Water level measuring device and method for producing water level data

Abstract

This invention provides a water level measuring device that can measure water levels more reliably, and a method for producing water level data. [Solution] In the water level measuring device, the determination unit 21 of the processing unit 2 determines whether or not it is possible to distinguish the boundary line between the water surface-related structure and the water surface in the measurement image acquired by the camera 1. If the determination unit 21 determines that it is not possible to distinguish between the water surface-related structure and the water surface, the setting unit 22 sets a point on the boundary line between the water region and a region in the non-water region that is at a different location from the fixed structure as the boundary point. The setting unit 22 sets the point where a virtual horizontal line passing through the boundary point intersects the water surface-related structure as the water level determination point and sets at least one position reference point. Based on the positions of the water level determination point and the position reference point in the measurement image, the extraction unit 24 extracts a reference image corresponding to the measurement image from a plurality of reference images as an extracted image.

Description

Technical Field

[0004] ,

[0006] , , , , , ,

[0005] , , , , , ,

[0001] The present disclosure relates to a water level measuring device and a method for producing water level data.

Background Art

[0002] Patent Document 1 discloses a determination device that calculates the river width in a determination image and determines the water level from the river width using a function representing the relationship between the river width and the water level.

Prior Art Documents

Patent Documents

[0003]

Patent Document No. 1

Summary of the Invention

Problems to be Solved by the Invention

Means for Solving the Problems

[0006] The water level measuring device according to this disclosure includes a camera that captures an image of the target area as a measurement image by photographing the target area in which a plurality of fixed structures exist, including water surface-related structures fixed in a state intersecting the water surface of the target to be measured, and a processing unit that processes the measurement image, the processing unit having determined whether or not the boundary line between the water surface-related structures and the water surface can be identified in the measurement image, an image storage unit that stores a plurality of images as a plurality of reference images, which show the target area when the water levels are different from each other with the same field of view as the measurement image and in which the determination unit has determined that the boundary line between the water surface-related structures and the water surface can be identified, and the determination unit has determined that it is not possible to identify the water surface-related structures and the water surface in the measurement image. In this case, the system includes a setting unit which defines the area where the water surface is shown in the measurement image as the water region, and the area other than the water region as the non-water region, and sets a boundary point on the boundary line between the water region and an area in the non-water region that is at a different location from the fixed structure, and sets the point where a virtual horizontal line passing through the boundary point intersects the water surface-related structure as the water level determination point, and sets at least one point on the fixed structure away from the water surface in the measurement image as the position reference point, an extraction unit which extracts a reference image corresponding to the measurement image from a plurality of reference images based on the positions of the water level determination point and at least one position reference point in the measurement image, and a data production unit which uses the extracted image as a specific image and produces water level data that identifies the water level of the water surface based on the specific image. Furthermore, the water level measuring device according to this disclosure includes a camera that captures an image showing the target area as a measurement image by photographing the target area in which a plurality of fixed structures exist, including water surface-related structures fixed in a state intersecting the water surface of the object to be measured, and a processing unit that processes the measurement image. The processing unit includes a determination unit that determines whether or not the boundary line between the water surface-related structures and the water surface can be identified in the measurement image, an image storage unit that stores a plurality of images as a plurality of reference images, which show the target area when the water levels are different from each other with the same field of view as the measurement image and in which the determination unit has determined that the boundary line between the water surface-related structures and the water surface can be identified, if the determination unit has determined that the boundary line between the water surface-related structures and the water surface cannot be identified in the measurement image, an extraction unit that extracts at least one reference image from the plurality of reference images as an extracted image, in which the similarity of each reference image to the measurement image is equal to or greater than a set threshold, and a data production unit that uses the extracted image as a specific image and produces water level data that identifies the water level of the water surface based on the specific image. [Effects of the Invention]

[0007] According to this disclosure, water levels can be measured more reliably. [Brief explanation of the drawing]

[0008] [Figure 1] This is a functional block diagram showing a water level measuring device according to Embodiment 1. [Figure 2] This is an explanatory diagram showing an example of a measurement image acquired by the camera in Figure 1. [Figure 3] This is a functional block diagram showing the processing unit in Figure 1. [Figure 4] Figure 2 is an explanatory diagram showing an example of a measurement image illustrating a situation where the water level gauge is hidden by ice due to the freezing of the river. [Figure 5] This is an explanatory diagram showing an example of a measurement image illustrating the state in which each of the water surface-related structures in Figure 2, namely the water level gauge, embankment, and bridge piers, is hidden by ice due to freezing. [Figure 6]This is an explanatory diagram showing an example of a reference image corresponding to the measurement image in Figure 5. [Figure 7] This flowchart shows the method for producing water level data using the water level measuring device shown in Figure 1. [Figure 8] This is a functional block diagram showing the processing unit of the water level measuring device according to Embodiment 2. [Figure 9] Figure 8 is a flowchart showing the method for producing water level data using the water level measuring device. [Figure 10] This is a configuration diagram showing a first example of a processing circuit that realizes the functions of the processing unit according to Embodiments 1 and 2. [Figure 11] This is a configuration diagram showing a second example of a processing circuit that realizes the functions of the processing unit according to Embodiments 1 and 2. [Modes for carrying out the invention]

[0009] The embodiments for carrying out the subject matter of this disclosure will be described with reference to the attached figures. In each figure, the same or corresponding parts are denoted by the same reference numerals, and redundant explanations are simplified or omitted as appropriate. The subject matter of this disclosure is not limited to the following embodiments, and any modification of any component of the embodiments or omission of any component of the embodiments is possible without departing from the spirit of this disclosure.

[0010] Embodiment 1. Figure 1 is a functional block diagram showing a water level measuring device according to Embodiment 1. In the figure, the water level measuring device includes a camera 1, a processing unit 2, an input unit 3, and a display unit 4.

[0011] Camera 1 acquires an image of the target area as a measurement image by photographing the target area. The target area includes a river, which is the target of water level measurement by the water level measuring device, and several fixed structures.

[0012] The processing unit 2 is connected to the camera 1 either by wire or wirelessly. The processing unit 2 processes the measurement-time image from the camera 1. By processing the measurement-time image, the processing unit 2 produces water level data that identifies the water level of the river surface shown in the measurement-time image. Therefore, the processing unit 2 measures the water level of the river existing in the target range by processing the measurement-time image acquired by the camera 1.

[0013] Each of the input unit 3 and the display unit 4 is connected to the processing unit 2 either by wire or wirelessly. The input unit 3 is designed to be operated by a supervisor or the like. As the input unit 3, a keyboard, a mouse, etc. are used. The processing unit 2 makes settings necessary for processing the measurement-time image according to the operation on the input unit 3. The display unit 4 receives the water level data produced by the processing unit 2 from the processing unit 2 and displays the water level specified by the received water level data.

[0014] FIG. 2 is an explanatory diagram showing an example of the measurement-time image acquired by the camera 1 in FIG. 1. In the target range photographed by the camera 1, there is a river 11, and a water gauge 12, a levee 13, a plurality of bridge piers 14, a bridge girder 15, and a plurality of road lights 16 exist as a plurality of fixed structures.

[0015] The plurality of fixed structures include the water gauge 12, the levee 13, and the plurality of bridge piers 14 as a plurality of water surface-related structures, and the bridge girder 15 and the road lights 16 are included as a plurality of water surface-unrelated structures. Each water surface-related structure is a fixed structure fixed in a state of intersecting the water surface of the river 11. Each water surface-unrelated structure is a fixed structure fixed at a position separated from the water surface of the river ll. Therefore, a boundary line with the water surface of the river 11 is formed on each of the water gauge 12, the levee 13, and the plurality of bridge piers 14 which are water surface-related structures.

[0016] The levee 13 is a water surface-related structure arranged along the length direction X of the river 11. In the present embodiment, the inner wall surface of the levee 13 is inclined with respect to the horizontal plane.

[0017] The water level gauge 12 is a water surface related structure fixed to the inner wall surface of the levee 13. In the present embodiment, the water level gauge 12 is arranged along the inner wall surface of the levee 13. Therefore, in the present embodiment, the water level gauge 12 is inclined with respect to the horizontal plane. A scale for measuring the water level determined by the water surface of the river 11 is displayed on the water level gauge 12.

[0018] The plurality of bridge piers 14 are water surface related structures fixed at intervals in the width direction Y of the river 11. The width direction Y of the river 11 is a horizontal direction orthogonal to the length direction X of the river 11.

[0019] The bridge girder 15 is a water surface unrelated structure supported by the plurality of bridge piers 14. The bridge girder 15 is fixed at a position above the water surface of the river 11 in the vertical direction Z. Each of the length direction X and the width direction Y of the river 11 is orthogonal to the vertical direction Z. The bridge girder 15 is arranged along the width direction Y of the river 11.

[0020] The road lamp 16 is a water surface unrelated structure fixed to the upper part of the levee 13. The road lamp 16 has a support part 16a and a lighting part 16b. The support part 16a is fixed upright on the upper part of the levee 13. The lighting part 16b is fixed to the upper end part of the support part 16a. The lighting part 16b protrudes obliquely from the upper end part of the support part 16a.

[0021] Figure 3 is a functional block diagram showing the processing unit 2 of Figure 1. The processing unit 2 has a determination unit 21, a setting unit 22, an image storage unit 23, an extraction unit 24, and a data production unit 25. <00001 of 9>

[0022] The determination unit 21 receives the measurement-time image from the camera 1. The determination unit 21 determines whether or not it is possible to identify the boundary lines between each water surface related structure, that is, the water level gauge 12, the levee 13, and each bridge pier 14, and the water surface of the river 11 in the measurement-time image by performing image processing on the measurement-time image.

[0023] The determination unit 21 makes a determination that the boundary line between the water surface structure and the water surface of the river 11 can be identified when it is possible to identify the boundary line between the water surface structure and the water surface of the river 11. Conversely, the determination unit 21 makes a determination that the boundary line cannot be identified when it is impossible to identify the boundary line between the water surface structure and the water surface of the river 11. For example, if the water surface around the water level gauge 12 is frozen, the determination unit 21 cannot identify the boundary line between the water level gauge 12 and the water surface of the river 11, and makes a determination that the boundary line cannot be identified for the water level gauge 12.

[0024] The determination unit 21 prioritizes determining whether or not it is possible to identify the boundary line between the water level gauge 12 and the water surface of the river 11 among the multiple water surface-related structures, over other water surface-related structures besides the water level gauge 12.

[0025] If the determination unit 21 determines that the boundary line is identifiable for the water level gauge 12, it determines whether the markings displayed on the water level gauge 12 are identifiable. If the markings on the water level gauge 12 are identifiable, the determination unit 21 determines that the markings are identifiable for the water level gauge 12. If the markings on the water level gauge 12 are not identifiable, the determination unit 21 determines that the markings are not identifiable for the water level gauge 12. If the determination unit 21 determines that the markings are identifiable for the water level gauge 12, it does not make any determinations regarding the boundary line between water surface-related structures other than the water level gauge 12 and the water surface of the river 11.

[0026] If the determination unit 21 determines that boundary line identification is impossible for the water level gauge 12, it then determines whether or not it is possible to identify the boundary lines between the water surface-related structures other than the water level gauge 12, namely the embankment 13 and each bridge pier 14, and the water surface of the river 11. An example of when the determination unit 21 determines that boundary line identification is impossible for the water level gauge 12 is when the water level gauge 12 is hidden by snow, ice, etc., within the target area.

[0027] Here, Figure 4 is an explanatory diagram showing an example of a measurement image showing the state in which the water level gauge 12 is hidden by ice due to the freezing of the river 11 in Figure 2. In the example shown in Figure 4, the water surface around the water level gauge 12 is frozen, while the water surface around the bridge pier 14 is not frozen. As a result, in the example shown in Figure 4, the determination unit 21 determines that the boundary line cannot be identified for the water level gauge 12, and determines that the boundary line can be identified for the bridge pier 14.

[0028] If the determination unit 21 determines that the boundary line cannot be identified or the scale cannot be identified for the water level gauge 12, the setting unit 22 sets at least one position reference point and a water level determination point for determining the water level of the river 11 in the measurement image.

[0029] In this embodiment, the setting unit 22 sets two position reference points A and B in the measurement image. Each position reference point A and B is set to avoid locations where identification is expected to be difficult due to snow cover, freezing, etc. In the example shown in Figure 4, one of the two position reference points A and B is set at the bottom of the bridge girder 15, and the other position reference point B is set at the bottom of the lighting unit 16b.

[0030] The reference point data that identifies the setting positions of each reference point A and B in the measurement image is pre-registered in the setting unit 22 by an operation on the input unit 3. The setting unit 22 sets each reference point A and B in the measurement image based on the pre-registered reference point data.

[0031] In this embodiment, the setting unit 22 sets one water level determination point in the measurement image. When the setting unit 22 sets the water level determination point C in the measurement image, it defines the area in the measurement image where the water surface of the river 11 is shown as a water region, and the area other than the water surface of the river 11 as a non-water region.

[0032] The setting unit 22 has an AI (Artificial Intelligence) unit that stores a trained model that outputs images identifying water regions and non-water regions in response to an input image showing the target area. The trained model is created by performing machine learning based on multiple past images obtained by camera 1 capturing the target area. The setting unit 22 identifies water regions and non-water regions in the measurement image by processing the measurement image using the algorithm of the trained model stored in the AI ​​unit.

[0033] If the determination unit 21 determines that a boundary line can be identified for at least one of the water surface-related structures, the setting unit 22 sets a point on the boundary line between the water area and the area where the water surface-related structure for which a boundary line identification determination was made is indicated, within the non-water area, as the water level determination point C. In the example shown in Figure 4, the point on the boundary line between the water area and the non-water area where the bridge pier 14 is indicated is set as the water level determination point C by the setting unit 22.

[0034] On the other hand, if the determination unit 21 determines that it is impossible to identify the boundary line for any of the water surface-related structures, the setting unit 22 will set a point on the boundary line between the water area and a non-water area at a location different from that of the fixed structure as the boundary point. In other words, if the determination unit 21 determines that it is impossible to identify the boundary line between each water surface-related structure and the water surface of the river 11, the setting unit 22 will set a point on the boundary line between the water area and a non-water area at a location different from that of the area where the fixed structure is indicated as the boundary point.

[0035] An example of a case in which the determination unit 21 determines that the boundary line cannot be identified for any of the water surface-related structures is when each of the water surface-related structures in the target area is hidden by ice, snow, etc. In this case, it is thought that the formations of ice, snow, etc. due to the freezing of the river 11 extend from the embankment 13 toward the center of the river 11. Since the center of the river 11 is less likely to freeze due to the flow of water, a boundary line is created between the formations of ice, snow, etc. and the water surface of the river 11. The height of the boundary line between the formations of ice, snow, etc. and the water surface of the river 11 coincides with the water level of the river 11.

[0036] Here, Figure 5 is an explanatory diagram showing an example of a measurement image in which each of the water surface-related structures in Figure 2, namely the water level gauge 12, the embankment 13, and each bridge pier 14, is hidden by ice due to freezing. In Figure 5, in the measurement image, ice 17 due to the freezing of the river 11 extends from the embankment 13 toward the center of the river 11. Since the ice 17 is formed by the freezing of the river 11, it is exposed from the water surface of the river 11 at a different location from the fixed structures in the target area.

[0037] The non-water region identified by the setting unit 22 in the measurement image includes the region where ice 17 is shown. If the determination unit 21 determines that the boundary line cannot be identified for any of the water surface-related structures, the setting unit 22 identifies a point on the boundary line between the region where ice 17 is shown and the water region, which is a region in the non-water region at a different location from the fixed structure, as boundary point D.

[0038] The position of the boundary point D in the longitudinal direction X of the river 11 is determined to be a coordinate position predetermined as a fixed coordinate value in the longitudinal direction X of the river 11. In this embodiment, the coordinate value indicating the position of the water level gauge 12 in the longitudinal direction X of the river 11 is predetermined as a fixed coordinate value.

[0039] The setting unit 22 has pre-registered position coordinate data that shows the positions of multiple points on the water level gauge 12, associated with coordinate values ​​in the width direction Y and the vertical direction Z of the river 11. The registration of coordinate values ​​that show the positions of multiple points on the water level gauge 12 can be performed by operating the input unit 3.

[0040] The setting unit 22 identifies the boundary point D in the measurement image, and then sets the point where a virtual horizontal line E passing through the boundary point D intersects with a water surface-related structure as the water level determination point C. The virtual horizontal line is a line on a virtual plane perpendicular to the vertical direction Z. The position of the water level determination point C is determined by the setting unit 22 based on position coordinate data that has been registered in the setting unit 22 in advance. In this embodiment, the point where the virtual horizontal line E passing through the boundary point D and along the width direction Y of the river 11 intersects with the water level gauge 12 is set as the water level determination point C.

[0041] The image storage unit 23 stores multiple reference images for comparison with the measurement image. The multiple reference images stored in the image storage unit 23 are images that show the target range when the water levels of the river 11 are different, with the same field of view as the measurement image. Furthermore, the multiple reference images are images for which the determination unit 21 has previously determined that the boundary line between the water surface-related structures and the water surface of the river 11 can be identified. In addition, the multiple reference images are images for which the determination unit 21 has previously determined that the scale displayed on the water level gauge 12 can be identified. Therefore, each reference image is an image for which the determination unit 21 has determined that the boundary line can be identified and the scale can be identified.

[0042] In this embodiment, among a plurality of images showing the target range previously captured by camera 1, images in which the determination unit 21 has determined that the boundary line is identifiable and the scale is identifiable relative to the water level gauge 12 are stored in the image storage unit 23 as a plurality of reference images.

[0043] The extraction unit 24 extracts a reference image corresponding to the measurement image from a plurality of reference images stored in the image storage unit 23, based on the positions of the water level determination point and at least one position reference point in the measurement image. Therefore, in this embodiment, the extraction unit 24 extracts the extraction image from a plurality of reference images based on the positions of the water level determination point C and the two position reference points A and B in the measurement image. The extraction unit 24 extracts the extraction image from a plurality of reference images unless the determination unit 21 has made a boundary line identification determination and a scale identification determination for the water level gauge 12.

[0044] Figure 6 is an explanatory diagram showing an example of a reference image corresponding to the measurement image in Figure 5. The extraction unit 24 sets the water level determination point C and the two position reference points A and B in each reference image so that their respective positional relationships are the same as in the measurement image. The extraction unit 24 also extracts a reference image from multiple reference images as the extracted image, where the position of the water level determination point C is on the boundary line between the water surface-related structure and the water surface of the river 11.

[0045] The extraction unit 24 has an AI unit that stores a trained model that outputs a reference image corresponding to the positional relationship between the water level determination point C and the two positional reference points A and B, in response to the input of the positional relationship between the water level determination point C and the two positional reference points A and B. Based on the positions of the water level determination point C and the two positional reference points A and B set in the measurement image, the extraction unit 24 searches for multiple reference images using the algorithm of the trained model stored in the AI ​​unit, thereby extracting an image from multiple reference images.

[0046] The data production unit 25 acquires the extracted image extracted from multiple reference images by the extraction unit 24 as a specific image, unless the determination unit 21 has determined that the boundary line and scale markings are identifiable for the water level gauge 12. In addition, if the determination unit 21 has determined that the boundary line and scale markings are identifiable for the water level gauge 12, the data production unit 25 acquires the measurement image from the determination unit 21 as a specific image.

[0047] The data production unit 25 produces water level data that identifies the water level of the river 11 in a specific image, based on the specific image. In this embodiment, the data production unit 25 reads the relationship between the position of the boundary line between the water level gauge 12 and the water surface of the river 11 in the specific image and the scale of the water level gauge 12, thereby producing water level data that identifies the water level of the river 11. If the water level determination point C set on the water level gauge 12 remains in the specific image, the data production unit 25 can also produce water level data that identifies the water level of the river 11 by reading the scale of the water level gauge 12 that coincides with the position of the water level determination point C. The water level data produced by the data production unit 25 is transmitted to the display unit 4.

[0048] Next, a method for producing water level data that identifies the water level on the surface of the river 11 using a water level measuring device will be described. Figure 7 is a flowchart showing the method for producing water level data using the water level measuring device shown in Figure 1. In the water level data production method, in step S1, when camera 1 photographs the target area and acquires a measurement image, the measurement image data is transmitted from camera 1 to processing unit 2. When processing unit 2 receives the measurement image data from camera 1, processing unit 2 proceeds to step S2. In the water level data production method, the processing in step S1 is the photography step.

[0049] In step S2, the determination unit 21 determines whether or not the boundary line between the water level gauge 12 and the water surface of the river 11 can be identified in the measurement image.

[0050] If, in step S2, the determination unit 21 determines that it is possible to identify the boundary line between the water level gauge 12 and the water surface of the river 11, that is, it makes a determination that the boundary line for the water level gauge 12 is identifiable, then the processing unit 2 proceeds to step S3.

[0051] In step S3, the determination unit 21 determines whether or not the scale of the water level gauge 12 can be identified in the measurement image.

[0052] If the determination unit 21 determines in step S3 that the scale of the water level gauge 12 is identifiable, that is, if it determines that the scale of the water level gauge 12 is identifiable, the processing unit 2 proceeds to step S4. In step S4, the determination unit 21 transmits the measurement image to the data production unit 25, and the data production unit 25 receives the measurement image as a specific image.

[0053] If, in step S2, the determination unit 21 determines that it is impossible to identify the boundary line between the water level gauge 12 and the water surface of the river 11, that is, it determines that the boundary line for the water level gauge 12 cannot be identified, the processing unit 2 proceeds to step S5.

[0054] In step S5, the determination unit 21 determines whether it is possible to identify the boundary line between the water surface-related structures other than the water level gauge 12 and the water surface of the river 11. In the water level data production method, steps S2, S3, and S5 are each determination steps.

[0055] If, in step S5, the determination unit 21 determines that it is possible to identify the boundary line between the water surface-related structures other than the water level gauge 12 and the water surface of the river 11, that is, it determines that the boundary line of the water surface-related structures other than the water level gauge 12 is identifiable, the processing unit 2 proceeds to step S6. Also, if, in step S3, the determination unit 21 determines that it is impossible to identify the scale of the water level gauge 12, that is, it determines that the scale of the water level gauge 12 is identifiable, the processing unit 2 proceeds to step S6.

[0056] In step S6, the setting unit 22 sets a point on the boundary line between the non-water area and the water area in the measurement image as a water level determination point. The water level determination point is a point on the boundary line between the water area and the area where the determination unit 21 has determined that the boundary line with the water surface of the river 11 can be identified within the non-water area. In step S6, the setting unit 22 also sets at least one point of a fixed structure away from the water surface of the river 11 in the measurement image as a position reference point. In step S6 of this embodiment, as shown in Figure 4, the setting unit 22 sets one water level determination point C on the boundary line and two position reference points A and B in the measurement image. After this, the processing unit 2 proceeds to step S8.

[0057] On the other hand, if in step S5 the determination unit 21 determines that it is impossible to identify the boundary line between water surface-related structures other than the water level gauge 12 and the water surface of the river 11, that is, if the determination unit 21 determines that the boundary line for water surface-related structures other than the water level gauge 12 cannot be identified, the processing unit 2 proceeds to step S7.

[0058] In step S7, the setting unit 22 sets a point on the boundary line between the non-water area and the water area in the measurement image as a boundary point, and sets the point where a virtual horizontal line passing through the boundary point intersects a water surface-related structure as the water level determination point. In step S7, the setting unit 22 also sets at least one position reference point in the measurement image. In step S7 of this embodiment, as shown in Figure 5, the setting unit 22 sets the point where a virtual horizontal line E passing through boundary point D intersects the water level gauge 12 as the water level determination point C in the measurement image, and sets two position reference points A and B in the measurement image. After this, the processing unit 2 proceeds to step S8. In the water level data production method, the processing in steps S6 and S7 are setting steps.

[0059] In step S8, the extraction unit 24 extracts a reference image corresponding to the measurement image from a plurality of reference images as an extracted image, based on the positions of the water level determination point and at least one position reference point set in the measurement image in step S6 or step S7. In this embodiment, in step S8, the extraction unit 24 sets the water level determination point C and the two position reference points A and B in each reference image so that the positional relationship between the water level determination point C and the two position reference points A and B is the same as in the measurement image. Also in step S8, the extraction unit 24 extracts a reference image from a plurality of reference images as an extracted image, in which the position of a point on the boundary line between the water surface-related structure and the water surface of the river 11 coincides with the position of the water level determination point C. In the water level data production method, the processing in step S8 is the extraction step.

[0060] After this, the processing unit 2 proceeds to step S9. In step S9, the extraction unit 24 transmits the extracted image to the data production unit 25, and the data production unit 25 receives the extracted image as a specific image.

[0061] When the data production unit 25 receives a specific image in step S4 or step S9, the processing unit 2 proceeds to step S10. In step S10, the data production unit 25 produces water level data that identifies the water level of the river 11 in the specific image based on the specific image. In the water level data production method, steps S4, S9, and S10 are data production steps.

[0062] The water level data produced by the data production unit 25 is transmitted from the data production unit 25 to the display unit 4. When the display unit 4 receives the water level data, the water level identified by the data is displayed on the display unit 4.

[0063] In this type of water level measuring device, the determination unit 21 determines whether it is possible to distinguish the boundary line between the water surface-related structure and the water surface of the river 11 in the measurement image. If the determination unit 21 determines that it is impossible to distinguish between the water surface-related structure and the water surface of the river 11 in the measurement image, the setting unit 22 sets a point on the boundary line between the water region and a region in the non-water region that is at a different location from the fixed structure as boundary point D. The setting unit 22 also sets the point where a virtual horizontal line passing through boundary point D intersects the water surface-related structure as water level determination point C, and sets points on the fixed structure that are far from the water surface of the river 11 in the measurement image as position reference points A and B. Based on the positions of water level determination point C and position reference points A and B in the measurement image, the extraction unit 24 extracts a reference image corresponding to the measurement image from among multiple reference images stored in the image storage unit 23 as an extracted image. Based on the extracted image, the data production unit 25 produces water level data that identifies the water level of the water surface of the river 11 in the extracted image.

[0064] Therefore, even if it becomes impossible to distinguish the boundary between water surface-related structures and the water surface of the river 11 due to freezing or snow accumulation in the river 11, water level data for the river 11 can be produced more reliably by, for example, distinguishing the boundary between the ice 17 formed in the river 11 and the water surface of the river 11. This allows for more reliable measurement of the water level of the river 11.

[0065] Furthermore, one of the multiple water surface-related structures is a water level gauge 12 with a scale for reading the water level. The determination unit 21 prioritizes determining whether the boundary line between the water level gauge 12 and the water surface of the river 11 can be identified over other water surface-related structures. If the determination unit 21 determines that the boundary line between the water level gauge 12 and the water surface of the river 11 can be identified, it then determines whether the scale of the water level gauge 12 can be identified in the measurement image. Therefore, if the scale of the water level gauge 12 can be identified, the relationship between the position of the boundary line between the water level gauge 12 and the water surface of the river 11 and the scale of the water level gauge 12 can be easily read in the measurement image without having to extract an image from multiple reference images. This makes it possible to easily produce water level data for the river 11 and easily measure the water level of the river 11.

[0066] Furthermore, in this method of producing water level data, the determination step determines whether or not it is possible to identify the boundary line between the water surface-related structure and the water surface of the river 11 in the measurement image. In the setting step, if the determination step determines that it is impossible to identify the boundary line between the water surface-related structure and the water surface of the river 11 in the measurement image, a point on the boundary line between the water area and a non-water area at a location different from the fixed structure is set as boundary point D. Also in the setting step, the point where a virtual horizontal line passing through boundary point D intersects the water surface-related structure is set as water level determination point C, and points on the fixed structure that are far from the water surface of the river 11 in the measurement image are set as position reference points A and B. In the extraction step, based on the respective positions of water level determination point C and position reference points A and B in the measurement image, a reference image corresponding to the measurement image is extracted as an extracted image from among multiple reference images stored in the image storage unit 23. In the data production step, the extracted image is set as a specific image, and water level data that identifies the water level of the water surface of the river 11 in the extracted image is produced based on the specific image.

[0067] Therefore, even if the boundary between water surface-related structures and the water surface of the river 11 becomes impossible to distinguish due to freezing or snow accumulation in the river 11, water level data for the river 11 can be produced more reliably. This allows for more accurate measurement of the water level of the river 11.

[0068] In Embodiment 1, a water level gauge 12, which displays a scale for measuring water level, is present in the target area as one of several water surface-related structures. However, the water level gauge 12 does not necessarily have to be present in the target area. In this case, multiple reference images are pre-stored in the image storage unit 23 in association with water level data that identifies the water level of the river 11. In this way, even if the water level gauge 12 is not shown in a specific image, the water level data can be produced by the data production unit 25 by reading the water level data associated with the specific image.

[0069] Furthermore, in Embodiment 1, the extraction unit 24 has an AI unit. However, instead of using the trained model of the AI ​​unit, the extraction image may be extracted from multiple reference images by searching for multiple reference images based on the positions of the water level determination point C and the two position reference points A and B set in the measurement image.

[0070] Furthermore, in Embodiment 1, the fixed coordinate value indicating the position of the boundary point D in the longitudinal direction X of the river 11 is the coordinate value indicating the position of the water level gauge 12. However, the fixed coordinate value is not limited to the coordinate value indicating the position of the water level gauge 12. For example, the fixed coordinate value indicating the position of the boundary point D in the longitudinal direction X of the river 11 may be the coordinate value indicating the inner wall surface of the embankment 13.

[0071] Furthermore, in Embodiment 1, the number of position reference points set by the setting unit 22 in the measurement image is two, position reference points A and B. However, the number of position reference points set by the setting unit 22 in the measurement image is not limited to two, but may be one or three or more. The more position reference points set in the measurement image, the more accurate the extraction of the reference image corresponding to the measurement image from multiple reference images can be improved.

[0072] Embodiment 2. Figure 8 is a functional block diagram showing the processing unit of the water level measuring device according to Embodiment 2. The processing unit 2 includes a determination unit 21, an image storage unit 23, an extraction unit 24, and a data production unit 25. The functions of the image storage unit 23 and the data production unit 25 are the same as in Embodiment 1.

[0073] The determination unit 21 determines whether or not the boundary line between the water level gauge 12 and the water surface of the river 11 can be identified in the measurement image. If the boundary line between the water level gauge 12 and the water surface of the river 11 can be identified, the determination unit 21 makes a determination that the boundary line can be identified for the water level gauge 12. If the boundary line between the water level gauge 12 and the water surface of the river 11 cannot be identified, the determination unit 21 makes a determination that the boundary line cannot be identified for the water level gauge 12.

[0074] If the determination unit 21 determines that the boundary line is identifiable for the water level gauge 12, it then determines whether the markings displayed on the water level gauge 12 are identifiable. If the markings on the water level gauge 12 are identifiable, the determination unit 21 determines that the markings are identifiable for the water level gauge 12. If the markings on the water level gauge 12 are not identifiable, the determination unit 21 determines that the markings are not identifiable for the water level gauge 12.

[0075] If the determination unit 21 determines that it is impossible to distinguish the boundary line between the water level gauge 12 and the water surface of the river 11 in the measurement image, the extraction unit 24 calculates the similarity of each reference image to the measurement image. In addition, if the determination unit 21 determines that it is impossible to distinguish the scale displayed on the water level gauge 12, the extraction unit 24 also calculates the similarity of each reference image to the measurement image.

[0076] The extraction unit 24 calculates the similarity of each reference image to the measurement image, and then extracts at least one reference image from the multiple reference images as the extracted image, where the similarity of each reference image to the measurement image is equal to or greater than a set threshold. In other words, if the determination unit 21 determines that the boundary line cannot be identified or the scale cannot be identified for the water level gauge 12, the extraction unit 24 extracts at least one reference image from the multiple reference images as the extracted image, where the similarity of each reference image to the measurement image is equal to or greater than a set threshold.

[0077] In this embodiment, if there are multiple reference images whose similarity to the measurement image is greater than or equal to a set threshold, the single reference image with the highest similarity to the measurement image is extracted from the multiple reference images by the extraction unit 24 as the extracted image.

[0078] An example of when the determination unit 21 determines that the boundary line is indistinguishable or the scale is indistinguishable for the water level gauge 12 is when, due to heavy rain, blizzards, dense fog, etc., the boundary line between the water level gauge 12 and the surface of the river 11, and the outline of the water level gauge 12, are faintly visible in the measurement image. The extraction unit 24 calculates the similarity of each reference image to the measurement image based on the boundary line and outline that are faintly visible in the measurement image.

[0079] The extraction unit 24 has an AI unit that stores a trained model that, in response to an input image indicating a target range, outputs an image in which the similarity of each reference image to the input image is equal to or greater than a set threshold. The extraction unit 24 performs image matching between the measurement image and each reference image using the algorithm of the trained model stored in the AI ​​unit, and calculates the matching rate of each reference image to the measurement image as the similarity of each reference image to the measurement image. The extraction unit 24 also extracts at least one reference image from among the multiple reference images in which the matching rate is equal to or greater than a set threshold as the extracted image. In this embodiment, the extraction unit 24 extracts the one reference image with the highest matching rate to the measurement image from among the multiple reference images as the extracted image. The other configurations are the same as in Embodiment 1.

[0080] Next, a method for producing water level data that identifies the water level on the surface of the river 11 using a water level measuring device will be described. Figure 9 is a flowchart showing the method for producing water level data using the water level measuring device shown in Figure 8. In the water level data production method of this embodiment, similar to Embodiment 1, in step S1, camera 1 acquires a measurement image, and the measurement image data is transmitted from camera 1 to processing unit 2. Therefore, in this embodiment, similar to Embodiment 1, the processing in step S1 is the shooting step in the water level data production method. When processing unit 2 receives the measurement image data from camera 1, processing in processing unit 2 proceeds to step S2.

[0081] In step S2, the determination unit 21 determines, in the same manner as in Embodiment 1, whether or not the boundary line between the water level gauge 12 and the water surface of the river 11 can be identified in the measurement image. If the determination unit 21 determines in step S2 that the boundary line between the water level gauge 12 and the water surface of the river 11 can be identified, that is, if it determines that the boundary line for the water level gauge 12 can be identified, the processing unit 2 proceeds to step S3.

[0082] In step S3, the determination unit 21 determines whether or not the scale of the water level gauge 12 can be identified in the measurement image, similar to the first embodiment. If the determination unit 21 determines in step S3 that the scale of the water level gauge 12 can be identified, that is, if it determines that the scale of the water level gauge 12 can be identified, the determination unit 21 transmits the measurement image to the data production unit 25, and the processing unit 2 proceeds to step S4. In step S4, the data production unit 25 receives the measurement image as a specific image. In this embodiment, the processing in steps S2 and S3 constitutes the determination steps in the water level data production method.

[0083] If, in step S2, the determination unit 21 determines that it is impossible to identify the boundary line between the water level gauge 12 and the water surface of the river 11, that is, it determines that the boundary line of the water level gauge 12 cannot be identified, the determination unit 21 transmits the measurement image to the extraction unit 24, and the processing unit 2 proceeds to step S11. Similarly, if, in step S3, the determination unit 21 determines that it is impossible to identify the scale of the water level gauge 12, that is, it determines that the scale of the water level gauge 12 cannot be identified, the determination unit 21 transmits the measurement image to the extraction unit 24, and the processing unit 2 proceeds to step S11.

[0084] In step S11, the extraction unit 24 extracts at least one reference image from the plurality of reference images as the extracted image, wherein the similarity of each reference image to the measurement image is equal to or greater than a set threshold, and the processing unit 2 proceeds to step S9. In this embodiment, in step S11, the extraction unit 24 extracts the reference image with the highest similarity to the measurement image from the plurality of reference images as the extracted image. In this embodiment, the processing in step S11 is the extraction step in the water level data production method.

[0085] In step S9, the data production unit 25 receives the extracted image as a specific image.

[0086] When a specific image is received by the data production unit 25 in step S4 or step S9, the processing unit 2 proceeds to step S10. In step S10, the data production unit 25 produces water level data in the same manner as in Embodiment 1. In this embodiment, the processing in steps S4, S9, and S10 constitutes the data production steps in the water level data production method. The subsequent processing is the same as in Embodiment 1.

[0087] In this type of water level measuring device, if the determination unit 21 determines that the boundary line of the water level gauge 12 cannot be identified, the extraction unit 24 extracts at least one reference image from the multiple reference images as the extracted image, provided that the similarity of each reference image to the measurement image is equal to or greater than a set threshold. Therefore, even if the boundary line between the water level gauge 12 and the water surface of the river 11 is unclear in the measurement image due to heavy rain, blizzards, dense fog, etc., the data production unit 25 can identify the water level of the river 11 and produce water level data. This allows for more reliable measurement of the water level of the river 11.

[0088] Furthermore, even if the determination unit 21 determines that the scale markings on the water level gauge 12 are unidentifiable, the extraction unit 24 extracts at least one reference image from the multiple reference images as an extracted image, provided that the similarity of each reference image to the measurement image is equal to or greater than a set threshold. Therefore, even if the scale markings on the water level gauge 12 are unclear in the measurement image, the data production unit 25 can determine the water level on the surface of the river 11 and produce water level data.

[0089] Furthermore, in this method of producing water level data, in the extraction step, at least one reference image is extracted from multiple reference images as the extracted image, provided that the similarity of each reference image to the measurement image is equal to or greater than a set threshold. In the data production step, the extracted image is designated as the specific image in step S9, and water level data is produced based on the specific image in step S10. Therefore, even if the boundary between the water level gauge 12 and the water surface of the river 11 is unclear in the measurement image due to heavy rain, blizzards, dense fog, etc., the water level of the river 11 can be identified, and water level data can be produced. This allows for more reliable measurement of the water level of the river 11.

[0090] In each of the above embodiments, the object for which the water level is measured is the river 11. However, the object for which the water level is measured is not limited to the river 11. For example, harbors, lakes, dams, reservoirs, etc., may also be used as objects for which the water level is measured.

[0091] Furthermore, in Embodiment 2, if the determination unit 21 determines that it is impossible to distinguish the boundary line between the water level gauge 12 and the water surface of the river 11, the extraction unit 24 extracts an image from a plurality of reference images. However, if the determination unit 21 determines that it is impossible to distinguish the boundary line between water surface-related structures other than the water level gauge 12 and the water surface of the river 11, the extraction unit 24 may be configured to extract an image from a plurality of reference images. In this case, a plurality of reference images are pre-stored in the image storage unit 23 in association with water level data that identifies the water level of the water surface of the river 11. In this way, even if the water level gauge 12 is not shown in the specific image, the water level data can be produced by the data production unit 25 by reading the water level data associated with the specific image.

[0092] In Embodiment 2, the extraction unit 24 extracts one reference image with the highest similarity to the measurement image as the extracted image from the multiple reference images. However, it is not limited to this. For example, the extraction unit 24 may extract two or more reference images from the multiple reference images as two or more extracted images, each of which has a similarity to the measurement image equal to or greater than a set threshold. In this case, the data production unit 25 may use the two or more extracted images as two or more specific images, and the average value of the water levels identified in the two or more specific images as the water level of the river 11. Alternatively, in this case, the data production unit 25 may use the two or more extracted images as two or more specific images, and the sum of the values ​​obtained by multiplying the water levels identified in the two or more specific images by their individual matching rates, divided by the sum of the matching rates, as the water level of the river 11.

[0093] Furthermore, the functions of the processing unit 2 according to each of the above embodiments are realized by a processing circuit. Figure 10 is a configuration diagram showing a first example of a processing circuit that realizes the functions of the processing unit 2 according to embodiments 1 and 2. The processing circuit 100 in the first example is dedicated hardware.

[0094] Furthermore, the processing circuit 100 may include, for example, a single circuit, a composite circuit, a programmed processor, a parallel programmed processor, an ASIC (Application Specific Integrated Circuit), an FPGA (Field Programmable Gate Array), or a combination thereof.

[0095] Figure 11 is a configuration diagram showing a second example of a processing circuit that realizes the functions of the processing unit 2 according to Embodiments 1 and 2. The processing circuit 200 of the second example includes a processor 201 and a memory 202.

[0096] In the processing circuit 200, the functions of the processing unit 2 are realized by software, firmware, or a combination of software and firmware. The software and firmware are written as programs and stored in memory 202. The processor 201 realizes the functions of the processing unit 2 by reading and executing the programs stored in memory 202.

[0097] A program stored in memory 202 can be said to cause the computer to execute the procedures or methods described above. Here, memory 202 refers to non-volatile or volatile semiconductor memory such as RAM (Random Access Memory), ROM (Read Only Memory), flash memory, EPROM (Erasable Programmable Read Only Memory), and EEPROM (Electrically Erasable and Programmable Read Only Memory). Magnetic disks, flexible disks, optical disks, compact disks, minidiscs, DVDs, etc., also fall under the category of memory 202.

[0098] Furthermore, some of the functions of the processing unit 2 described above may be implemented using dedicated hardware, while others may be implemented using software or firmware.

[0099] Thus, the processing circuit can realize the functions of the processing unit 2 described above through hardware, software, firmware, or a combination thereof.

[0100] The configurations shown in the embodiments described above are merely examples of the content of this disclosure. The embodiments can be combined with other known technologies. Some parts of the configurations of the embodiments can be omitted or modified without departing from the gist of this disclosure. [Explanation of Symbols]

[0101] 1 Camera, 2 Processing unit, 11 River (object for measuring water level), 12 Water level gauge (water surface related structure), 13 Embankment (water surface related structure), 14 Bridge pier (water surface related structure), 15 Bridge girder (fixed structure), Street light (fixed structure), 21 Judgment unit, 22 Setting unit, 23 Image storage unit, 24 Extraction unit, 25 Data production unit.

Claims

1. A camera that captures an image of the target area as a measurement image by photographing the target area in which multiple fixed structures exist, including water surface-related structures fixed in a position intersecting the water surface of the target to be measured, A processing unit that processes the measurement image and Equipped with, The aforementioned processing unit, A determination unit that determines whether or not the boundary line between the water surface-related structure and the water surface can be identified in the measurement image, An image storage unit stores multiple images as multiple reference images, which show the target range when the water levels are different from each other with the same field of view as the measurement image, and for which the determination unit has determined that the boundary line between the water surface-related structure and the water surface can be identified. If the determination unit determines that it is impossible to distinguish between the water surface-related structure and the water surface in the measurement image, the setting unit sets the area in the measurement image where the water surface is shown as a water area, the area other than the water area as a non-water area, a point on the boundary line between the water area and an area in the non-water area at a different location from the fixed structure as a boundary point, the point where a virtual horizontal line passing through the boundary point intersects the water surface-related structure as a water level determination point, and sets at least one point on the fixed structure away from the water surface in the measurement image as a position reference point. An extraction unit extracts a reference image corresponding to the measurement image from a plurality of reference images, based on the positions of the water level determination point and at least one of the position reference points in the measurement image, A data production unit that uses the extracted image as a specific image and produces water level data that identifies the water level of the water surface based on the specific image. A water level measuring device equipped with the following features.

2. The aforementioned area includes multiple water surface-related structures, Any of the aforementioned water surface-related structures is a water level gauge on which a scale for reading the water level is displayed, The determination unit determines whether the boundary line between the water level gauge and the water surface can be identified, prioritizing this determination over any other water surface-related structures, and if it determines that the boundary line between the water level gauge and the water surface can be identified, it determines whether the scale can be identified in the measurement image. The water level measuring device according to claim 1, wherein the data production unit, when it is determined by the determination unit that the scale can be identified in the measurement image, designates the measurement image as a specific image and produces the water level data based on the specific image.

3. A camera that captures an image of the target area as a measurement image by photographing the target area in which multiple fixed structures exist, including water surface-related structures fixed in a position intersecting the water surface of the target to be measured, A processing unit that processes the measurement image and Equipped with, The aforementioned processing unit, A determination unit that determines whether or not the boundary line between the water surface-related structure and the water surface can be identified in the measurement image, An image storage unit stores multiple images as multiple reference images, which show the target range when the water levels are different from each other with the same field of view as the measurement image, and for which the determination unit has determined that the boundary line between the water surface-related structure and the water surface can be identified. If the determination unit determines that it is impossible to distinguish the boundary line between the water surface-related structure and the water surface in the measurement image, the extraction unit extracts at least one reference image from the plurality of reference images as an extracted image, wherein the similarity of each reference image to the measurement image is equal to or greater than a set threshold. A data production unit that uses the extracted image as a specific image and produces water level data that identifies the water level of the water surface based on the specific image. A water level measuring device equipped with the following features.

4. The aforementioned water surface-related structure is a water level gauge on which a scale for reading the water level is displayed. If the determination unit determines that it is impossible to distinguish the boundary line between the water level gauge and the water surface in the measurement image, it determines whether or not the scale markings can be distinguished in the measurement image. The water level measuring device according to claim 3, wherein, if the determination unit determines that it is impossible to identify the scale in the measurement image, the extraction unit extracts at least one reference image from the plurality of reference images, wherein the similarity of each reference image to the measurement image is equal to or greater than a set threshold, as the extracted image.

5. A method for producing water level data that identifies the water level of the water surface using a water level measuring device according to claim 1 or claim 2, A shooting step in which an image showing the target area is obtained as a measurement image by photographing the target area, A determination step of determining whether or not the boundary line between the water surface-related structure and the water surface can be identified in the measurement image, If, in the determination step, it is determined that it is impossible to distinguish the boundary line between the water surface-related structure and the water surface in the measurement image, the area in the measurement image where the water surface is shown is designated as a water area, and the area other than the water area is designated as a non-water area. Within the non-water area, a point on the boundary line between the area at a different location from the fixed structure and the water area is designated as a boundary point, and the point where a virtual horizontal line passing through the boundary point intersects the water surface-related structure is set as the water level determination point. In the measurement image, at least one point on the fixed structure that is away from the water surface is set as a position reference point. An extraction step of extracting a reference image corresponding to the measurement image from the plurality of reference images, based on the positions of the water level determination point and at least one of the position reference points in the measurement image, A data production step in which the extracted image is designated as a specific image, and water level data is produced based on the specific image to determine the water level of the water surface. A method for producing water level data that includes the following features.

6. A method for producing water level data, which produces water level data that identifies the water level of the water surface using a water level measuring device according to claim 3 or claim 4, A shooting step in which an image showing the target area is obtained as a measurement image by photographing the target area, A determination step of determining whether or not the boundary line between the water surface-related structure and the water surface can be identified in the measurement image, If the determination unit determines that it is impossible to distinguish the boundary line between the water surface-related structure and the water surface in the measurement image, the extraction step involves extracting at least one reference image from the plurality of reference images, wherein the similarity of each reference image to the measurement image is equal to or greater than a set threshold, as the extracted image. A data production step in which the extracted image is designated as a specific image, and water level data is produced based on the specific image to determine the water level of the water surface. A method for producing water level data that includes the following features.

Images