown in FIG. Means 1, still image acquisition means 2 for acquiring image data from the photographing means 1 as a still image, water level measurement means 3 for measuring the water level by image processing, and water level display means 4 for longitudinally displaying the water level And a warning / distribution means 5 for issuing a warning to a river manager or the like.

The photographing means 1 mainly uses existing CCTV cameras installed for monitoring predetermined water areas such as rivers, dams, and lakes, and transmits video data using a digital camera or the like as necessary. Needless to say, it can be done.

Further, the still image acquisition unit 2 captures, for example, a subject transmitted through an optical fiber network, that is, a measurement object (including a reflection target) such as a water surface and a structure existing in a river. The image data from the means 1 is acquired as a still image.

Further, the water level measurement means 3 detects the water level by image processing such as CCTV, and measures the water level using the later-described three-dimensional spatial information or reference image measured in advance and the position of the water surface. A device 6 is provided.

Moreover, the water level display means 4 measures and summarizes water level information at a plurality of points at the same time, and displays the water level in a longitudinal direction. A CCTV camera is provided on the longitudinal section of the target river (main river and tributaries separately). In addition to displaying the name, time and water level, main structures such as bridges, distance markers, levee heights, warning water levels, etc. are also displayed, so which points in the whole upstream and downstream are at risk of overflow, etc. Can be overlooked.

Furthermore, the warning / distribution means 5 is for warning the river manager, disaster prevention personnel, and other related parties, but the river manager, the disaster prevention personnel, etc. change from moment to moment. It may be programmed so that it can be distributed in real time as information that was necessary for carrying out appropriate flood control activities from river information such as rainfall and water level.

In particular, when the warning water level is reached or expected to be reached, an alarm signal is automatically notified to the local platform including the relevant river water area as disaster prediction information to inform the danger of flooding and evacuation. It may be constructed so that recommendations and evacuation guidance can be encouraged.

For example, the distribution means is to automatically send out river information related to rain and related information on flood hazard maps from the distribution server (data distribution), and is an existing platform that can use the Internet, personal computer communication, and network (PCs, digital information home appliances, mobile phones, PHS, car navigation, etc.) are used.

On the other hand, the water level measurement device 6 is displayed with an image information composition unit 6a that synthesizes the acquired digital image and the three-dimensional spatial information of the subject, and at least one attribute information consisting of the latitude, longitude, and altitude of the subject. An attribute information adding unit 6b to be applied to each pixel of the digital image, a luminance information acquiring unit 6c for acquiring luminance information on the subject surface from the digital image, and a physical property determining unit 6d for determining the type and state of the object based on the luminance information. A three-dimensional surface information extraction unit 6e composed of a group of latitude, longitude, and altitude of the subject surface, and an attribute information drawing unit 6f that draws the attribute information of the subject surface obtained by the three-dimensional surface information extraction unit 6e on a two-dimensional image And a semantic information creating unit 6g that creates spatial data given to the 3D spatial information of the subject, and separates each pixel of the 2D image into the same spatial data or the same 3D spatial information. And a pixel information classifying unit 6h to.

Furthermore, the attribute information drawing unit 6f draws contour lines that connect the same elevation attributes on the subject, and the luminance information acquisition unit 6c acquires luminance information for each contour line added to the digital image. A change curve creation unit 6i that creates a luminance information change curve based on the information and a liquid level boundary information extraction unit 6j that extracts elevation information of the liquid level boundary line of the subject based on the luminance information change curve are provided.

In addition, the liquid surface boundary information extraction unit 6j scans the average degree change curve by designating a certain section, and detects the section most correlated with the above-defined model curve shape from the parameters of the curve in the scanned section. To do. In addition, with reference to the predefined curve features,
The most correlated elevation position is detected from the detected section.

Here, a method for acquiring luminance information from a digital image of a subject and specifying the water level of a river will be briefly described. Contour lines are drawn on the digital image by the attribute information drawing unit 6f, and the luminance information acquisition unit The luminance information of the subject surface is acquired from the digital image by 6c. For such luminance information, an average luminance value is obtained for each contour line, and a graph (not shown) of an average luminance change curve (continuous data curve of luminance information) along the elevation direction is created.

Then, the position of the contour line corresponding to the river water level is detected from such an average brightness change curve. As shown in FIG. 13, when the water surface boundary is detected in this average brightness change curve, it is abrupt. By detecting the position of the local pattern corresponding to the river water level from the luminance feature point causing the luminance attenuation, the water level can be quickly detected from the digital image.

Thus, in the water level observation system of the present invention, the river water level can be obtained based on the luminance information obtained from the digital image, so the difference in weather conditions (especially the decrease in illuminance at night)
Therefore, even when it is difficult to read an image, a more accurate image analysis can be performed.

Moreover, since the water level observation system of the present invention obtains the average brightness for each digital image of the river, the longer the observation width in the horizontal direction, the more the brightness information in the horizontal direction on the surface of the subject represented by the contour line. It can enhance and effectively remove noise on the digital image.

The water level observation system of the present invention configured as described above includes, for example, a digital image of a subject acquired by a surveillance camera (CCTV camera) fixed on a riverbed (not shown).
The acquired image signal of the subject is stored in a data storage medium such as a hard disk from a receiving device such as a personal computer and displayed on a monitor screen.
Furthermore, by designating a predetermined range on the digital screen, contour lines at regular intervals are drawn in such a range, and the water level is observed by the above-described image processing.

Specifically, an average luminance change curve along the altitude on the subject surface is separately calculated from digital images at time (T ₁ ) and (T ₁ + Δt). Next, the local curve pattern corresponding to the water level at time (T ₁ ) is referred to based on the luminance information at time (T ₁ ) accumulated in the luminance information acquisition unit 6c that accumulates data obtained by measuring luminance information over time. Thus, the curve section whose shape is most correlated is extracted from the curve at time (T ₁ + Δt). The altitude position corresponding to the minimum value of the average luminance in the extracted section is the water level at that time (T ₁ + Δt).

Thus, since the river water level can be measured based on the past luminance information accumulated in the luminance information acquisition unit 6c in the physical property determination unit 6d, compared with the case where the river water level is measured based on the pre-input data. Highly accurate measurement can be performed.

Furthermore, when processing a time-series digital image acquired from the fixed photographing means 1,
Instead of repeatedly calculating each fixed parameter every time a new image is read, high-speed processing of a digital image can be realized by substituting only the luminance information parameter into the equation and calculating the image analysis result.

In addition, when the water level of a river or the like being measured exceeds a preset height, the river water level is programmed to increase the frequency of measurement, so real-time water level information can be obtained, for example, evacuation guidance Evacuation advisories or warning warnings, adjustment of river flow by dam discharge, etc.
Appropriate and quick response is possible.

Next, the operation procedure of this water level observation system will be described in more detail. First, before introducing this water level observation system to the river office B, we will collect information on CCTV cameras that can be used as water level observation targets, survey information such as 3D spatial information, create models of target structures, and set various parameters. Must be done (system introduction).

Next, the water level observation target and CCTV camera are selected. The CCTV camera checks the set position and determines the measurement target structure while viewing each image. The location of CCTV cameras can be used as an object for observing the water level of bridges, etc. around the location from the office jurisdiction map or topographical map, for example, by receiving information from equipment management departments such as river offices that manage rivers Check if there is a structure. In addition, it is necessary to check the entire river to see if there is any bias.

In addition, the CCTV camera control device and display device installed in the disaster prevention room of the river office, etc., the target structure is in the proper direction while rotating and zooming the camera.
Check if you can shoot in the range.

Incidentally, as for the shooting direction of the camera, when the backlight or illumination light directly enters the image, the brightness of the object is lowered, which causes deterioration of the water level measurement degree. The range does not have to be enlarged, and whether or not the water surface can be photographed including matching points for image matching to be described later is examined. For example, when a bridge pier is targeted, it is desirable to have a size that allows a single image to be taken from the bottom of the pier where the water surface is level during flat water to the bridge girder used as a matching point.

Next, three-dimensional spatial information data and a reference image are acquired. Specifically, in the vicinity of the target structure and in the river office B, as shown in FIGS. 2 to 5, the measurement image 7 which is the three-dimensional data of the structure and the reference image 8 used for the image matching 13 are used. To get.

As the three-dimensional data of the target structure, either the ground laser scanner 9 or the total station 10 is selected according to the shape of the target structure. Incidentally, it is preferable to use the ground laser scanner 9 when the shape of the target structure is cylindrical or complicated, and the surveying method using the total station 10 when the target structure is a substantially flat plane.

  Next, the ground point laser scanner 9 or the total station 10 is set with the reference point arbitrarily set near the target structure, and the positions of the target structure and the reflection target installed in the vicinity are measured. In addition, the positional relationship between a river distance marker and a reference point is measured. From these survey results, the three-dimensional absolute coordinates 11 of the object and the reflection target are obtained.

The reference image 8 fixes a CCTV camera in a state where the target structure and the reflection target are reflected, and shoots several tens of minutes of data. When the reflected light or shadow changes due to the movement of the sun or surrounding illumination light, data acquisition is performed several times a day according to the change. Note that the measurement image (three-dimensional data) 7 and the reference image 8 are preferably acquired at the time of a normal level when the water level is low.

Next, a terrain perspective model is created. Specifically, the measurement screen 7 that is three-dimensional data and the reference image 8 are coupled via a reflection target to create a topographical perspective model. An irregular triangle network with three measurement points as one unit (excluding unnecessary parts from the acquired 3D data)
The three-dimensional surface data 12 of the target structure based on the (TIN) model is created.

It is desirable that the three-dimensional surface data 12 be an image viewed from the CCTV camera as much as possible. Reflection target and 3D surface data 1 in reference image 8
The reference image 8 is geometrically corrected so that the two reflection targets overlap, and the pixels of the reference image 8 and the absolute coordinates of the three-dimensional surface data 12 are linked.

Next, create a basic map of the entire water system to be displayed on the system, the installation position of the CCTV camera, and a longitudinal section of the main river and tributaries, and input them to the system together with the reference image 8 and the topographical perspective model (introduction to the office) . For example, feature points (such as corners of structures and rivets) on the reference image are set as matching points with the measurement image 7. In addition, the range for obtaining the image luminance value is set.

In the present water level observation system, “designated water level”, “warning water level”, “dangerous water level”, etc. can be set as warning conditions. However, since there are many cases where these are not set at the CCTV camera installation position, the water level as a warning condition is determined in consultation with the river office. After various settings, the system is connected to the LAN, and it is tested whether processing such as CCTV camera image capture can be performed.

These operations are performed only once by the trustee when the system is introduced, with the cooperation of the river office. In the event of river channel repairs or structural changes, the trustee will make corrections as necessary.

Next, when performing water level observation constantly, start up the software and perform CCT for water level observation.
By determining the V camera, confirming the video, and capturing a still image, setting the automatic observation mode makes it possible to easily perform continuous water level observation.

Specifically, when the software is started, as shown in FIG.
The installation position 15 and name 16 of the CCTV camera are displayed. When water level observation is possible, the installation position is represented by a black circle mark, and at the same time, the name is displayed in the list of the observation site 17 on the left.

Next, select the CCTV camera you want to see from the basin map 14 or the observation site list 17,
By pressing the “video playback” button 18, as shown in FIG.
The video 20 can be confirmed.

To set the automatic water level observation mode, capture still images from the video being played,
The still image capture button 19 is pressed, and an image analysis window 21 is displayed as shown in FIG. Next, the automatic observation mode setting button 22 is pressed or the ON check box of the automatic observation mode 23 on the automatic observation setting tab is checked, and the analysis state update button 24 of the image window is pressed. Needless to say, when automatic observation is performed for a plurality of CCTV cameras, it is necessary to repeat the above operation for each camera.

As a method for confirming the observation data of such water level, A. Basin plan, B. Water level fluctuation map (hydrograph) at a single observation point, C.I. Water level longitudinal section of each river, D. There are four of the latest water level information at a single observation point. Each feature is described below, and any method is programmed to display a warning when the water level exceeds a certain level.

A. As shown in Fig. 9, the basin plan 25 shows the CCT on the topographic map including the entire basin of the target water system.
The V camera installation position is marked, and the entire basin can be confirmed from a bird's-eye view. The mark is colored in blue (during water level observation, below warning water level), black (during water level observation preparation), red (during water level observation, above warning water level), and gray (water level observation not possible). In addition, CCTV
The camera name, water level, time, etc. can also be displayed.

B. Hydrograph (Water level fluctuation map at a single observation point)
As shown in FIG. 10, the hydrograph 26 can confirm the change of the water level in time series. Depending on the setting, it is possible to select and view three types of graphs: an average value every minute, an average value every 10 minutes, and an average value every hour. In addition to the dike height and riverbed height, the water level necessary for river managers to provide information to local governments, such as the planned high water level, dangerous water level, special warning water level, warning water level, and designated water level, is displayed. River managers can cope with future fluctuations based on time series changes in water levels.

C. Water level vertical cross-sectional view of each river As shown in FIG. 11, the water level vertical cross-sectional view 27 displays the CCTV camera name 16, time 28, and water level 29 on the vertical cross-sectional view of the target river (main river and tributaries separately). Is. In addition, the longitudinal view shows the main structures such as bridges, distance markers, levee heights, warning water levels, etc., so it is possible to see which points are in danger of overflow, etc. can do.

D. Latest Water Level Information at a Single Observation Point The latest water level information 30 displays the difference between the water level and the warning water level as a numerical value, as shown in FIG. Compared to Figs. 8-11, the entire basin cannot be seen.
. Since it can be confirmed as a specific number of 87 meters, it is effective basic data for issuing flood forecast warnings.

Here, the automatic water level observation mode of the water level observation system will be briefly described. First, as an image approach, all images (with CCTV cameras) that are observing the water level every few seconds to several tens of seconds are taken into the system.

In the image matching 13, the captured image (measurement image 7) and the reference image 8 are automatically matched. Feature points on the reference image 8 are set in advance as matching points, and three or more identical points are automatically extracted from the measurement image 7 to obtain an affine transformation equation 13a (see FIG. 2).

Next, the measurement image 7 and the terrain perspective model are linked 13 using the affine transformation formula 13a.
b, create a water level extraction model 13c for image analysis (create a water level extraction model),
An average image degree value is calculated for each image of the same elevation from a preset image luminance extraction range to obtain a luminance change curve (extraction of image luminance value).

Next, the water level is measured. Specifically, a feature point such as peak brightness is obtained from a brightness change curve and the water level is automatically observed. As shown in FIG. 13, the brightness feature point = water surface boundary is shown. It is devised so that the recognition of the shape of the feature point can be automatically changed depending on the direction of the object, the color of the object, and the attached matter.

Next, if you want to stop the automatic water level observation mode, select the CCTV camera you want to stop the automatic observation, uncheck the check box for the automatic extraction mode 23 from the automatic observation setting tab 31 at the bottom right of the screen, and analyze the image window The user can simply stop by clicking the update button 24 (see FIG. 11). If you want to exit this water level observation system (software), simply select “File”-“Exit” from the menu and click the “Exit” button.

It should be noted that the water level observation system of the present invention is not limited to this embodiment, and can be freely modified within the scope of the object of the present invention, and the present invention encompasses all of them. .

For example, a river information database is a GIS (Geographic Information System: Geographi).
c) Supports the preparation of map data by Information System)
It has river channel data acquisition function and hydrological data acquisition function to acquire river channel data on the network, and prepares river specification data necessary for calculation of river channel water level and cross-sectional view display.
A floodplain database, runoff analysis function, river channel water level prediction function, breach point inflow calculation function, flood analysis function, feedback correction function, etc. may be added as necessary.

In addition, radar rainfall information, telemeter rainfall, water level, water quality, dam quantities, etc., dam body deformation, levee body level, river water level, inland water level by optical fiber sensor, map information by GIS, GPS (
Location information and CCTV by Global Positioning System)
Digital archive data that collects data such as video and still image information from digital cameras and moving objects, and predicts floods, floods, sediment disasters, etc., and the results and past data are stored , As appropriate, including alarm information and evacuation guidance support information,
You may construct | assemble so that it can deliver to platforms, such as a personal computer, digital information household appliances, a mobile telephone, PHS, and a car navigation, via networks (including the internet), such as a satellite (CS) system and a ground system.

Incidentally, the flood plain database has a flood plain data acquisition function for acquiring data on flood plains on the network in response to the development of the national spatial data infrastructure. This flood plain data acquisition function calculates the inundation depth from the obtained land information and calculates the evacuation route. Specifically, the flood plain elevation and land use data required for the flood analysis, The data on the location of the evacuation site and the attribute data necessary for studying the evacuation plan will be prepared.

In addition, the runoff analysis function may include a river inflow calculation means for calculating a flow rate flowing into the river using an estimated value or actual measured value of the upstream average rainfall or an arbitrarily set rainfall value as an index.

Furthermore, the river water level prediction function consists of two models: runoff analysis and river unsteady flow. It can be determined by rainfall that has been set arbitrarily, such as rainfall of major floods in the past and planned rainfall used to calculate basic high water flow. Perform runoff analysis calculations. The river level in the river channel is calculated by indeterminate flow calculation using the calculated runoff.

The function of calculating the amount of inflow at the breach point is to calculate the flow rate flowing into the floodplain from the breach point, using the crossover formula from the river level at the breach point, the water level at the floodplain, and the breach width. calculate. still,
The bank breakage width and how it spreads are based on, for example, the civil engineering research manual.

Furthermore, the inundation analysis function is to perform inundation analysis using the flow rate flowing into the floodplain from any bank breakage point. As an inundation analysis method, for example, two-dimensional that can accurately reproduce the movement of the inundation flow An indefinite flow model is mentioned.

The feedback correction function includes, for example, a local information feedback function that is calculated based on information from the local area, a water depth feedback function that is calculated from the inundation depth by the inundation sensor, and an inundation area that is calculated from the inundation area based on observation data such as satellites. Feedback function,
The internal water discharge amount feedback function that is calculated from the discharge amount of the internal water pump, etc. is mentioned, and various observation information that changes every moment at the time of the levee is fed back as the initial value of the flood prediction calculation, and the flood analysis is performed with high accuracy It is.

It is a whole explanatory view showing the outline of the water level observation system by the image processing concerning the present invention. It is a flowchart which shows preparation of the water level extraction model in the same water level observation system. It is explanatory drawing which shows acquisition of the image for image matching reference in the water level observation system. It is explanatory drawing which shows preparation of the topographical perspective model in the same water level observation system. It is explanatory drawing which shows the matching of the image for a reference, and the image for a measurement in the water level observation system. It is explanatory drawing which shows the watershed map screen in the same water level observation system. It is explanatory drawing which shows the CCTV video screen in the water level observation system. It is explanatory drawing which shows the image analysis window screen in the same water level observation system. It is explanatory drawing which shows the watershed top view screen in the water level observation system. It is explanatory drawing which shows the hydrograph display screen in the same water level observation system. It is explanatory drawing which shows the display of the river longitudinal cross-sectional view screen in the same water level observation system. It is explanatory drawing which shows the water level latest information screen in the water level observation system. It is explanatory drawing which shows the water level extraction table | surface by the image processing in the water level observation system.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Image | photographing means 2 Still image acquisition means 3 Water level measurement means 4 Water level display means 5 Warning / distribution means 6 Water level measurement apparatus 6a Image information synthesis | combination part 6b Attribute information provision part 6c Luminance information acquisition part 6d Physical property discrimination | determination part 6e Three-dimensional surface information extraction 6f Attribute information drawing unit 6g Semantic information creation unit 6h Pixel information classification unit 6i Change curve creation unit 6j Liquid boundary information extraction unit 7 Measurement image 8 Reference image 9 Ground laser scanner 10 Total station 11 Three-dimensional absolute coordinate 12 Three-dimensional Surface data 13 Image matching
13a Affine transformation formula
13b link
13c Water level extraction model 14 Basin map of the entire water system 15 Installation position of CCTV camera 16 Name 17 Observation site 18 Video playback button 19 Still image capture button 20 CCTV video 21 Image analysis window
22 Automatic Observation Mode Setting Button 23 Automatic Observation Mode 24 Update Button 25 Basin Plan 26 Hydrograph 27 Water Level Longitudinal View 28 Time 29 Water Level 30 Latest Water Level Information 31 Automatic Observation Setting Tab

Claims (2)

In a water level observation system for observing the water level using a photographing means such as a CCTV camera or a digital camera installed to monitor a predetermined water area such as a river, a dam or a lake,
A still image acquisition means for acquiring, as a still image, image data from the photographing means, which is a photograph of a water surface transmitted via a network and a measurement object such as a structure existing in a river (hereinafter simply referred to as a subject); Detect water level by image processing such as CCTV, water level measurement means to measure water level using water surface position and pre-measured 3D spatial information or reference image, and water level information at multiple points simultaneously Select all or one of the water level display means to display the level and / or vertical water level, and the warning / distribution means to warn concerned parties such as river managers and disaster prevention personnel. Or a water level observation system by image processing, characterized by being combined.   The water level measuring means is an image information combining section that combines the acquired image data from the photographing means and three-dimensional spatial information, and an image displaying at least one attribute information consisting of a group of latitude, longitude, and altitude of the subject. An attribute information adding unit to be applied to each pixel, a luminance information acquiring unit for acquiring luminance information on the surface of the subject from the image data, a physical property determining unit for determining the type and state of the object based on the luminance information, A three-dimensional surface information extraction unit that extracts three-dimensional surface information composed of a group of latitude, longitude, and altitude, and an attribute information drawing unit that draws attribute information of the subject surface obtained by the three-dimensional surface information extraction unit on a two-dimensional screen And a semantic information creation unit that creates spatial data given meaning to the three-dimensional spatial information of the subject, and pixel information classification that classifies each pixel of the two-dimensional image according to the same spatial data or the same three-dimensional information A change curve creation unit that creates a luminance information change curve based on the luminance information acquired for each contour line, and a liquid level boundary information extraction unit that extracts elevation information of the liquid level boundary line of the subject based on the luminance information change curve The water level observation system by image processing according to claim 1, comprising a water level measuring device that selects or combines all or any of the above.