A vision-based floating buoy water level monitoring system and method

Abstract

The present invention discloses a vision-based floating buoy water level monitoring system and method. A floating water level gauge is used to monitor water level information. Compared with the existing water immersion scale, it avoids large deviations in the collected water level information caused by surface reflection, and greatly improves the accuracy and precision of water level monitoring. An infrared supplementary light is added, which can well cope with the situation of unclear pictures caused by poor visibility and low brightness, greatly improving the compatibility of the vision-based floating water level monitoring system and enabling it to well adapt to various situations. By controlling the pan-tilt camera to work, it ensures that the float is at the center of the high-definition camera acquisition frame, providing a good basis for subsequent image recognition and greatly improving the accuracy and stability of image recognition.

Description

Technical Field

[0001] The present invention relates to the technical field of water level monitoring, and particularly relates to a visual-based buoy-type water level monitoring system and method. Background Art

[0002] Water level monitoring is an important means for flood forecasting, flood control command, safe dispatching of water conservancy projects, water resource management and protection. China has a vast territory and numerous rivers, lakes and reservoirs. The water conservancy department needs to manually detect the water level during the flood season, which poses a great potential safety hazard. Currently, in actual work, the use of a scale is relatively common. However, since the scale is submerged in water, it is easy to cause reading errors due to the reflection of the water surface, making it impossible for the water conservancy department to understand the accurate real situation. There are also applications of machine vision monitoring in water level monitoring. However, generally, the scale submerged in water is directly read for water level monitoring. However, due to the reflection of the water surface, errors occur in the video processing of the image, and when the change range of the water level is large, the water surface will deviate from the field of view of the high-definition camera, resulting in the camera being unable to collect suitable images for monitoring, and the monitoring effect is not good. Summary of the Invention

[0003] The technical problem to be solved by the present invention is to provide a visual-based buoy-type water level monitoring system and method, which can effectively collect images of the buoy-type water level gauge, and then obtain water level information by identifying the position of the float in the scale table, greatly improving the accuracy and precision of water level monitoring.

[0004] To achieve the above object, the technical solution adopted by the present invention is: A visual-based buoy-type water level monitoring system, including a buoy-type water level gauge, a visual monitoring device and a server platform. The visual monitoring device is used to detect the data state of the buoy-type water level gauge to obtain water level information. The buoy-type water level gauge includes a hollow housing jointly composed of a stainless steel outer shell and a transparent plastic plate. A scale table is vertically arranged inside the hollow housing, and a card slot is vertically arranged outside the hollow housing. A float that can move up and down is arranged in the card slot. The visual detection device includes a fixed block, a high-definition camera, an infrared supplementary light, a camera pan-tilt, a 4G Internet access module, a controller and a power supply module. The camera pan-tilt, the 4G Internet access module and the controller are arranged inside the fixed block. The output end of the camera pan-tilt extends out of the fixed block to fix the high-definition camera. The infrared supplementary light is used to supplement light for the image of the detected buoy-type water level gauge. The controller is electrically connected to the camera pan-tilt, the 4G Internet access module, the high-definition camera, the infrared supplementary light and the power supply module respectively. The server platform is wirelessly communicatively connected to the 4G Internet access module.

[0005] Preferably, the float is a sealed "I-shaped" floating ball made of PVC plastic, and the float is brightly colored and easy to be recognized.

[0006] Preferably, a plurality of water holes are evenly arranged at the bottom and on both sides of the hollow shell, and the inside of the water holes is closed by a fine wire mesh.

[0007] Preferably, the length of each grid of the scale is 10CM*N, and each grid consists of numbers and 10 equally spaced black and white striped scales, where the vertical length of each striped scale is 1CM*N, and N is a non-zero natural number.

[0008] Preferably, the power supply module includes a power controller, a solar panel and a storage battery. The power controller and the storage battery are arranged in the fixed block, and the solar panel is arranged on the top of the fixed block. The solar panel, the storage battery, the power controller and the controller are electrically connected in sequence.

[0009] A vision-based floating water level monitoring method includes the following steps:

[0010] S1. The controller controls the high-definition camera to collect pictures of the floating water level gauge, and obtains the brightness of the floating water level gauge and the position information of the float.

[0011] S2. The controller judges whether to turn on the fill light according to the brightness of the floating water level gauge.

[0012] S3. The controller controls the camera pan-tilt to drive the high-definition camera to rotate according to the position information of the float, so that the float is at the center of the acquisition frame of the high-definition camera.

[0013] S4. The controller controls the high-definition camera to collect pictures of the floating water level gauge again, and transmits the collected pictures, the recognized Mask of the float, and the Mask information of the floating water level gauge to the server platform through the 4G Internet access module.

[0014] S5. The server platform receives the pictures and the position information of the recognized float on the scale, calculates the water level information through image recognition technology, and publishes it on the website for customers to view.

[0015] Preferably, in step S3, the controller obtains the mask Masks of the floating water gauge and the float from the pictures through a target detection algorithm, and then controls the camera pan-tilt to rotate.

[0016] Preferably, in step S5, the server uses an image segmentation algorithm to extract the mask Mask of the scale table image from the picture, and uses a deep learning algorithm to extract the numbers on the strip scale of the scale table and their masks Mask; calculates the average mask height Havg, and obtains the mask information of the scale with the smallest number through arrangement. The water level information is calculated by comparing the mask of the scale with the mask of the float; if the difference between the horizontal pixel position X at the upper left corner of the mask of the smallest number scale Mask and the horizontal pixel position X at the upper left corner of the mask of the float is less than the average mask height Havg, the mask of the smallest number is extracted, and the number of black and white scales in the scale is read through the gray histogram algorithm to calculate the water level information; if the difference between the two is greater than Havg, the part between the horizontal pixel position at the upper left corner of the float mask Mask and the horizontal pixel position at the upper left corner of the smallest scale mask Mask is extracted, and the number of black and white scales in the grid is read through the gray histogram algorithm to calculate the water level information.

[0017] According to the above technical solution, the beneficial effects of the present invention are:

[0018] 1. This application uses a floating water level gauge to monitor water level information. Compared with the existing scale immersed in water, it avoids large deviations in the collected water level information caused by surface reflection, and greatly improves the accuracy and precision of water level monitoring.

[0019] 2. This application adds an infrared fill light, which can well handle the situation where the picture is unclear due to poor visibility and low brightness, greatly improves the compatibility of the vision-based floating water level monitoring system, and can well adapt to various situations.

[0020] 3. This application works by controlling the pan-tilt camera to ensure that the float is at the center of the high-definition camera acquisition frame, providing a good basis for subsequent image recognition, and greatly improving the accuracy and stability of image recognition. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 is the usage state diagram of the present invention;

[0022] Figure 2 is the structural schematic diagram of the floating water level gauge of the present invention;

[0023] Figure 3 is the structural schematic diagram of the vision monitoring device of the present invention.

[0024] In the figure: 1. Floating water level gauge, 11. Scale table, 12. Float, 13. Water passing hole, 14. Hollow shell, 2. Vision monitoring device, 21. Solar panel, 22. Infrared fill light, 23. High-definition camera, 24. Fixed block. DETAILED DESCRIPTION OF THE INVENTION

[0025] The present invention will be further described in detail below in conjunction with the accompanying drawings and specific embodiments.

[0026] As Figures 1-3 shown, a visual-based floating buoy water level monitoring system and method. The system includes a floating buoy water level gauge 1, a visual monitoring device 2, and a server platform. The floating buoy water level gauge 1 includes a hollow housing 14 jointly composed of a stainless steel outer shell and a transparent plastic plate. A scale table 11 is vertically arranged inside the hollow housing 14. Each grid of the scale table 11 has a length of 10CM*N, and each grid consists of numbers and 10 equally spaced black and white striped scales. The vertical length of each striped scale is 1CM*N, where N is a non-zero natural number. The size of N is selected according to specific circumstances.

[0027] A card slot is vertically arranged outside the hollow housing 14, and a float 12 that can move up and down is arranged in the card slot. The float 12 is a sealed "I-shaped" floating ball made of PVC plastic, and the float 12 is brightly colored and easy to identify. A plurality of water holes 13 are evenly arranged at the bottom and on both sides of the hollow housing. The inside of the water holes 13 is closed with a fine wire mesh to prevent dirt from entering the inside of the water gauge and blocking the up and down movement of the float 12.

[0028] The visual detection device includes a fixed block 24, a high-definition camera 23, an infrared fill light 22, a camera pan-tilt, a 4G Internet access module, a controller, and a power supply module. The power supply module includes a power controller, a solar panel 21, and a storage battery. The camera pan-tilt, the 4G Internet access module, the power controller, the storage battery, and the controller are arranged inside the fixed block 24. The output end of the camera pan-tilt extends outside the fixed block 24 to fix the high-definition camera 23. The solar panel 21, the infrared fill light 22, and the power supply module are arranged on the top of the fixed block 24. The infrared fill light 22 is used to fill light for the image of the floating buoy water level gauge 1 being detected. The controller is electrically connected to the camera pan-tilt, the 4G Internet access module, the high-definition camera 23, the infrared fill light 22, and the power supply module respectively. The server platform is wirelessly communicatively connected to the 4G Internet access module. The solar panel 21, the storage battery, the power controller, and the controller are electrically connected in sequence.

[0029] During the specific working process of the system, the following steps are included:

[0030] S1. The controller controls the high-definition camera 23 to collect pictures of the floating buoy water level gauge 1, and obtains the brightness of the floating buoy water level gauge 1 and the position information of the float 12.

[0031] S2. The controller determines whether to turn on the fill light according to the brightness of the floating buoy water level gauge 1.

[0032] S3. The controller obtains the masks Mask of the float type water gauge 1 and the float 12 from the picture through the target detection algorithm, and then controls the camera pan-tilt to rotate so that the float 12 is at the center of the acquisition frame of the high-definition camera 23;

[0033] S4. The controller controls the high-definition camera 23 to collect the picture of the float type water level gauge 1 again, and transmits the collected picture, the recognized Mask of the float 12, and the Mask information of the float type water level gauge 1 to the server platform through the 4G Internet access module;

[0034] S5. The server platform receives the picture and the position information of the scale table 11 where the recognized float 12 is located, uses the image segmentation algorithm to extract the mask Mask of the scale table 11 image from the picture, and extracts the numbers and their masks Mask of the strip-shaped scales on the scale table 11 through the deep learning algorithm; calculates the average mask height Havg, and obtains the mask information of the scale with the smallest number through arrangement. The water level information is calculated by comparing the mask of the scale with the mask of the float 12; if the difference between the horizontal pixel position X at the upper left corner of the mask of the smallest number scale Mask and the horizontal pixel position X at the upper left corner of the mask of the float 12 is less than the average mask height Havg, the mask of the smallest number is extracted, and the number of black and white scales in the scale is read through the gray histogram algorithm to calculate the water level information; if the difference between the two is greater than Havg, the part between the horizontal pixel position at the upper left corner of the mask of the float 12 Mask and the horizontal pixel position at the upper left corner of the mask of the smallest scale Mask is extracted, and the number of black and white scales in the grid is read through the gray histogram algorithm to calculate the water level information.

[0035] It should be noted that the parts not described in detail in this application are all prior arts.

[0036] The above are only the preferred embodiments of the present invention, and do not impose any form of limitation on the present invention. Although the present invention has been disclosed above with the preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some changes or modifications within the scope of the technical solution of the present invention to make equivalent embodiments with equivalent changes, but as long as it does not deviate from the content of the technical solution of the present invention, any simple modification, equivalent change and modification made to the above embodiments according to the technical essence of the present invention should be covered by the protection scope of the present invention.

Claims

1. A vision-based float-type water level monitoring system, characterized in that: It includes a buoy-type water level gauge (1), a visual monitoring device (2), and a server platform. The visual monitoring device (2) is used to detect the data status of the buoy-type water level gauge (1) to obtain water level information. The buoy-type water level gauge (1) includes a hollow shell (14) jointly composed of a stainless steel outer shell and a transparent plastic plate. A scale table (11) is vertically arranged inside the hollow shell (14). A card slot is vertically arranged outside the hollow shell (14), and a float (12) that can move up and down is arranged in the card slot. The visual monitoring device includes a fixed block (24), a high-definition camera (23), an infrared fill light (22), a camera pan-tilt, a 4G Internet access module, a controller, and a power supply module. The camera pan-tilt, the 4G Internet access module, and the controller are arranged inside the fixed block (24). The output end of the camera pan-tilt extends outside the fixed block (24) to fix the high-definition camera (23). The infrared fill light (22) is used to fill light for the image of the buoy-type water level gauge (1) being detected. The controller is electrically connected to the camera pan-tilt, the 4G Internet access module, the high-definition camera (23), the infrared fill light (22), and the power supply module respectively. The server platform is wirelessly communicatively connected to the 4G Internet access module. Among them, the server uses an image segmentation algorithm to extract the mask Mask of the scale table (11) image from the picture, and uses a deep learning algorithm to extract the numbers on the strip-shaped scale of the scale table (11) and their masks Mask. Calculate the average mask height Havg, and obtain the mask information of the scale with the smallest number through arrangement. Calculate the water level information by comparing the mask of the scale with the mask of the float (12). If the difference between the horizontal pixel position of the upper left corner of the Mask of the scale with the smallest number and the horizontal pixel position of the upper left corner of the Mask of the float (12) is less than the average mask height Havg, then extract the mask of the smallest number, and read the number of black and white scales in the scale through the gray histogram algorithm to calculate the water level information. If the difference between the two is greater than Havg, then extract the part between the horizontal pixel position of the upper left corner of the Mask of the float (12) and the horizontal pixel position of the upper left corner of the Mask of the smallest scale mask, and read the number of black and white scales in the grid through the gray histogram algorithm to calculate the water level information.

2. The visual-based float-type water level monitoring system according to claim 1, characterized in that: The float (12) is a sealed "I-shaped" floating ball made of PVC plastic, and the float (12) is brightly colored and easy to be recognized.

3. The visual-based float-type water level monitoring system according to claim 1, characterized in that: A plurality of water holes (13) are evenly arranged at the bottom and on both sides of the hollow shell, and the inside of the water holes (13) is closed with a fine wire mesh.

4. The visual-based float-type water level monitoring system according to claim 1, characterized in that: The length of each grid of the scale table (11) is 10CM*N, and each grid is composed of a number and 10 equally spaced black and white strip-shaped scales. The vertical length of each strip-shaped scale is 1CM*N, and N is a non-zero natural number.

5. The visual-based float-type water level monitoring system according to claim 1, characterized in that: The power supply module includes a power controller, a solar panel (21), and a storage battery. The power controller and the storage battery are arranged inside the fixed block (24), and the solar panel (21) is arranged on the top of the fixed block (24). The solar panel (21), the storage battery, the power controller, and the controller are electrically connected in sequence.

6. A vision-based floating buoy water level monitoring method, which uses a vision-based floating buoy water level monitoring system according to any one of claims 1 to 5, characterized in that: It includes the following steps: S1. The controller controls the high-definition camera (23) to capture pictures of the buoy-type water level gauge (1), and obtains the brightness of the buoy-type water level gauge (1) and the position information of the float (12). S2. The controller determines whether to turn on the fill light according to the brightness of the buoy-type water level gauge (1). S3. The controller controls the camera pan-tilt to drive the high-definition camera (23) to rotate according to the position information of the float (12), so that the float (12) is at the center of the acquisition frame of the high-definition camera (23). S4. The controller controls the high-definition camera (23) to capture pictures of the buoy-type water level gauge (1) again, and transmits the captured pictures, the Mask of the identified float (12), and the Mask information of the buoy-type water level gauge (1) to the server platform through the 4G Internet access module. S5. The server platform receives the pictures and the position information of the scale table (11) where the identified float (12) is located, calculates the water level information through image recognition technology, and publishes it on the website for customers to view.

7. A vision-based float-type water level monitoring method according to claim 6, characterized in that: In step S3, the controller obtains the mask Masks of the buoy-type water level gauge (1) and the float (12) from the pictures through the target detection algorithm, and then controls the camera pan-tilt to rotate.

Images