Water level estimation system and water level estimation method

Abstract

To provide a water level estimation system and a water level estimation method which accurately estimate a water level of a water storage part with a simple procedure.SOLUTION: A water level estimation system 30 stores a first image obtained by imaging a water storage part from a prescribed direction, sets a boundary between the water storage part and water showing a water level of the water storage part on the first image for the plurality of water levels, specifies a variation direction of the water level on the first image in the water storage part on the first image by setting a reference point on each set boundary, acquires a second image obtained by imaging the water storage part from the prescribed direction at set timing, specifies a portion of the water level on the acquired second image, estimates the water level in the second image on the basis of the boundary at the plurality of water levels on the first image, the variation direction of the water level on the first image and the specified portion of the water level on the second image, and outputs information indicating the estimated water level.SELECTED DRAWING: Figure 3

Description

[Technical Field]

[0001] The present invention relates to a water level estimation system and a water level estimation method. [Background technology]

[0002] The water level in reservoirs such as rivers is often determined by visual inspection using water level gauges and surveillance cameras. However, in recent years, there has been a growing need for an accurate and inexpensive method of determining water levels due to the frequent occurrence of severe wind and water disasters.

[0003] Systems for determining water levels have been developed in the past. For example, Patent Document 1 describes a river management device that includes a processing unit that generates a reference water level composite image by superimposing a reference water level line, which indicates a reference water level that serves as a reference for determining flood prevention activities against rising water levels and flooding of rivers and for evacuation actions of residents, on a river image acquired by an image acquisition device, and an output unit that outputs the reference water level composite image generated by the processing unit. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Patent Publication No. 2021-124370 Summary of the Invention [Problem to be solved by the invention]

[0005] However, it is not easy to build a system that can be provided with practical accuracy and at a reasonable price. For example, Patent Document 1 describes extracting and using a portion of a captured video as an image. However, when using video to determine water levels, the amount of communication data increases, which can cause problems in terms of performance (such as rapid response to emergency disasters) and management costs.

[0006] Furthermore, the placement of surveillance cameras capturing images of rivers varies depending on the river and the area being photographed, and the direction of river flow in each image often differs, which can also reduce the accuracy of water level determination.

[0007] The present invention has been made in consideration of the above circumstances, and its purpose is to provide a water level estimation system and a water level estimation method that can accurately estimate the water level in a water storage section using simple procedures. [Means for solving the problem]

[0008] One aspect of the present invention to solve the above problem is to provide a method for capturing an image of a water storage area from a predetermined direction. each a storage device for storing a first image; each A boundary between the water reservoir and the water on the first image, which indicates the water level of the water reservoir. and the straight line segments or curved portions are approximately parallel to each other. Action to be set of , The process is performed for each of a plurality of water levels including at least a first water level and a second water level higher than the first water level. Processing and the set , at least for each of the first and second water levels On each boundary As a predetermined reference point in the boundary, the position of the center of gravity of a rectangle whose diagonal is the line segment of the straight line or a rectangle whose diagonal is a line segment connecting both ends of the curved line is set as the reference point of the boundary. By doing so, on the first image in the water storage section , at least from the first water level to the second water level Water level Rising waters a process of acquiring a second image of the water storage unit taken from the predetermined direction at a set timing, and identifying a portion of the water surface on the acquired second image; and a process of identifying the boundary portion at the plurality of water levels on the first image. The area of ​​the boundary in the coordinate system on the first image based on the reference point , the water level on the first image Rising waters and estimating the water level in the second image based on the direction and the identified portion of the water surface on the second image. In this case, it is determined whether the boundary portion for the first water level coincides with any part of the water surface of the second image, and if the boundary portion is even partially on the water surface, a series of processes for making the determination for the boundary portion for the second water level is repeatedly performed for the boundary portion for each water level in an order according to the rising water direction until the boundary portion no longer coincides with the water surface at all, and when the boundary portion for a certain water level no longer coincides with the water surface at all, the water level is estimated to be the water level of the second image. and a processing device that executes a process of outputting information indicating the estimated water level. [Effects of the Invention]

[0009] According to the present invention, the water level of a water reservoir can be estimated with high accuracy using a simple procedure. Configurations and effects other than those described above will become apparent from the following description of the embodiments. [Brief explanation of the drawings]

[0010] [Figure 1] 1 is a diagram illustrating an example of the configuration of a water level monitoring system according to an embodiment of the present invention. [Figure 2] FIG. 2 is a diagram illustrating an example of hardware included in the water level estimation system and functions of the water level estimation system. [Figure 3] FIG. 10 is a flow diagram illustrating an outline of a river water level monitoring process. [Figure 4] FIG. 10 is a flowchart illustrating details of the water level boundary line setting process. [Figure 5] 10A and 10B are diagrams illustrating an example of a normal image and a water level boundary line set on the normal image. [Figure 6] FIG. 10 is a diagram showing an example of a normal image in which a plurality of water level boundary lines are set. [Figure 7] A flow chart explaining the details of the water level rise direction determination process. [Figure 8] 10 is a diagram showing an example in which a circumscribing rectangle and a water level reference point are set for a water level boundary line of a straight line segment. FIG. [Figure 9] 10 is an example of a threshold line parameter DB. [Figure 10] FIG. 10 is a flowchart illustrating details of an initial setting information setting process. [Figure 11] FIG. 10 is a diagram illustrating an example of initial setting information. [Figure 12] FIG. 10 is a flowchart illustrating details of an estimated start time setting process. [Figure 13] FIG. 10 is a flowchart illustrating details of a water level determination process. [Figure 14] FIG. 10 is a flowchart illustrating the details of the water level estimation process. [Figure 15] FIG. 10 is a diagram showing an example of a transition estimation result screen. DETAILED DESCRIPTION OF THE INVENTION

[0011] An embodiment of the present invention will be described with reference to the drawings.

[0012] 1 is a diagram showing an example of the configuration of a water level monitoring system 1 according to this embodiment. The water level monitoring system 1 includes one or more imaging devices 10, a water level management system 20, and a water level estimation system 30.

[0013] The photographing device 10 is installed at a predetermined location along the river 40 with a predetermined photographing direction set. The photographing device 10 photographs images of the river 40 at predetermined timings (for example, at predetermined times or predetermined time intervals). One or more photographing devices 10 are installed in one or more predetermined photographing areas.

[0014] The water level management system 20 is an information processing system consisting of one or more information processing devices used by a manager or the like who manages the river 40. The water level management system 20 acquires and stores images captured by the photographing device 10. The water level management system 20 also displays information (described below) related to the water level of the river 40 estimated by the water level estimation system 30.

[0015] The water level estimation system 30 is an information processing system consisting of one or more information processing devices. When the water level estimation system 30 receives estimation request information requesting a water level estimation from the water level management system 20, it estimates the water level of the river 40 captured in an image stored in the water level management system 20. Then, the water level estimation system 30 transmits information on the estimated water level to the water level estimation system 30.

[0016] The photographing device 10, the water level control system 20, and the water level estimation system 30 are connected to each other via, for example, the Internet, a LAN (Local Area Network), a WAN (Wide Area Network), or a dedicated Communication is possible via a wired or wireless communication network such as a utility line.

[0017] Next, FIG. 2 is a diagram showing an example of hardware included in the water level estimation system 30 and functions of the water level estimation system 30. As shown in FIG.

[0018] The water level estimation system 30 comprises the following functional units: a water level boundary line setting unit 101 , a water rise direction determination unit 102 , an initial setting information setting unit 103 , an estimation start time setting unit 104 , and a water level determination unit 105 .

[0019] The water level boundary line setting unit 101 sets the boundary line (hereinafter referred to as the water level boundary line) between the reservoir part (low water channel, levee, etc.) of the river 40 and the water on the image of the river 40 photographed by the photographing device 10 for a plurality of normal water levels of the river 40. The water level boundary line is, for example, the boundary line between the water surface of the river 40 and the slope of the levee on the river front side on the image.

[0020] The water level rise direction determination unit 102 identifies the direction of change in the water level of the river 40 (in this embodiment, the direction of rise) on the image of the river 40 by setting a predetermined reference point (hereinafter referred to as a water level reference point) on each water level boundary line set by the water level boundary line setting unit 101.

[0021] Information on the water level boundary line and water level reference point is recorded in the threshold line parameter DB 300 .

[0022] The initial setting information setting unit 103 sets the initial setting information 113, which is information that defines the conditions under which a notification regarding the water level of the river 40 is sent to the water level management system 20.

[0023] The estimation start time setting unit 104 sets information on the timing for starting monitoring of the water level of the river 40 (estimation start time).

[0024] When the estimation start time arrives, the water level determination unit 105 acquires a current image of the river 40 and identifies the water surface portion in the acquired image (hereinafter referred to as the current image) using a water surface determination model 111, which will be described later.The water level determination unit 105 then estimates the water level in the current image based on the water level boundary lines and rising direction, and the information on the water surface portion.The estimation result is recorded in the estimation result DB 200.

[0025] The water surface determination model 111 is a trained model that, when an image of the river 40 is input, identifies the water surface portion of the river 40 and other portions in the image, for example, by utilizing semantic segmentation. The water surface determination model 111 is constructed by machine learning the feature quantities of each image of rivers across the country using deep learning. The water surface determination model 111 is a neural network that has an input layer to which pixel information of the image (contents and coordinates of pixels) is input, one or more intermediate layers (hidden layers) that extract and output the feature quantities of the image from the pixel information, and an output layer that outputs the coordinates of the water surface part from the feature quantities of the image. For example, a CNN (Convolution Neural Network) is used as this neural network. Network), SVM (Support Vector Machine), Bayesian network, regression tree, etc. can be applied.

[0026] Next, the water level estimation system 30 includes a CPU (Central Processing Unit) as hardware. A processing device 31 (processor) such as a DSP (Digital Signal Processor), a GPU (Graphics Processing Unit), or an FPGA (Field-Programmable Gate Array), a main memory device 32 (memory) such as a ROM (Read Only Memory), a RAM (Random Access Memory), and an HDD (Hard Disk Drive). , an auxiliary storage device 33 such as an SSD (Solid State Drive), an NIC (Network Interface Card), a wireless communication module, a USB (Universal Serial Interface) module, or a serial A communication device 34 configured with a communication module or the like, an input device 35 configured with a mouse, keyboard, or the like, and a liquid crystal display or organic EL (Electro-Luminescence) display or the like. The water level control system 20 also includes similar hardware.

[0027] The functions of each functional unit of the water level estimation system 30 are controlled by a processing unit 31, a main memory 32, or an auxiliary memory 33. The water level estimation system 30 is realized by reading and executing each program stored in the storage device 33. Each program can also be recorded on a recording medium and distributed, for example. Note that all or part of each information processing device may be realized using virtual information processing resources provided using virtualization technology, process space separation technology, or the like, such as a virtual server provided by a cloud system. Furthermore, all or part of the functions provided by the water level estimation system 30 may be realized by a service provided by a cloud system via an API (Application Programming Interface), for example. Next, the processing performed by the water level monitoring system 1 will be described.

[0028] <River water level monitoring processing> FIG. 3 is a flow diagram outlining the process of estimating the current water level of a river (river water level monitoring process) performed by the water level monitoring system 1.

[0029] First, the photographing devices 10 photograph an image of the river 40 (s10). Specifically, each photographing device 10 photographs an image of the river 40 when the water level is within a normal range (hereinafter referred to as a normal image or a first image). Then, each photographing device 10 transmits the photographed normal image to the water level management system 20.

[0030] The water level estimation system 30 receives each normal image from the water level management system 20 and executes a water level boundary line setting process s20 to set a water level boundary line for each received normal image. The water level boundary line setting process s20 will be described in detail later.

[0031] Then, for each normal image, the water level estimation system 30 executes a water level rise direction determination process s30 for determining the water level rise direction of the river 40 on the normal image based on the water level boundary line set in the water level boundary line setting process s20. The water level rise direction determination process s30 will be described in detail later.

[0032] The water level estimation system 30 also executes an initial setting information setting process s40 for determining the initial setting information 113.

[0033] The water level estimation system 30 also executes an estimation start time setting process s50 for setting an estimation start time.

[0034] Thereafter, when the estimation start time set in the estimation start time setting process s50 arrives, the water level estimation system 30 executes the water level determination process s60 to estimate the water level of the river 40. The water level determination process s60 is repeatedly executed as needed. Each process will be described in detail below.

[0035] <Water level boundary setting process> FIG. 4 is a flow diagram illustrating the details of the water level boundary line setting process s20.

[0036] The water level boundary line setting unit 101 acquires a normal image (s21). For example, the water level boundary line setting unit 101 starts a predetermined image editing program, accepts a designation of a normal image from a user, reads the normal image from the water level management system 20, and displays it on the screen. Note that the following describes a case where the water level boundary line setting unit 101 processes one normal image captured by one particular imaging device 10.

[0037] The water level boundary line setting unit 101 sets a boundary line (water level boundary line) between the river 40 and the reservoir (levee, etc.) on the normal image acquired in s21, and records information on the coordinates of the set water level boundary line on the normal image in the threshold line parameter DB300 (s22).

[0038] 5 is a diagram showing an example of a normal image and a water level boundary line set on the normal image. As shown in the figure, this normal image 400 captures a water surface 401 of a river 40, a levee 402 which is the water reservoir of the river 40, and a structure 403 outside the levee 402. This normal image 400 is an image captured from a camera 10 (not shown) fixed at a height overlooking the river 40. The height of the water surface 401 of the river 40 shown in this normal image 400 reaches a water level line 404 of a certain water level on the slope of the levee 402 on the river-front side.

[0039] Here, for example, the image editing program receives a drawing of a line segment (which may be a curve or a straight line) along the water level line 404 from the user as a water level boundary line via the input device 35. This line segment does not need to be the entire water level line 404 of the normal image 400, but may be a part of it.

[0040] If the water level line 404 is a curve, the image editing program may accept a drawing of a straight line approximating the curve from the user as the water level boundary line. In this embodiment, as shown in Figure 4, a line segment approximating a part of the curved water level line 404 with a straight line is set as the water level boundary line 406.

[0041] Here, the water level boundary line may be automatically set by the water level estimation system 30 or may be suggested to the user. For example, the water level boundary line setting unit 101 may recognize the boundary between water and other parts (e.g., the water level line 404) based on the output value obtained by inputting the normal image into the water surface determination model 111, and may display the recognized boundary by superimposing it on the normal image 400, thereby suggesting the recognized boundary to the user as a candidate for the water level boundary line to be set, or may set the water level boundary line automatically.

[0042] Next, as shown in FIG. 3, the water level boundary setting unit 101 receives a drawing of a water level boundary line on the normal image when the water level rises above the water level of the water level boundary line set in s22, and records the coordinates of the water level boundary line (coordinates on the normal image) in the threshold line parameter DB 300 (s23).

[0043] For example, the water level boundary line setting unit 101 accepts input from the user to draw a new water level boundary line while continuing to display the normal image from s22. Note that the water level boundary line setting unit 101 may accept from the user a designation of another normal image of the river 40 that shows a water level higher than the water level shown in the normal image displayed in s22, and may accept input to draw a water level boundary line by reading that normal image from the water level management system 20 and displaying it.

[0044] Then, the water level boundary line setting unit 101 confirms whether or not to input a water level boundary line at a water level higher than the water level indicated by the water level boundary line drawn in s23 (s24). For example, the water level boundary line setting unit 101 receives a confirmation input from the user as to whether or not to input a water level boundary line at a higher water level.

[0045] If a water level boundary line at a higher water level is to be input (s24: NO), the water level boundary line setting unit 101 executes the process of s23, and if a water level boundary line at a higher water level is not to be input (s24: YES), the water level boundary line setting process s20 ends. The water level boundary line setting unit 101 may also be configured to automatically execute the process of s23 a predetermined number of times.

[0046] Here, Fig. 6 is a diagram showing an example of a normal image 400 in which a plurality of water level boundary lines are set. In this normal image 400, a water level boundary line 406 (first water level boundary line) for the lowest water level (first set) explained in Fig. 5, a water level boundary line 407 (second water level boundary line) for when the water level rises further, and a water level boundary line 408 (third water level boundary line) for when the water level rises further to the limit water level of the normal water level. The first water level boundary line, second water level boundary line, and third water level boundary line are set. The boundary lines are set parallel or approximately parallel to each other and have the same or approximately the same length. The extension directions of the first water level boundary line, the second water level boundary line, and the third water level boundary line are parallel or approximately parallel to the flow direction 409 of the river 40, and the arrangement directions thereof are perpendicular or approximately perpendicular to the flow direction 409 of the river 40.

[0047] In this way, it is preferable that each water level boundary line is drawn with the same length and aligned with each other, and is suggested to the user by the water level estimation system 30 or set automatically by the water level estimation system 30.

[0048] <Water level rise direction determination process> Next, FIG. 7 is a flow diagram illustrating the details of the water level increase direction determination process s30. The flood direction determining unit 102 reads the normal image (hereinafter referred to as the water level threshold line image) on which the water level boundary line has been set in the water level boundary line setting process s20 (s31).

[0049] The water level rise direction determination unit 102 sets the water level indicated by the water level boundary line drawn on the water level threshold line image (hereinafter referred to as water level level) for the water level threshold line image, and stores the set in the threshold line parameter DB 300 (s32). For example, the water level rise direction determination unit 102 associates and sets information such as "Level 1," "Level 2," "Level 3," etc. for each water level boundary line in the order set in the water level boundary line setting process s20 (i.e., in order of increasing water level).

[0050] Next, the flood direction determination unit 102 sets water level reference points, which are representative points or reference points for the water level of the river 40, on each water level boundary line of the water level threshold line image, and stores them in the threshold line parameter DB 300 (s33).

[0051] For example, the water level rise direction determination unit 102 calculates the coordinates of each vertex of a rectangle (e.g., a rectangle; hereinafter referred to as a circumscribed rectangle) whose diagonal is the straight line connecting both ends of the water level boundary line, calculates the center of gravity of the circumscribed rectangle based on the calculated coordinates of each vertex, and sets this as the water level reference point.Furthermore, for example, the water level rise direction determination unit 102 calculates the midpoint between the two coordinates of both ends of the water level boundary line on each water level threshold line image as the water level reference point. FIG. 8 shows an example in which a circumscribing rectangle 411 and a water level reference point 412 are set for the water level boundary line 406 of the straight line segment shown in FIG.

[0052] Here, the water level boundary line 406 does not necessarily have to be a straight line, and the center of gravity of the circumscribed rectangle can be calculated even if the water level boundary line 406 is a curved portion (a curve with both ends). Specifically, the water level rise direction determination unit 102 determines the circumscribed rectangle to be a rectangle whose diagonal is a straight line segment connecting both ends of the curved portion.

[0053] The method for setting the water level reference point described here is merely an example, and the method for determining the water level boundary line may be any method as long as it conforms to certain standards as described above.

[0054] Next, as shown in Figure 7, the flood direction determination unit 102 stores the water level reference points in ascending order of their water level (level 1, level 2, ...) and stores these as the rising direction (flood direction) of the water level of the river 40 (s34). This completes the flood direction determination process s30.

[0055] For example, the water rise direction determination unit 102 identifies as the water rise direction the average direction of the direction from the water level reference point of level 1 to the water level reference point of level 2 and the water rise direction from the water level reference point of level 2 to the water level reference point of level 3. Note that if the lengths of each water level boundary line are set to be the same and the positions of each water level boundary line are aligned in the water level boundary line setting process s20, the water rise directions between each level will all be the same, making the process simple.

[0056] (Threshold line parameter DB) 9 is an example of the threshold line parameter DB 300. The threshold line parameter DB 300 has the following data items: a normal image number 301, a camera 302 in which the identifier of the imaging device 10 that captured the normal image is set, a group 303 in which the identifier of the imaging area is set, a level 304 in which the water level associated with the normal image is set, a start point coordinate 305 in which the coordinate of one end of the water level boundary line in the normal image is set, an end point coordinate 306 in which the coordinate of the other end of the water level boundary line in the normal image is set, and a rectangle center coordinate 307 in which the coordinate of the water level reference point in the normal image is set.

[0057] <Initial setting information setting process> Next, FIG. 10 is a flow diagram illustrating the details of the initial setting information setting process s40.

[0058] The initial setting information setting unit 103 displays an initial setting input screen that accepts input of the initial setting information 113 (s41). Then, the initial setting information setting unit 103 accepts input of initial settings from the user through the initial setting input screen (s42). The initial setting information setting unit 103 displays a screen that allows the user to confirm whether the initial settings entered by the user are correct (s43). When there is confirmation input from the user, the initial setting information setting unit 103 registers the entered initial settings in the initial setting information 113 (s44). Note that the initial setting information setting process s40 may be executed by the water level management system 20 calling the water level estimation system 30.

[0059] (Initial setting information) 11 is a diagram showing an example of the initial setting information 113. The initial setting information 113 stores setting information 1131 indicating that the timing for notifying the water level is to be before the water level exceeds a reference water level. If the setting information 1131 is not particularly specified, a notification of the maximum water level is made when the water level exceeds the reference water level (default value 1132). The initial setting information 113 may be set for each imaging device 10 or each imaging area.

[0060] <Estimated start time setting process> FIG. 12 is a flow diagram illustrating the details of the estimation start time setting process s50.

[0061] The estimation start time setting unit 104 transmits estimation request information 112 to the water level management system 20 (s51). Then, the estimation start time setting unit 104 receives the estimation start time from the water level management system 20 (s52). The estimation start time setting unit 104 stores the received estimation start time (s53). Note that the estimation start time setting process s50 may be executed by the water level management system 20 calling the water level estimation system 30.

[0062] <Water level determination process> FIG. 13 is a flow diagram illustrating the details of the water level determination process s60.

[0063] The water level determination unit 105 acquires the current time (s61). The water level determination unit 105 determines whether the current time is the estimated time of the water level of the river 40 (the estimation start time or the second or subsequent estimation time) (s62).

[0064] If the current time is the water level estimation time, the water level determination unit 105 executes a water level estimation process s63 to estimate the current water level of the river 40. The water level estimation process s63 will be described in detail later.

[0065] The water level determination unit 105 receives the next water level of the river 40 from the water level management system 20 at any time. The time information is waiting to be received (s64).

[0066] The water level determination unit 105 determines whether or not to estimate the water level of the river 40 again (s65). For example, the water level determination unit 105 determines whether or not information on the next water level estimation time has been received within a certain period of time.

[0067] If the water level of the river 40 is to be estimated again (s65: NO), the water level determination unit 105 executes the process of s61, and if the estimation of the water level of the river 40 is to be ended (s65: YES), the water level determination process s60 ends.

[0068] The water level determination unit 105 may determine the timing for making an estimation based on information preset therein.

[0069] <Water level estimation processing> FIG. 14 is a flow diagram illustrating the details of the water level estimation process s63. The water level determination unit 105 selects one of the photographing devices 10 and transmits a request to acquire a current image of the river 40 (hereinafter referred to as a current image or a second image) photographed by the photographing device 10 from the water level management system 20 (s631). Then, the water level determination unit 105 receives the current image of the river 40 (s632).

[0070] The water level determination unit 105 analyzes the current image and identifies the water surface portion of the river 40 on the current image (s633).

[0071] Specifically, the water level determination unit 105 inputs the current image into the water surface determination model 111, and outputs information (hereinafter referred to as pixel information) indicating whether each coordinate (pixel) on the current image is part of the water surface of the river 40 or some other part.

[0072] The water level determination unit 105 estimates the current water level of the river 40 based on the water surface portion of the river 40 on the current image identified in s633 and each water level boundary line in the normal image corresponding to the current image (normal image captured by the same photographing device 10) (s634).

[0073] Specifically, the water level determination unit 105 determines the water level of the current image based on the water level boundary lines of each level on the normal image, information on the direction of rising water on the normal image, and pixel information of the current image.

[0074] For example, the water level determination unit 105 identifies the area of ​​the water level boundary line for each level in the coordinate system on the normal image based on the coordinates of the water level boundary line for each level acquired from the threshold line parameter DB 300. Then, based on the identified area and pixel information of the current image, the water level determination unit 105 determines whether the water level boundary line for level 1 does not match any part of the water surface of the current image (whether the water level boundary line is completely above the water surface). If even a part of the water level boundary line is above the water surface, the unit repeats the process for the water level boundary line of the next level (level 2, 3, ...) until the water level boundary line no longer matches the water surface at all. Then, when the water level boundary line for a certain level no longer matches the water surface at all, the unit determines that the level at that time is the water level of the current image. If such a level cannot be identified, the unit 105 determines that the water level of the current image exceeds the normal water level.

[0075] The water level determination unit 105 determines whether the result of estimating the water level in s634 satisfies the notification conditions indicated by the initial setting information 113. If the initial setting information 113 satisfies the conditions, the water level determination unit 105 transmits information about the result of estimating the water level in s634 (for example, the identifier, group, and water level of the image capturing device 10) to the water level management system 20. (s635) After that, the water level control system 20 displays the received information on the screen.

[0076] Thereafter, the water level determination unit 105 checks whether or not there is an unselected photographed image (s636). If there is an unselected photographed image (s636: NO), the water level determination unit 105 selects the unselected photographed image and executes the process of s631. If there is no unselected photographed image (s636: NO), the water level estimation process s63 ends.

[0077] 15 is a diagram showing an example of a screen (water level estimation result screen) showing information about the water level of the river 40 displayed by the water level management system 20. The water level estimation result screen 500 displays the name 501 of the river 40, the identifier 502 of the imaging device 10, the name 503 of the group, the current water level 504 of the river 40, and information 505 indicating whether the water level is within the normal water level range.

[0078] As described above, the water level estimation system 30 of this embodiment sets the boundary (water level boundary line) between the levee 402 and the water of the river 40, which indicates the water level of the river 40, for a plurality of water levels on a normal image taken from a predetermined direction of the river 40, which is a reservoir, and specifies the direction of fluctuation of the water level of the river 40 on the normal image by setting a water level reference point on each water level boundary line.The water level estimation system 30 then specifies the water surface portion in the current image and estimates the water level in the current image based on the water level boundary line at each water level on the normal image, the direction of fluctuation of the water level on the normal image, and the water surface portion in the current image.

[0079] In this way, the water level estimation system 30 of this embodiment can calculate the current water level shown in the image of the river 40 by setting each water level line of the river 40 and a reference point for the direction of water level fluctuation on the normal image. In other words, the water level estimation system 30 of this embodiment can accurately estimate the water level of the reservoir with a simple procedure. For example, the water level of the river can be accurately estimated at any timing regardless of the position of the imaging device 10.

[0080] In addition, the water level estimation system 30 of this embodiment sets a straight line segment or curve as the water level boundary line, and sets the position of the center of gravity of a rectangle whose diagonal is the straight line segment or a rectangle whose diagonal is a line segment connecting both ends of the curve as the water level reference point.

[0081] By this process, the water level reference point that represents the fluctuation of the water level can be easily set.

[0082] Furthermore, the water level estimation system 30 of this embodiment sets straight line segments or curved portions approximately parallel to each other as the water level boundary lines.

[0083] This allows an accurate water level reference point to be set as a point representing the reference for water level fluctuations.

[0084] Furthermore, the water level estimation system 30 of this embodiment identifies the corresponding water level from among the water levels set in the normal image as the water level of the river 40 in the current image.

[0085] In this way, the user can easily grasp the current state of the water level of the river 40, for example, whether there is a risk of flooding.

[0086] As described above, the present invention is not limited to the above-described embodiments, and can be implemented using any components within the scope of the gist of the present invention. The above-described embodiments and modifications are merely examples, and the present invention is not limited to these contents as long as the characteristics of the invention are not impaired. Furthermore, although various embodiments and modifications have been described above, the present invention is not limited to these contents. Other aspects that are conceivable within the scope of the technical idea of ​​the present invention are also included in the present invention. included within the scope of the invention.

[0087] For example, some of the functional units provided in each device of this embodiment may be provided in another device, or functional units provided in another device may be provided in the same device.

[0088] In addition, although this embodiment is intended to estimate the water level of a river, it can also be applied to estimating the water level of a specified water storage area (water storage structure) such as a lake, swamp, dam, or sea (specifically, a port, etc.) where water is stored.

[0089] Furthermore, in this embodiment, the water level estimation system 30 determines the direction of rising water, but it may also determine the direction of falling water.

[0090] Furthermore, in this embodiment, the water surface determination model 111 is a trained model created by machine learning, but other types of numerical models (for example, a model that performs image pattern matching) may be used as long as they are capable of distinguishing between the water surface and other parts.

[0091] In addition, in this embodiment, the first water level boundary line, the second water level boundary line, and the third water level boundary line are straight lines and parallel to each other (FIG. 6), but they do not necessarily have to be parallel to each other. For example, as a result of the user inputting the water level boundary lines in accordance with the state of the water surface of the river 40 and the slope of the levee on the river-front side, the first water level boundary line, the second water level boundary line, and the third water level boundary line do not necessarily have to be parallel to each other. [Explanation of symbols]

[0092] 1. Water level monitoring system 10 Imaging equipment 20 Water Level Control System 30 Water Level Estimation System

Claims

1. a storage device that stores first images of a water storage section that stores water, the first images being captured from a predetermined direction; a process of setting a boundary between the water storage portion and the water on each of the first images, the boundary being a straight line segment or a curved line segment that are approximately parallel to each other, and indicating the water level of the water storage portion, for each of a plurality of water levels including at least a first water level and a second water level that is higher than the first water level; a process of specifying a direction of water level increase from at least the first water level to the second water level on the first image of the reservoir by setting the position of the center of gravity of a rectangle whose diagonal is the straight line segment or a rectangle whose diagonal is a line segment connecting both ends of the curved section as the reference point of the boundary portion for at least the first and second water levels; At a set timing, a process of acquiring a second image of the water storage section taken from the predetermined direction and identifying a portion of the water surface in the acquired second image; a process of estimating the water level in the second image based on the area of ​​the boundary in the coordinate system on the first image, the rising direction of the water level on the first image, and the specified water surface part on the second image, which is based on the reference point of the boundary at the plurality of water levels on the first image, determining whether the boundary for the first water level coincides with any part of the water surface of the second image, and if the boundary is even partially on the water surface, repeating a series of processes of making the determination for the boundary for the second water level for each water level in an order according to the rising direction until the boundary no longer coincides with the water surface at all, and estimating the water level as the water level of the second image when the boundary for a certain water level no longer coincides with the water surface at all; a process of outputting information indicating the estimated water level; A water level estimation system comprising a processing device that executes the above.

2. The water level estimation system according to claim 1 , wherein the processing device identifies a corresponding water level from among the water levels set in the first image as the water level in the second image.

3. The water level estimation system according to claim 1 , wherein the processing device receives a setting of the boundary on the first image from a user via an input device.

4. The water level estimation system according to claim 1 , wherein the processing device recognizes the boundary between the water and the non-water portion in the first image based on a predetermined numerical model, and outputs the recognized boundary as a candidate for the boundary portion.

5. The information processing device storing a first image of the water storage unit in which water is stored, taken from a predetermined direction; a process of setting a boundary between the water storage portion and the water on each of the first images, the boundary being a straight line segment or a curved line segment that are approximately parallel to each other, and indicating the water level of the water storage portion, for each of a plurality of water levels including at least a first water level and a second water level that is higher than the first water level; a process of specifying a direction of water level increase from at least the first water level to the second water level on the first image of the reservoir by setting the position of the center of gravity of a rectangle whose diagonal is the straight line segment or a rectangle whose diagonal is a line segment connecting both ends of the curved section as the reference point of the boundary portion for at least the first and second water levels; At a set timing, a process of acquiring a second image of the water storage section taken from the predetermined direction and identifying a portion of the water surface in the acquired second image; a process of estimating the water level in the second image based on the area of ​​the boundary in the coordinate system on the first image, the rising direction of the water level on the first image, and the specified water surface part on the second image, which is based on the reference point of the boundary at the plurality of water levels on the first image, determining whether the boundary for the first water level coincides with any part of the water surface of the second image, and if the boundary is even partially on the water surface, repeating a series of processes of making the determination for the boundary for the second water level for each water level in an order according to the rising direction until the boundary no longer coincides with the water surface at all, and estimating the water level as the water level of the second image when the boundary for a certain water level no longer coincides with the water surface at all; and outputting information indicating the estimated water level. Water level estimation method.

Images