A shipborne high-precision rainfall measurement device and method based on video image and attitude synchronization

Abstract

This invention relates to the field of marine meteorological observation equipment technology, specifically a shipborne high-precision rainfall measurement device and method based on video image and attitude synchronization. The device includes a rain collection head, a main body, a multi-tube array rain gauge, an optical observation module, a main control module, and a temperature and humidity sensor. The main body, rain collection head, and bottom plate form a Faraday cage. The rain gauge is equipped with a damping and liquid-sealing structure to suppress liquid surface oscillations, and a variable diameter design amplifies liquid level changes. The measurement method synchronizes ship data with the attitude sensor, triggers image acquisition at the alignment moment, extracts the water level line using an algorithm, and calculates the rainfall using a variable diameter formula. The temperature and humidity sensor is linked to achieve temperature compensation and snow melting. This invention integrates attitude calibration and video image technology to form a closed-loop measurement process, possesses electromagnetic shielding and full-scene adaptability, and can effectively improve the accuracy and stability of rainfall measurement in shipborne environments. It is suitable for various ship observation scenarios in nearshore and ocean-going environments.

Description

Technical Field

[0001] This invention relates to the field of marine meteorological observation equipment technology, specifically to a shipborne high-precision rainfall measurement device and method based on video image and attitude synchronization. Background Technology

[0002] Oceans cover more than 70% of the Earth's surface, and changes in marine meteorological information have a crucial impact on land. In particular, the increasingly frequent occurrence of severe natural disasters and extreme weather phenomena such as typhoons and storm surges in recent years has a vital impact on global shipping, marine aquaculture, and land-based disaster prevention and mitigation. Research and weather forecasting related to these natural disasters rely on accurate basic meteorological observation data. Marine rainfall measurement is a crucial part of marine meteorological information observation. While the measurement platforms for marine rainfall observation are largely the same as those used along the coast and on some islands, more representative and broader information sources are generally volunteer vessels or various anchored buoys at sea. These mobile platforms can measure information over a wider area, providing representative data from different regions for meteorological and climatological forecasting and scientific research, ensuring the accuracy and generalizability of forecasts and scientific theories.

[0003] Unlike stable observations on land, marine vessels and buoys, as mobile offshore platforms, experience pitching and rolling due to wave effects. Marine rainfall measurements must not only cope with these pitching and rolling effects but also withstand the following adverse influences when observing extreme weather phenomena: 1) Damage to sensors caused by induced lightning, strong lightning strikes, and high-energy electromagnetic pulses accompanying extreme or disastrous weather phenomena such as extreme cloud cover, heavy rain, and typhoons; 2) Fluctuations in rainfall water quality (such as salinity) caused by strong winds carrying seawater droplets; and variations in local rainfall salinity due to different regional climate conditions and human industrial and agricultural activities. Sensors need to achieve meteorological-level stable and accurate measurements under the combined influence of these factors, i.e., rainfall error ≤ 0.1 mm, and the measurement range must cover different ranges from light rain to heavy rain.

[0004] To address the aforementioned requirements, achieving meteorological-level rainfall measurement under swaying conditions at sea and maintaining high accuracy across a wide measurement range requires sensors to be able to withstand strong electromagnetic interference such as induced lightning and strong electromagnetic pulse impacts, and to be insensitive to changes in rainwater composition. These are the problems that rainfall sensors must solve in unstable marine environments.

[0005] Although existing patents CN112630866A (a rainfall measurement device) and CN201520290126 (a marine rainfall measurement device) indicate that sensors suitable for accurate rainfall measurement in the dynamic marine environment are rare, the capacitive rain gauges mentioned in these patents are meteorological-grade sensors specifically designed for marine use. Their water-vapor volume ratio measurement principle is unaffected by unstable sea surfaces, enabling stable, meteorological-grade rainfall measurement. Other sensors, such as tipping bucket and siphon rain gauges, load-bearing rain gauges (e.g., pressure method, weighing method), and liquid level sensors (e.g., ultrasonic ranging, infrared ranging), are either unsuitable due to their underlying principles or the numerous interference factors at sea, making them difficult to modify for engineering applications. The capacitive principle described in the aforementioned patent is applicable under normal marine weather conditions. However, when dealing with strong electromagnetic interference and changes in precipitation salinity, its principle becomes inherently limited: capacitive sensing requires the outer casing to be made of non-metallic material; otherwise, it is difficult to resolve the impact of parasitic capacitance of the metal casing on the measurement. Under strong electromagnetic interference at sea, without metal casing shielding, the internal components of the sensor will be directly damaged by environmental interference, affecting measurement accuracy and potentially causing permanent damage or even destruction to internal circuits and components. Furthermore, the capacitive principle is sensitive to the salinity of precipitation components. The salinity of marine precipitation varies in different regions, and existing patents do not mention how to correct for salinity, leading to doubts about the measurement results. Summary of the Invention

[0006] The purpose of this invention is to provide a shipborne high-precision rainfall measurement device and method based on video image and attitude synchronization, so as to solve the problems of low accuracy and poor adaptability of traditional shipborne rainfall measurement as mentioned in the background art.

[0007] To achieve the above objectives, the present invention provides the following technical solution: A shipborne high-precision rainfall measurement device based on video image and attitude synchronization includes a rain collection head, a main body, and a multi-tube array rain measuring tube; The rain collection head has a funnel-shaped structure, integrating a diversion port, a one-way valve, and a heating wire. The rain collection head is sealed to the main body. The multi-tube array rain gauge is located inside the main body and consists of multiple symmetrically arranged thin tubes connected at the bottom. The equivalent cross-sectional area of ​​the multi-tube array rain gauge... , with the cross-sectional area of ​​the rainwater collection head Satisfy the equivalent diameter ratio formula Actual rainfall With the liquid level height inside the rain gauge Satisfy conversion formula ; The main body is also equipped with an optical observation module opposite to the multi-tube array rain gauge, a main control module electrically connected to the optical observation module, a base plate for supporting the multi-tube array rain gauge, an electromagnetic / manual water discharge valve set on the bottom surface of the base plate, and a temperature and humidity sensor set on the inner side of the base plate. The base plate is provided with a positioning hole for setting the temperature and humidity sensor and a ventilation hole for ventilation.

[0008] Preferably, the multi-tube array rain gauge includes a centrally located observation main tube, a first water inlet tube and a second water inlet tube symmetrically distributed around the periphery of the observation main tube, and a spare observation tube with an integrated siphon structure. The spare observation tube is equipped with a siphon stop tube, which is located 2-3 cm above the liquid seal tube; the top of all the thin tubes is sealed and an air inlet is opened to balance the air pressure moving up and down on the liquid surface.

[0009] Preferably, the bottom of the multi-tube array rain gauge is connected to a liquid seal tube, the inner height of which is 1 cm, and the depth to which the main observation tube and the spare observation tube are inserted into the upper surface of the liquid seal tube is 0.7-0.8 cm. The bottom surface of the liquid seal tube is a funnel shape that slopes towards the center, and the inner wall edge at the connection with the water inlet pipe is provided with a 1mm symmetrical and uniform chamfer to form an air trap; the bottom of the thin tube of the multi-tube array rain gauge is provided with a damping structure or filled with a damping medium to form a coupled damping system to suppress liquid surface oscillation.

[0010] Preferably, the equivalent inner diameter of the multi-tube array rain gauge is... It is 4-5cm, which is the same as the inner diameter of the rain collection head. This results in an equivalent radius ratio of 1:4 to 1:5, corresponding to an equivalent variable diameter ratio. The ratio is 1:16 to 1:25; The thin tubes of the multi-tube array rain gauge are made of transparent hydrophobic material, and the actual inner diameter of a single thin tube is 2cm.

[0011] Furthermore, the thin tube of the multi-tube array rain gauge can be further divided into four smaller-diameter tubes, preferably four tubes with an inner diameter of 1 cm, arranged at equal intervals around the center on the same arc. This finer tube diameter physically suppresses the amplitude of liquid surface sloshing. Figure 7 As shown.

[0012] Preferably, the optical observation module is a wide-angle infrared camera with a field of view. Its field of view covers the entire height of the multi-tube array rain gauge from the bottom to the top of the effective observation point; The non-mirror area of ​​the infrared camera is protected by a metal casing, and the camera's pixel resolution is no less than 1920×1080.

[0013] Furthermore, the optical observation module is a wide-angle visible light camera, with other parameters and settings the same as the infrared camera, and visible light supplement lamps are added at appropriate positions on the inner wall and top of the main body (2) to achieve visible light image capture.

[0014] Preferably, the main control module integrates a 3-axis, 6-axis, or 9-axis pitch and roll attitude sensor based on MEMS principles, and supports RS485, RJ45, CAN, or I2C communication protocols to receive attitude data from external platforms. The main control module is fully enclosed in a metal casing for protection. Its built-in algorithm supports YOLOv5 / v8 image algorithm acceleration and is used to perform edge detection and water level line recognition on the acquired liquid level images.

[0015] Preferably, the heating wire inside the rain collection head has a circular structure for snow melting to adapt to measurement in all rain and snow scenarios; The electromagnetic / manual drain valve is connected to the liquid seal pipe through a through hole on the base plate and is used for draining accumulated water after the last rainfall, assisting drainage during heavy rain, and emergency manual drainage.

[0016] Preferably, the connection surfaces of the main body, the rain collection head, and the base plate are all provided with sealing rubber gaskets; the power supply end of the main control module integrates a lightning protection module, and the metal shell of the main control module (5), the metal shell of the optical observation module (4), and the Faraday cage structure of the main body together form a double electromagnetic shielding protection. The main body has a sealed dark cavity inside to isolate external natural light interference and ensure the measurement stability of the optical observation module. The main body is made of salt spray corrosion resistant metal material, and together with the rain head and the base plate, it forms a closed Faraday cage structure to shield external electromagnetic interference.

[0017] Preferably, two conductive electrodes are inserted into the top of the liquid seal tube, with the electrode ends inserted to a depth of half the height of the internal cavity of the liquid seal tube; when the liquid level rises to the electrode position, the measurement process is triggered; the main control module also supports receiving external forced measurement commands through the communication interface.

[0018] Furthermore, this invention can be simplified to directly measure the liquid level inside a large-volume cavity under swaying conditions, without requiring a diameter ratio amplification: the optical observation module (4), main control module (5), and observation main tube (O) are integrated into a sealed shell to form an independent dynamic liquid level measurement module. The observation main tube (O) is connected to the liquid cavity inside an external container via a connecting pipe. This module is suitable for installation on an external container via a mechanical interface to directly measure the liquid level inside the container, such as... Figure 8 As shown.

[0019] On the other hand, the present invention also provides a measurement method for a shipborne high-precision rainfall measurement device based on video image and attitude synchronization as described above, comprising the following steps: a. Calibration: Pre-calibrate the correspondence between the vertical coordinate of the image pixels and the actual liquid level height. Set the position where the liquid level is slightly higher than the upper side of the liquid seal pipe by 0.5mm as the zero point of the first rainfall level, and calibrate the rainfall amount corresponding to this position. ; b. Attitude acquisition: Acquire the pitch and roll data of the carrier through the attitude sensor integrated in the main control module or an external platform; c. Image acquisition: When the pitch and roll angles approach 0° and return to center, the optical observation module is triggered to capture liquid level images including the main observation tube and the backup observation tube; d. Liquid level calculation: The acquired image is cropped for the region of interest (ROI), converted to grayscale, binarized, and edge detected. The mean ordinate of multiple key points on the water level line is extracted, and then combined with the equivalent diameter ratio formula. Convert to actual rainfall; when the liquid level exceeds the zero point of the initial rainfall level, the cumulative rainfall is: ; e. Siphon circulation: When the liquid level rises to the siphon trigger height, the siphon structure starts to drain the liquid. The liquid level stops after it drops to the height of the siphon stop pipe. This height is used as the new reference zero point for subsequent measurements. f. Data Traceability and Redundancy Assurance: All original liquid level images, attitude data, and temperature and humidity data are retained as the basis for data traceability; ambient temperature is collected in real time through temperature and humidity sensor 8, based on the formula... Temperature compensation was applied to the converted rainfall, whereby... Rainfall after temperature compensation. For the initial calculation of rainfall, The actual ambient temperature is used, with 25℃ as the standard compensation benchmark, to correct for errors caused by water volume expansion or contraction due to temperature changes; when the attitude sensor fails, an image-based attitude substitution algorithm is activated to obtain the equivalent pitch and roll angles by fitting the water level line angles, ensuring measurement continuity. g. Drainage reset: If the liquid level does not change after more than 24 hours or 10 consecutive measurements, open the electromagnetic / manual drain valve to drain the accumulated water and wait for the next rainfall measurement.

[0020] Furthermore, the simplified liquid level measurement method of the present invention includes the following steps: a. Calibration: Pre-calibrate the correspondence between the vertical coordinates of image pixels and the actual liquid level height; b. Attitude acquisition: Acquire the pitch and roll data of the carrier through the attitude sensor integrated in the main control module or an external platform; c. Image acquisition: When the pitch and roll angles approach 0° and return to center, the optical observation module is triggered to capture an image of the liquid level in the observation main pipe; d. Liquid level calculation: The acquired image is cropped to a region of interest (ROI), converted to grayscale, binarized, and edge detected. The mean ordinate of multiple key points on the water level line is extracted to calculate the liquid level. Compared with the prior art, the beneficial effects of this invention are: 1. This invention features groundbreaking structural design innovation, effectively addressing the core pain points of traditional shipborne rainfall measurement devices, such as surface oscillation and insufficient measurement sensitivity. The core device, a multi-tube array rain gauge, employs a symmetrical layout and bottom-connected design. Combined with a damping structure at the bottom of the thin tube and a filling medium, it forms a coupled damping system that can quickly suppress surface disturbances caused by the ship's pitching and rolling. The dual structure of the liquid seal tube, combining air trap and liquid seal, avoids bubble interference and pressure imbalance, allowing the liquid surface to quickly stabilize under complex swaying conditions. Simultaneously, through the equivalent diameter ratio design of the rain collection head and the rain gauge, the formula K=(d / D)² is used to amplify the liquid level change of minute rainfall by a multiple of the square of the diameter ratio, significantly improving the measurement sensitivity for light and minute rainfall. Combined with the siphon circulation structure of the backup observation tube D, continuous rainfall measurement without interruption is achieved, requiring no manual intervention, greatly optimizing the rationality and stability of the measurement structure.

[0021] 2. This invention employs a measurement technology that simultaneously fuses video images and attitude data to construct a high-precision data processing system, overcoming the limitations of traditional measurement methods due to attitude interference. The optical observation module uses a wide-angle infrared camera, combined with the main body's sealed dark cavity design, to clearly capture liquid level images in low-light and low-temperature environments. Through YOLOv5 / v8 algorithms and Canny edge detection technology, the coordinates of key points on the water level line are accurately extracted. Furthermore, the optical observation module uses a visible light camera with supplementary lighting, also adapted to the main body's sealed dark cavity design. The MEMS attitude sensor integrated into the main control module collects real-time pitch and roll data of the hull. Through a mechanism that triggers image acquisition at the moment of alignment, combined with multi-frame image fitting and attitude substitution algorithms, the measurement error caused by attitude interference is controlled within ±0.1mm. Then, through temperature compensation formulas and the diameter conversion formula H=Kh, accurate calculation of rainfall is achieved, forming a closed-loop process of attitude calibration-image acquisition-algorithm processing. The measurement accuracy and data reliability far exceed those of traditional devices.

[0022] 3. This invention possesses strong adaptability to shipboard environments, balancing application in all scenarios with the need for long-term stable operation. The main body is preferably made of salt spray-resistant stainless steel, forming a Faraday cage structure with the rain collection head and bottom plate. Combined with a lightning protection module at the main control module's power supply end, the outer shell of the main body and the metal shell of the internal circuit components provide double electromagnetic shielding protection, resisting strong electromagnetic interference from the sea surface. The rain collection head's built-in heating wire, linked to a temperature and humidity sensor, enables snow melting in low-temperature environments, adapting to all rain and snow measurement scenarios. The dual design of the electromagnetic / manual water discharge valve for both automatic and emergency operation, along with the device's overall shock-absorbing mounting base and lightweight structure, allows it to adapt to different wave levels in nearshore and open ocean environments, while also meeting the needs of portable power supply and long-term observation on ships. Furthermore, the design of dual redundant data links and original image retention further ensures data continuity and traceability, adapting to diverse shipboard scenarios such as oceanographic research and nearshore patrol. Attached Figure Description

[0023] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are explained in detail together with the embodiments of the invention, but do not constitute a limitation thereof.

[0024] Figure 1 This is a vertical cross-sectional view of the present invention along the pipe BOC; Figure 2 This is a cross-sectional view of the present invention (AA section). Figure 3 This is a diagram showing the positional and structural relationships of the main observation pipe, backup observation pipe, siphon drainage pipe, liquid seal pipe, and siphon shut-off conduit of the present invention. Figure 4 This is a top view of the shape and position of the siphon shut-off catheter of the present invention; Figure 5 This is a schematic diagram showing the positions of the base plate, siphon drainage pipe, and exhaust hole of the present invention; Figure 6 This is a schematic diagram showing the installation position of the temperature and humidity sensor of the present invention on the base plate; Figure 7 This is a schematic diagram of a hybrid array in which the central tube variant is a local multi-tube micro-tube array, as shown in the AA cross-sectional view of the present invention. Figure 8 This is a schematic diagram of the simplest measuring unit of the present invention for assisting in the direct measurement of liquid level in other large tanks; The meanings of the labels in the diagram are as follows: 1. Rain collection head, 2. Main body, 3. Rain measuring tube, 4. Integrated camera, 5. Main control module, 6. Base plate, 7. Electromagnetic / manual drain valve, 1-1. Diverter port, 1-2. One-way valve, 1-3. Heating wire, θ. Camera pitch angle and field of view, 2-1. Low-level fixing plate, 2-2. High-level fixing plate, O. Main observation pipe, B. First water inlet pipe, C. Second water inlet pipe, D. Spare observation pipe, E. Siphon drain pipe, 3-1. Liquid seal pipe, 3-2. Siphon shut-off conduit, 6-1. Vent hole, 6-2. Positioning hole, 8. Temperature and humidity sensor. Detailed Implementation

[0025] The technical solutions of the present invention will be clearly and completely described below with reference to the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0026] The core concept of this invention lies in constructing a robust measurement system that isolates external natural light interference, physically suppresses liquid surface instability, utilizes optical capture, and employs electromagnetic immunity, rather than relying on traditional electronic sensing principles. This system, through a unique rain gauge mechanical structure, suppresses interference such as liquid level sloshing and flutter at the physical source, and uses intelligent strategies to capture evidential images at the optimal time. The sensor architecture employs a double-shielded metal shell, fundamentally overcoming the inherent defects of existing technologies. Simultaneously, the observation chamber is constructed as a closed structure to isolate natural light interference, providing a stable observation environment for optical observation. The specific invention is as follows: This invention discloses a shipborne high-precision rainfall measurement device and method based on video image and attitude synchronization, aiming to solve the problems of low measurement accuracy, poor adaptability, and insufficient data reliability caused by the influence of ship roll, electromagnetic interference, and sea surface temperature and humidity changes on traditional shipborne rainfall measurement devices. By integrating a multi-tube array rainfall measurement structure, video image observation and attitude synchronization calibration technology, and combining electromagnetic shielding and damping stabilization design, high-precision rainfall measurement can be achieved in all rain and snow scenarios on board. The specific implementation of the device structure and measurement method is described in detail below.

[0027] like Figures 1-8 As shown, the device of the present invention adopts a modular assembly design, including a rain head 1, a main body 2, and a multi-tube array rain gauge 3, which constitute the basic measurement unit. It is equipped with an optical observation module 4, a main control module 5, a base plate 6, an electromagnetic / manual water discharge valve 7, and a temperature and humidity sensor 8 to realize data acquisition, processing and equipment control. All components work together to meet the needs of complex shipboard environments.

[0028] The rain collection head 1 adopts a funnel-shaped integrated molding structure and is made of polypropylene material that is resistant to salt spray and aging. Its inner diameter D at the large end is set to 20cm, corresponding to a cross-sectional area S1=π(D / 2)²=π×10²=100πcm², which can efficiently collect rain or snow. The rain collection head 1 integrates a diversion port 1-1, a one-way valve 1-2, and a heating wire 1-3. Two diversion ports 1-1 are symmetrically arranged and evenly distributed on the lower part of the inner wall of the funnel. They are used to smoothly guide the collected rainwater and snow into different inlet pipes of the multi-tube array rain gauge 3, avoiding water flow impact that could cause liquid surface oscillation. The one-way valve 1-2 adopts a rubber sealing structure and is installed above the diversion port 1-1. It only allows water to flow from top to bottom into the rain gauge, while reducing water vapor evaporation from the rain gauge into the atmosphere outside the rain sensor, ensuring the accuracy of rainfall measurement data. The heating wire 1-3 is made of a ring-shaped nickel-chromium alloy with a diameter that matches the inner wall of the rain collection head 1. Its power is set to 50W. It is controlled by the temperature and humidity sensor 8 through the main control module 5. It automatically starts when the ambient temperature is below 0℃, melting the snow into liquid water for measurement, achieving full-scene adaptability to rain and snow.

[0029] The main body 2 is made of 304 stainless steel and treated with salt spray resistant coating, with a thickness of 5mm. Its top is sealed to the rain collection head 1 via a flange, and its bottom is bolted to the base plate 6. Fluororubber gaskets are installed on all connection surfaces. After assembly, it forms a closed cavity and constitutes a Faraday cage structure, shielding external electromagnetic interference with a shielding effectiveness of ≥40dB. This provides a stable working environment for the internal optical observation module 4, main control module 5, and multi-tube array rain gauge 3. The interior of the main body 2 is designed as a sealed dark cavity to isolate external natural light interference and prevent light reflection from affecting the liquid level recognition accuracy of the optical observation module 4.

[0030] The multi-tube array rain gauge 3 is the core measuring component of the device. It is made of transparent high borosilicate glass with a hydrophobic coating and a contact angle ≥110°. It consists of multiple thin tubes symmetrically arranged and connected at the bottom, and is fixed to the base plate 6. It includes a central observation tube O, a first water inlet tube B and a second water inlet tube C symmetrically distributed around it, and a spare observation tube D with an integrated siphon structure. The actual inner diameter of a single thin tube is 2cm. Through array combination, it forms an equivalent inner diameter d, with an equivalent cross-sectional area S2=π(d / 2)². It satisfies the equivalent diameter ratio formula K=(d / D)² with the cross-sectional area S1 of the rain collection head 1. The actual rainfall H and the liquid level height h in the rain gauge satisfy the conversion formula H=S2 / S1·h=Kh. The diameter variation design amplifies the liquid level change and improves the measurement sensitivity of micro-rainfall.

[0031] By using a multi-microtube array to create an equivalent diameter smaller than the rain-collecting head diameter, the equivalent diameter ratio K is defined as the ratio of the rain-collecting head diameter to the square of the second of the equivalent diameter of the multi-array. Equivalent to the area ratio of the rain collection tube to the rain measuring tube, it uses multiple small-diameter (relative to the large diameter of a traditional single-tube rain measuring tube) cylindrical bodies made of transparent, hydrophobic material, preferably PC material. The multiple thin tubes of the rain measuring tube are preferably of equal diameter, with an equivalent inner diameter d preferably of 4-5 cm, based on area conversion. R is the maximum inner diameter of the rain collection funnel. Let n be the area encompassing the inner diameter of the rain collection funnel, and n be the number of capillary tubes. Ideally, the capillary tubes should form an equivalent radius ratio of 1:4 to 1:5 and an equivalent diameter ratio of 1:16 to 1:25 with the rain collection head to achieve both rainfall amplification and error reduction effects. With a rain collection funnel inner diameter of 200mm, a diameter ratio of 25, and n=4, the equivalent inner diameter d of the multiple capillary tubes in the rain collection funnel is 4cm, and the actual inner diameter of each capillary tube is 2cm.

[0032] Let the cross-sectional area of ​​the rain head be... The equivalent cross-sectional area of ​​the combined multiple thin tubes of the rain gauge is The liquid level in the rain gauge is The actual rainfall was ,but: ; Amplification effect: when , hour, That is, the height of the liquid level inside the rain gauge. =2.5mm corresponds to the actual rainfall H=0.1mm, enabling the amplified measurement of minute rainfall; Error dilution: If the measurement error of the rain gauge level is... The actual rainfall error is... ,when At that time, the error was diluted to This significantly improves measurement accuracy.

[0033] The bottom of the multi-tube array rain gauge 3 is connected to a liquid seal tube 3-1, which is made of transparent acrylic material with an inner cavity height of 1cm. The main observation tube O and the spare observation tube D are inserted into the upper surface of the liquid seal tube 3-1 to a depth of 0.75cm, forming a liquid seal structure to prevent pressure imbalance inside the tube. The bottom surface of the liquid seal tube 3-1 is a funnel shape inclined towards the center at an angle of 15° to facilitate the drainage of accumulated water. The inner wall edge of the connection between the liquid seal tube 3-1 and the first water inlet tube B and the second water inlet tube C has a 1mm symmetrical and uniform chamfer to form an air trap structure, which avoids air bubbles generated by water flow impact from affecting the liquid level observation. At the same time, it facilitates the discharge of air when the water level in the liquid seal tube rises into the two water inlet tubes, reducing interference with the liquid level of the main observation tube O and the spare observation tube D. The bottom of the thin tube is equipped with a honeycomb damping structure filled with silicone damping medium to form a coupled damping system, which can effectively suppress the liquid surface oscillation caused by the pitching and rolling of the hull, so that the liquid surface quickly tends to stabilize and ensures the accuracy of observation.

[0034] The backup observation tube D integrates a U-shaped siphon structure and is equipped with a siphon stop conduit 3-2. The siphon stop conduit 3-2 is made of stainless steel, has a diameter of 8mm, and is located 2.5cm above the liquid seal tube 3-1. When the liquid level in the multi-tube array rain gauge 3 rises to the siphon trigger height (which is higher than the maximum height of the backup observation tube D), the siphon structure automatically drains the water. When the liquid level drops to the height of the siphon stop conduit 3-2, air enters the siphon structure, breaking the negative pressure and stopping the drainage. This height is used as the new zero point for subsequent measurements, enabling continuous rainfall measurement without manual intervention. All thin tubes are sealed at the top and have 2mm diameter air inlets to balance the air pressure inside the tubes during the up-and-down movement of the liquid surface, preventing pressure differences from affecting liquid level changes.

[0035] The optical observation module 4 employs a wide-angle infrared camera, mounted on the inner wall of the main body 2, and positioned opposite the multi-tube array rain gauge 3. It has a field of view θ = 165° and a pixel resolution of 2048 × 1536. Its field of view covers the entire height of the rain gauge from the bottom to the top of the effective observation tube, clearly capturing liquid level changes in the main observation tube O and the backup observation tube D. The non-mirror area of ​​the infrared camera is encapsulated and protected by a stainless steel shell, and the lens is equipped with a hydrophobic and dustproof film to prevent moisture adhesion from affecting image quality. The operating wavelength is set to 8-14 μm, enabling stable imaging in low-light and low-temperature environments, unaffected by external light interference.

[0036] The main control module 5 integrates a core control unit, attitude sensor, image processor, and communication module. It is fully encased in a metal shell and fixed to the top inner side of the main body 2. It is electrically connected to the optical observation module 4, temperature and humidity sensor 8, heating wires 1-3, and electromagnetic / manual drain valve 7 via wires. The attitude sensor is a 9-axis MEMS sensor supporting RS485 and CAN communication protocols. It can acquire real-time pitch and roll data of the carrier, with a measurement range of ±45° and an accuracy of ±0.1°. It can also receive attitude data from external ship platforms for synchronous attitude calibration. The image processor is equipped with a YOLOv8 algorithm acceleration chip, which performs ROI cropping, grayscale conversion, binarization, and edge detection on the liquid level images acquired by the optical observation module 4, quickly extracting the coordinates of key points on the water level line and calculating the actual liquid level height based on calibration parameters. The core control unit uses an STM32H7 series microcontroller with a processing speed ≥480MHz, enabling coordinated control, data processing, and storage of the various modules.

[0037] The base plate 6 is made of stainless steel with a thickness of 8mm. It supports the multi-tube array rain gauge 3, the electromagnetic / manual drain valve 7, and the temperature and humidity sensor 8. A through-hole is formed on its surface, connecting to the liquid seal pipe 3-1. The electromagnetic / manual drain valve 7 is installed below the through-hole, using a DC 12V electromagnetic drive with a rated flow rate ≥5L / min. It can automatically drain accumulated water after the last rainfall and also assist in drainage during heavy rain. A manual valve stem is also provided for emergency manual drainage. The temperature and humidity sensor 8 is a digital sensor with a measurement range of -40℃ to 85℃ and humidity of 0 to 100%RH, with accuracies of ±0.5℃ and ±2%RH respectively. It collects real-time temperature and humidity data inside the main body 2, providing a basis for the start / stop of the heating wires 1-3 and temperature compensation of the measurement data.

[0038] In addition, the power supply end of the main control module 5 integrates a lightning protection module (protection level ≥10kV). The metal shell of the main control module 5, the metal shell of the optical observation module 4, and the Faraday cage structure of the main body 2 together form a double electromagnetic shielding protection, which can adapt to the complex power supply environment of shipboard outdoor. The overall weight of the device is ≤15kg, and the bottom is designed with a shock-absorbing mounting base, which can be fixed to the ship deck or observation platform by bolts to reduce the impact of ship vibration on the measurement.

[0039] The measurement method based on the above-mentioned device is implemented through the following steps: a closed-loop process of calibration, attitude synchronization, image acquisition, liquid level calculation, siphon circulation, and data assurance is used to achieve high-precision rainfall measurement. Calibration: After the device is installed, pre-calibration is performed to establish the correspondence between the vertical coordinates of image pixels and the actual liquid level. Distilled water is injected into the multi-tube array rain gauge 3 so that the liquid level is slightly higher than the upper side of the liquid seal tube 3-1 by 0.5 mm. This position is set as the zero point of the liquid level for the first rainfall. The vertical coordinates of the pixels corresponding to the water level line in the image acquired by the optical observation module 4 at this time are recorded. Combined with the inner diameter of the rain gauge tube and the diameter ratio K, the rainfall H0 mm corresponding to this position is calibrated. H0 is the rainfall calculated by the sum of the volume of water in the liquid seal tube and the volume of water in the 0.5 mm high fine tube array. This calibration is permanently valid after one calibration. Subsequently, the water volume is gradually increased, and the pixel coordinates and actual rainfall corresponding to different liquid levels are recorded to establish a calibration database and correct image recognition errors. The calibration accuracy is ≤0.1 mm.

[0040] Attitude data acquisition: After the device is started, the main control module 5 controls the attitude sensor to collect the hull pitch and roll data in real time at a sampling frequency of 10Hz. At the same time, it receives attitude data from an external ship platform through the communication interface as a redundant backup. The two sets of data are subjected to mean filtering to remove abnormal noise and obtain the real-time attitude angles roll angle α and pitch angle β.

[0041] Image acquisition trigger: The main control module 5 judges the processed attitude angle. When the pitch and roll angles are close to 0° (error ≤ 0.5°) at the same time, the optical observation module 4 is triggered to capture liquid level images. The single capture time is 200ms, and 3 frames are continuously acquired as a group to avoid single imaging deviation. If the hull continues to sway (attitude angle deviation > 0.5°) for more than 5s, the image anti-shake algorithm is activated. Based on the multi-frame images, a stable water level line is fitted to ensure the effectiveness of the acquisition.

[0042] Liquid level calculation and rainfall conversion: The image processor preprocesses the three acquired images, first cropping the ROI region (keeping only the liquid level areas of the main observation pipe O and the backup observation pipe D), then performing grayscale and binarization processing (with adaptive threshold adjustment), extracting the water level contour using the Canny edge detection algorithm, calculating the mean of the pixel coordinates of 10 key points symmetrical to the left and right of the central axis of the observation pipe, and combining this with the calibration database to obtain the corresponding actual liquid level height h. Based on the equivalent diameter ratio formula H=Kh, the actual rainfall is calculated; when the liquid level exceeds the zero point of the first rainfall, the cumulative rainfall is... ,in Let n be the difference between the i-th measurement and the previous measurement, and n be the number of measurements. After the calculation is completed, the data will be stored in the local memory of the main control module 5.

[0043] Siphon circulation control: When continuous rainfall causes the liquid level in the multi-tube array rain gauge 3 to rise to the siphon trigger height (higher than the maximum height of the backup observation tube D), the siphon structure of the backup observation tube D is automatically activated to quickly drain the accumulated water; when the liquid level drops to the height of the siphon stop conduit 3-2 (2.5cm higher than the upper surface of the liquid seal tube 3-1), air enters the siphon structure to break the negative pressure, drainage stops, the main control module 5 sets this height as the new measurement reference zero point, updates the calibration parameters, and continues subsequent rainfall measurement to achieve continuous and uninterrupted monitoring of rainfall.

[0044] Data traceability and redundancy assurance: The main control module 5 retains all original liquid level images and corresponding attitude data and rainfall data, with a storage capacity of ≥16GB, supporting data export and traceability; when the attitude sensor fails, an image-based attitude replacement algorithm is automatically activated, which calculates the equivalent pitch and roll angle by fitting the angle between the water level lines in the main observation tube O and the backup observation tube D, correcting the liquid level measurement error and ensuring measurement continuity; the temperature and humidity sensor 8 collects data in real time and performs temperature compensation on the rainfall data, with the compensation formula being H comp =H×[1+0.002×(T-25)], where T is the actual temperature and 25℃ is the standard compensation reference), correcting the effect of temperature on water volume expansion.

[0045] Drainage Reset: When the device has no rainfall monitoring for 24 consecutive hours, the liquid level does not change, or the liquid level difference is ≤0.01mm after 10 consecutive measurements, the main control module 5 controls the electromagnetic drain valve 7 to open, draining the water accumulated in the multi-tube array rain gauge 3 and liquid seal pipe 3-1. After drainage is completed, the valve is closed, and the device returns to the initial standby state, waiting for the next rainfall to trigger the measurement. If the electromagnetic drain valve 7 fails, emergency drainage can be carried out by manually operating the valve stem.

[0046] To more clearly demonstrate the practical application effects and adaptability of the device and method of the present invention, and considering the different navigation environments in near-shore and open-ocean scenarios, two specific embodiments are provided below, corresponding to general-purpose and anti-sway application scenarios, respectively. Through targeted parameter settings, the adaptability of the present invention under different wind, wave, power supply and environmental interference conditions is demonstrated.

[0047] Example 1: Universal Measurement Device for Offshore Vessels and Its Application This embodiment is suitable for near-shore fishing vessels, patrol boats, and other near-shore operating vessels. These vessels navigate in areas with relatively small waves and have a roll / pitch angle of ≤15°. Therefore, the device requires high portability. Specific parameter settings are as follows: Rain collection head 1: inner diameter D=20cm, heating wire 1-3 power 50W, diversion port 1-1 quantity 3, one-way valve 1-2 is made of fluororubber; Multi-tube array rain gauge 3: equivalent inner diameter d=5cm, corresponding equivalent diameter ratio K=(5 / 20)²=1 / 16, consisting of one observation main tube O, one first water inlet tube B, one second water inlet tube C and one spare observation tube D. The bottom of the thin tube is equipped with a honeycomb damping structure and filled with silicone damping medium. The chamfer at the connection between the liquid seal tube 3-1 and the thin tube is strictly controlled to 1mm to ensure the air trap effect. Optical observation module 4: pixel resolution 1920×1080, field of view 160°, working wavelength 8-14μm; Main control module 5: Equipped with a 6-axis MEMS attitude sensor, supporting RS485 communication protocol, image processor adopts YOLOv5 algorithm, data storage capacity of 8GB, power supply end lightning protection module protection level adapted to near-shore environment is 8kV, overall power consumption is controlled within 15W, meeting the portable power supply needs of ships. Application results: Measurement range 0.1-200mm / h, measurement accuracy ±0.1mm, liquid surface oscillation suppression time ≤2s in nearshore wind and wave environment of level 3, can accurately capture rainfall changes from light rain to heavy rain, and data transmission delay ≤1s.

[0048] Example 2: Measurement device for strong sway resistance of ocean-going vessels and its application This embodiment is suitable for ocean-going freighters, research vessels, and other ocean-going vessels. These vessels navigate in areas with large waves and have a roll / pitch angle of ≤30°, requiring extremely high anti-interference and stability of the device. Specific parameter settings are as follows: Rain collection head 1: inner diameter D=25cm, heating wire 1-3 power 80W, diversion port 1-1 quantity 4, one-way valve 1-2 adopts metal sealing structure to enhance impact resistance; Multi-tube array rain gauge 3: equivalent inner diameter d=4cm, corresponding equivalent diameter ratio K=(4 / 25)²≈1 / 39.06. It adopts a redundant array consisting of 1 observation main tube O, 4 peripheral water inlet tubes and 2 symmetrically arranged spare observation tubes D. The bottom of the thin tube is filled with high viscosity silicone oil damping medium. The inner cavity height of the liquid seal tube 3-1 is increased to 1.2cm, the chamfer size is optimized to 1.2mm, and the adsorption capacity of the air trap structure for bubbles is significantly enhanced, further suppressing liquid surface disturbance. Optical observation module 4: Pixel resolution 2048×1536, field of view 170°, lens equipped with double-layer hydrophobic dustproof film, and added anti-vibration bracket; Main control module 5: Equipped with a 9-axis MEMS attitude sensor, supporting redundant backup of RS485 and CAN dual communication protocols, the image processor is equipped with a YOLOv8 algorithm acceleration chip, the data storage capacity is 32GB, the integrated GPS positioning module realizes real-time correlation between rainfall and location information, and the power supply end lightning protection module protection level is upgraded to 12kV, which is suitable for strong electromagnetic interference environment in the ocean. Application results: Measurement range 0.05-300mm / h, measurement accuracy ±0.1mm, in the ocean with level 5 wind and waves, liquid surface oscillation suppression time ≤1.5s, attitude calibration error ≤0.05°, can adapt to low temperature (-20℃), high humidity (95%RH) and strong electromagnetic interference environment, and continuous working stability ≥720 hours.

[0049] The two sets of embodiments, through the above-described structural design and measurement methods, possess the following core advantages compared to traditional shipborne rainfall measurement devices: First, the integrated design of the multi-tube array rain gauge 3, combined with the damping system and liquid seal pipe 3-1, effectively suppresses liquid surface oscillations caused by hull rolling. Combined with the precise adaptation of the variable diameter amplification formula K=(d / D)², it amplifies the liquid level changes of minute rainfall by 16-39 times, significantly improving measurement sensitivity. Second, the attitude synchronization calibration mechanism of the optical observation module 4 and the main control module 5 triggers image acquisition at the moment the hull returns to center. The algorithm correction controls the measurement error caused by attitude interference to within ±0.1mm; thirdly, the design of the Faraday cage structure of the main body 2, the built-in heating wires 1-3 of the rain collection head 1, and the dual redundant data link of the main control module 5 are fully adaptable to complex environments such as shipborne salt spray, low temperature, and strong electromagnetic fields, greatly improving the adaptability and data reliability of the device; fourthly, the siphon circulation of the backup observation tube D and the automatic drainage function of the electromagnetic / manual water discharge valve 7 are linked to achieve continuous rainfall measurement without interruption, without manual intervention, which can meet the needs of long-term shipborne observation scenarios such as oceanographic research.

[0050] The advantages of the shipborne high-precision rainfall measurement device and method based on video image and attitude synchronization proposed in this invention are as follows: This invention features groundbreaking structural design innovations, effectively addressing the core pain points of traditional shipborne rainfall measurement devices, such as surface oscillation and insufficient measurement sensitivity. The core device, a multi-tube array rain gauge 3, employs a symmetrical layout and bottom-connected design. Combined with a damping structure at the bottom of the thin tube and a filling medium, it forms a coupled damping system that quickly suppresses surface disturbances caused by the ship's pitching and rolling. The dual structure of the liquid seal tube 3-1, combining air trap and liquid seal, prevents bubble interference and pressure imbalance, allowing the surface to quickly stabilize under complex swaying conditions. Simultaneously, through the equivalent diameter ratio design of the rain collection head 1 and the rain gauge, the formula K=(d / D)² amplifies the liquid level changes of minute rainfall (typically 16-39 times), significantly improving the measurement sensitivity for light and minute rainfall. Combined with the siphon circulation structure of the backup observation tube D, continuous rainfall measurement without interruption is achieved, requiring no manual intervention, greatly optimizing the rationality and stability of the measurement structure.

[0051] This invention employs a measurement technology that simultaneously fuses video images and attitude data to construct a high-precision data processing system, overcoming the limitations of traditional measurement methods due to attitude interference. The optical observation module 4 uses a wide-angle infrared camera, combined with the sealed dark cavity design of the main body 2, to clearly capture liquid level images in low-light and low-temperature environments. Through the YOLOv5 / v8 algorithm and Canny edge detection technology, the coordinates of key points on the water level line are accurately extracted. The main control module 5 integrates a MEMS attitude sensor that collects real-time pitch and roll data of the hull. By triggering image acquisition at the moment of alignment, and combining multi-frame image fitting and attitude substitution algorithms, the measurement error caused by attitude interference is controlled within ±0.1mm. Then, through temperature compensation formulas and the diameter conversion formula H=Kh, accurate calculation of rainfall is achieved, forming a closed-loop process of attitude calibration-image acquisition-algorithm processing. The measurement accuracy and data reliability far exceed those of traditional devices.

[0052] This invention possesses strong adaptability to shipboard environments, balancing application in all scenarios with the need for long-term stable operation. The main body 2 is made of salt-spray resistant stainless steel, forming a Faraday cage structure with the rain collection head 1 and the base plate 6. Combined with the metal protective shell of the main control module 5 and the optical observation module 4, it forms dual electromagnetic shielding protection. Coupled with the lightning protection module at the power supply end, it can resist strong electromagnetic interference from the sea surface. The rain collection head 1 has heating wires 1-3 linked to temperature and humidity sensors 8, enabling snow melting in low-temperature environments, adapting to all rain and snow measurement scenarios. The electromagnetic / manual water discharge valve 7 features both automatic and emergency design, and the device's overall shock-absorbing mounting base and lightweight structure allow it to adapt to different wave levels in near-shore and open-ocean environments, while also meeting the needs of portable power supply and long-term observation on ships. Simultaneously, the dual redundant data link and original image retention design further ensure data continuity and traceability, adapting to diverse shipboard scenarios such as oceanographic research and near-shore patrol.

[0053] To further verify the measurement accuracy, adaptability, and other core technical effects of the technical solution of the present invention, the following calculation examples serve as supplementary verification materials for the technical effects of the present invention, and are used to prove the accuracy and adaptability of the technical solution described in the claims.

[0054] Calculation Example 1 – Setting Calculation Parameters The calculation sets the inner diameter of the rain collection head to 200mm, and the rain gauge array uses a multi-microtube array with an equivalent inner diameter of 40mm, preferably four tubes with a diameter of 20mm; the inner cavity of the liquid seal tube at the bottom of the rain gauge is 10mm high, and the inner diameter of the tube is 100mm. The main observation tube is 400mm higher than the backup observation tube. The backup observation tube is equipped with a siphon drainage pipe, and the siphon position is set at approximately 300mm high so that the siphon will be triggered for natural drainage when the water level reaches this position.

[0055] The equivalent diameter ratio of the rain collection head to the rain measuring tube array is K=1 / 25.

[0056] All the following calculations are based on the above data.

[0057] Calculation Example 2 – Comparison of Measurement Error Capability This calculation example uses an inner diameter of 2cm for the observation pipe and compares the limit measurement error of the liquid level at the time of return to center under the conditions of ±15° and ±20° longitudinal and transverse rocking. It also compares the limit measurement error of different rainfall amounts and rainfall intensities.

[0058] The basic premise of analysis and comparison: 1) Under ±15° / ±20° longitudinal and transverse swaying conditions, using the theoretical maximum difference in liquid surface tilt height as the absolute deviation limit for image measurement, even with random fluctuations in the liquid surface tilt at the time of straightening, the deviation will not exceed this theoretical value: the formula for the maximum difference in liquid surface tilt height is... (r is the radius of the thinner tube, (This refers to the sway angle), which is the geometric limit. When the sway angle is at its maximum, the tilt of the liquid surface reaches its peak. The process of returning to the normal position is "resetting from the peak to the reference." Random fluctuations will only be between 0 and the peak value, and it is impossible for them to exceed the theoretical maximum value.

[0059] 2) The liquid level is measured by image. The height of the liquid level line is calculated by taking the image at the time of alignment. The average value of the water level line image points is taken symmetrically on both sides of the center line of the rain gauge. Even if the water level line is not flat, theoretically, the deviation of the left and right sides can be used to take the average value of the vertical coordinate of the pixel points of the symmetrical measurement points. The deviation of the two sides of the water level line can be canceled out to a certain extent. The water level height calculated after taking the average value of multiple points of the water level line should theoretically be close to the true value of the water level line in a stable state. Even if there is a measurement error in the liquid level line due to image recognition, the 25-fold diameter ratio can greatly suppress the error. After taking the average value of the water level line, the error will be further compressed by the diameter ratio, and the error will be smaller.

[0060] 3) Calculate the theoretical measurement error of a 2cm diameter thin tube under different rainfall intensities. The data comparison is sufficient to support the rigor of the design.

[0061] 4) This calculation example is based on the maturity of image measurement technology: Image measurement of water levels can achieve sub-pixel accuracy, and after calibration, the actual water level measurement accuracy can reach below millimeter level. This patent further ensures this accuracy through three key designs: ① The narrow tube diameter physically constrains the liquid surface, significantly reducing geometric deformation and providing superior conditions for image measurement compared to large-diameter cylinders / natural water bodies; ② A sealed dark cavity is constructed to isolate external light and weather interference, ensuring a stable measurement environment; ③ A 25-fold diameter ratio suppresses measurement errors. Combined with the first two advantages, this patent can achieve rainfall measurement accuracy far exceeding existing technologies. Therefore, this calculation example is analyzed based on extreme geometric deviations, and the actual measurement results will be far better than the theoretical calculation values.

[0062] Based on the above settings and analysis, this calculation example is analyzed according to the maximum geometric error at the more extreme maximum longitudinal and transverse sway positions. That is, the theoretical maximum liquid surface tilt height difference is set as the measured liquid level measurement error of the rain gauge. The actual results are much better than this result when the analysis and comparison are based on this maximum geometric error.

[0063] According to the geometric deviation formula, calculate the maximum geometric deviation on both sides of the water level line under ±15° / ±20° longitudinal and transverse swaying respectively: Δh=r×sinθ, where r is the radius of the capillary tube and θ is the maximum swaying angle of longitudinal and transverse swaying; sin ≈0.3420. Table 1 compares the maximum liquid level tilt height difference at the theoretical maximum longitudinal and transverse sway angle positions of thin tubes with different inner diameters. Table 2 compares the rainfall intensity measurement deviation of thin tubes with an inner diameter of 2cm after being compressed by a 25-fold diameter ratio under different rainfall intensity levels. The actual situation will be much better than the results in the table.

[0064] Table 1. Theoretical maximum liquid level tilt height difference under ±15° / ±20° oscillation of capillary tubes with different inner diameters. Inner diameter of the thin tube (cm) ±15° tilt height difference Δh (mm) ±20° tilt height difference Δh (mm) 1 0.5×0.2588×10≈1.3 0.5×0.3420×10≈1.7 2 1.0×0.2588×10≈2.6 1.0×0.3420×10≈3.4 3 1.5×0.2588×10≈3.9 1.5×0.3420×10≈5.1 4 2.0×0.2588×10≈5.2 2.0×0.3420×10≈6.8 5 2.5×0.2588×10≈6.5 2.5×0.3420×10≈8.6 8 4.0×0.2588×10≈10.4 4.0×0.3420×10≈13.7 Table 2 shows the maximum error of a 2cm inner diameter thin tube under different rainfall intensities (after compression with a 25x diameter ratio). Rainfall Intensity Level ±15° original maximum deviation (mm) Measurement error (mm) after pressing with a 25x diameter ratio. ±20° original maximum deviation (mm) Measurement error (mm) after pressing with a 25x diameter ratio. Light rain 2.6 2.6÷25≈0.104 3.4 3.4÷25≈0.136 Moderate rain 2.6 0.104 3.4 0.136 heavy rain 2.6 0.104 3.4 0.136 rainstorm 2.6 0.104 3.4 0.136 In Table 2, based on a 2cm narrow tube diameter, under ±15° longitudinal and transverse swaying conditions, using the maximum geometric deviation of the liquid level as the measurement error, the measurement error after a 25-fold diameter change ratio basically reaches the meteorological level measurement error capability. However, under ±20° longitudinal and transverse swaying conditions, after a 25-fold diameter change ratio, the error slightly exceeds the meteorological level measurement error accuracy by 0.1mm. But at the instant the narrow tube diameter returns to the centered position, even with the added instability of the liquid level, its maximum theoretical geometric error will inevitably be much smaller than this maximum geometric error. Furthermore, by taking data points symmetrically on both sides of the water level line and calculating the average, the actual measurement error of the water level line inside the tube will be much lower than the maximum geometric deviation.

[0065] Even if we simply take the average deviation from the water level line to offset it, assuming it can offset half, the accuracy is better than under the two extreme pitch and roll angles, with its maximum error... The values ​​are 0.052mm and 0.068mm respectively, which already meet the meteorological level measurement error requirement of 0.1mm.

[0066] Different rainfall intensities do not affect geometric deviations. Essentially, the tilt of the liquid surface is determined by the device's sway angle, which is independent of the rate of liquid level rise. Therefore, the error after the variable diameter ratio compression remains constant under all rainfall intensities.

[0067] Regarding the different rainfall errors, although the rainfall amount is obtained by accumulating the rainfall intensity at different times, the total rainfall error is independent of the duration. This is because each rainfall intensity measurement is based on the instantaneous maximum liquid surface tilt deviation, which is a random deviation. The direction of the liquid surface swaying at the moment of returning to center is irregular, with deviations that are sometimes positive and sometimes negative. It is not a systematic deviation (it will not be unidirectionally high or low). The longer the time, the more fully the positive and negative deviations cancel each other out, and the total error tends to stabilize. The actual rainfall error, accumulated over time, will be lower than the aforementioned maximum geometric deviation. .

[0068] Calculation Example 3 – Measurement Accuracy of Replacing the Central Thin Tube with an Even Thinner Tube to Form a Hybrid Array This calculation example further optimizes the original array of four 20mm measuring tubes into a hybrid split structure of "three 20mm triangularly distributed outer tubes + four 10mm square arrays in the center," while maintaining the total cross-sectional area and ensuring the diameter ratio remains at 25: the three outer 20mm tubes retain their original triangular mounting layout, while the four 10mm tubes in the center are embedded in a 2×2 square array, replacing the original central 20mm tube. The centers of the four 10mm tubes are located on the inner diameter of the original 20mm tubes, and they are symmetrically distributed at equal intervals. All other structural elements remain unchanged. The vertical view of the sensor cross-section is shown below. Figure 7 .

[0069] Calculations show that after mixing and splitting, the total cross-sectional area of ​​the measuring tube remains 400π mm², maintaining a 25-fold diameter ratio with the cross-sectional area of ​​the rain head (200 mm inner diameter funnel), thus ensuring unchanged error suppression. Furthermore, the compact array of the four central 10 mm thin tubes further constrains the geometric deformation of the liquid surface, improving the accuracy and stability of the liquid level measurement along the central axis of the thin tubes. All measuring tubes are interconnected at their bottoms, forming a complete communicating vessel, ensuring level liquid levels.

[0070] In addition, one of the four central thin tubes can be selected as the main measuring tube, with the remaining three as spares, to further improve accuracy. At this point, at least one thinner main measuring tube, one spare thinner measuring tube, and the original spare D measuring tube (integrated siphon tube) can be obtained.

[0071] Referring to the tilt height difference corresponding to the 1cm inner diameter of the capillary tube in the first row of Table 1 in Example 2, even if this height difference is taken as the measurement error, and the average is taken symmetrically without passing through the water level line image points, after a 25-fold diameter change ratio, the rainfall measurement errors of ±15° and ±20° are 0.052mm and 0.068mm respectively, which directly meet the meteorological level rainfall measurement accuracy requirements. Furthermore, considering the physical suppression effect of the capillary tube diameter on the change of liquid level tilt posture at the actual return moment, and the further elimination of measurement error by taking the average of the liquid level line along the center line, the measurement accuracy capability of this invention for rainfall will far exceed the meteorological level measurement accuracy requirements.

[0072] Calculation Example 4 – Design of Connecting Cavity Structure and Analysis of Rainfall Accumulation Time The multi-tube measurement array of this invention, combined with a bottom liquid-sealed connecting cavity structure, achieves accurate rainfall measurement with a 25-fold diameter ratio, while also considering structural compactness and fluctuation suppression. The specific design and theoretical derivation are as follows: 1. Definition of core structural parameters Rain collection head: inner diameter D=200mm, cross-sectional area ; Rain gauge array: 4 tubes with an inner diameter d = 20 mm (2 cm), and a single tube cross-sectional area of... Total measured cross-sectional area ; Variable diameter ratio: K=25 times, which meets the requirement of significantly reducing rainfall error; Liquid seal tube connecting cavity: adopts a cylindrical structure, with a height H=10mm (1cm), and the inner diameter is adapted to the outer diameter of the measuring tube array, which is set to 100mm. Volume Four measuring tubes are vertically inserted into the upper surface of the connecting cavity, and the bottom is fully connected to the inside of the cavity.

[0073] 2. Theoretical Derivation Logic of Rainfall Accumulation Time The time it takes for rainwater to accumulate in the connecting cavity is essentially the time it takes for the rainwater collected at the rain inlet to fill the cavity. The derivation formula is based on the principle of fluid volume conservation. , in: Time taken for the connecting cavity to fill with water (unit: s); For the volume of the connecting cavity (78540) (fixed value) The inflow rate of the rainwater collection head per unit time (unit: / min), which is calculated by multiplying the rainfall intensity I (unit: mm / min) by the area of ​​the catchment inlet, i.e. .

[0074] Combining the common meteorological concept of rainfall intensity, and substituting the following minute-based rainfall intensity classifications (as an engineering reference classification for this invention) into the above formula, the theoretical time required for the connected cavity to fill with water under different rainfall intensities is derived, and the results are shown in the table below: Rainfall Intensity Level Rainfall intensity range (mm / min) Rainwater inlet flow rate per unit time (mm³ / min) Theoretical time required for the connecting cavity to be filled with water Theoretical Impact Analysis Light rain (no rain turning into light rain) 0~0.01 10000π×(0~0.01)≈0~314 25min~+∞ There is only a theoretical delay in the first rainfall event, with no measurement error. Light rain 0.01~0.2 10000π×(0.01~0.2)≈314~6280 1.2min~25min The initial rainfall was delayed, but it quickly entered a stable measurement phase. Moderate rain 0.2~0.5 10000π×(0.2~0.5)≈6280~15700 0.5min~1.2min Negligible delay heavy rain 0.5~1.0 10000π×(0.5~1.0)≈15700~31400 15s~30s No substantial impact rainstorm 1.0~2.0 10000π×(1.0~2.0)≈31400~62800 7.5s~15s No impact Heavy rain ≥2.0 10000π×2.0≥62800 ≤7.5s No impact 3. Explanation of the rationality of the theoretical derivation results All data in the table above are derived based on the structural parameters disclosed in this calculation example, the common definition of rainfall intensity in meteorology, and the principle of volume conservation in fluid mechanics. The derivation process can be repeatedly verified.

[0075] It should be noted that the data in the table are theoretical reference values ​​under ideal working conditions. In actual applications, they may be affected by factors such as ambient temperature (affecting rainwater viscosity), installation verticality (affecting rainwater flow path), and medium purity (affecting flow resistance), resulting in a theoretical fluctuation range. However, this fluctuation range will not change the core conclusion that the time consumed by the connecting cavity does not affect the measurement accuracy. This is because the connecting cavity only produces a "display delay" rather than a measurement error. The four measuring tubes and the connecting cavity always satisfy the principle of communicating vessels, and the water level changes synchronously. After zero-point calibration, the water level change in the measuring tube has a strict linear correspondence with the actual rainfall, and the error compression effect of the 25-fold diameter ratio is not affected in any way.

[0076] For rainfall intensities of moderate to heavy, the initial measurement may directly yield a result indicating that the liquid level inside the rain gauge is higher than the "zero point of the initial rainfall level." Theoretically, this result contains no measurement error: that is, the initial measurement already includes the rainfall amount corresponding to the zero point of the initial rainfall level. And the incremental height of the measured liquid level relative to the zero point of the first rainfall. That is, a single measurement yields the result. The actual rainfall was measured, and subsequent measurements were continuously taken based on this to measure any additional rainfall.

[0077] Example 5 – Calculation of Rainfall Measurement Duration This calculation example verifies the results for two questions: 1) Measurement duration of a single full-scale measurement of an effective liquid level of 300mm (total rain intensity); 2) The logic of repeated measurements without delay after the first rainfall, and the data are all based on a 25-fold diameter ratio (5 times the diameter ratio), a rain head funnel inner diameter of 200mm, and an observation array of 4 2cm thin tubes.

[0078] I. Core Parameters 1. Measuring side: 4 2cm tubes effectively measure liquid level 300mm, total measuring volume = 4×π×10²×300 = 120000π≈376991mm³.

[0079] 2. Rainwater collection port side: area 10000πmm², 25 times the diameter ratio → for every 1mm of rainwater collected by the rainwater collection port, the liquid level in the measuring tube rises by 25mm.

[0080] 3. Drainage trigger: The siphon tube height is set to 300mm. When the liquid level in the measuring tube reaches 300mm, the siphon will start to drain, draining to the liquid level near the siphon stop tube position. Subsequent measurements will be continuously cycled around this position.

[0081] 4. Connecting cavity: The 1cm height only delays the initial water storage, and there is always water in the cavity during subsequent rainfall, with no secondary delay, which does not affect the cyclic measurement.

[0082] II. Measurement duration of a single full-scale test for an effective liquid level of 300mm under different rainfall intensities. Single measurement duration = Measurement volume corresponding to 300mm liquid level ÷ Water inflow rate per unit time at the rainwater collection inlet This value directly reflects how long the next full-scale measurement can be performed for each rainfall intensity.

[0083] Rainfall Intensity Level Rainfall intensity range (mm / min) Rainwater inlet flow rate per unit time (mm³ / min) Theoretical measurement time for a single full-scale measurement of a 300mm liquid level. Measurement Cycle Instructions (for repeated measurements) light rain 0~0.01 0~314 ≈377min (6.3h) ~∞ Extremely slow water accumulation, almost no siphoning triggered, continuous metering Light rain 0.01~0.2 314~6280 18.8min~377min A full cycle occurs every 18 minutes to 6 hours; there is no delay in circulation after siphon drainage. Moderate rain 0.2~0.5 6280~15700 7.5min~18.8min It takes 7-19 minutes to fill once, and there is no delay in circulation after siphon drainage. heavy rain 0.5~1.0 15700~31400 3.8min~7.5min It fills every 4-8 minutes. During siphon drainage, the solenoid valve can be activated to assist and double the drainage speed. rainstorm 1.0~2.0 31400~62800 1.9min~3.8min It fills every 2-4 minutes. During siphon drainage, the solenoid valve can be activated to assist and double the drainage speed. The single flow rate covers the cumulative flow of heavy rain for 1 minute. Heavy rain ≥2.0 ≥62800 ≤1.9min (114 seconds) The flow rate is ≤1.9 minutes per full cycle. During siphon drainage, the solenoid valve can be activated to assist and double the drainage speed. A single flow rate can cover the cumulative flow of heavy rain within 1 minute. III. Delay and Loop Logic of Repeated Measurements 1. First rainfall: First, fill the liquid seal tube connecting cavity to a height of 1cm (time taken is shown in the table above). Then, the liquid level in the measuring tube rises from 0 to 300mm to complete the first full-range measurement. When the siphon drains for the first time, only the water level of 300mm in the measuring tube is drained. Water is always retained at a height of 1cm in the connecting cavity and will not be emptied. 2. Subsequent repeated measurements: Because there is always water in the connecting cavity, there is no secondary water storage delay, and rainfall directly enters the liquid level accumulation stage of the measuring tube; when the liquid level rises to 300mm, the siphon is triggered, and after the liquid level is discharged to the preset siphon stop safety level, the siphon stops, and the liquid level accumulates again, thus repeating the cycle without any delay.

[0084] 3. Cumulative measurement capability: The actual rainfall corresponding to a liquid level of 300mm is 300mm × 25 = 750mm (the rainfall corresponding to a single full-scale measurement), which far exceeds the cumulative rainfall of a single rainstorm. Cyclic measurement can cover rainfall of any duration.

[0085] It should be noted that the technical solution of the present invention can be flexibly adjusted based on the limitations of the claims: for example, the central observation tube of the multi-tube array rain gauge can be replaced with multiple thinner tubes (i.e., a hybrid array configuration), as long as the bottoms of all the thin tubes are connected and the total cross-sectional area satisfies the equivalent diameter ratio formula K=(d / D)², that is, the liquid surface distortion can be further suppressed by the thinner central tube, and its assembly method is consistent with the multi-tube array of this embodiment; as another example, the optical observation module, the main control module and the observation tube can be integrated into an independent dynamic liquid level measurement module, which only needs to be connected to the external container through a connecting tube, and its measurement logic (attitude synchronization + alignment time acquisition + image calculation) is the same as the rainfall measurement method of this embodiment, without the need for additional adjustments to the core structure. The above variations are all conventional adjustments made by those skilled in the art based on this embodiment, and can be implemented without creative effort. The basic principles, main features and advantages of the present invention have been shown and described above. Those skilled in the art should understand that this invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely preferred examples of this invention and are not intended to limit the invention. Various changes and improvements can be made to this invention without departing from its spirit and scope, including but not limited to simplifying the multi-tube array into a single-tube rain gauge when used in a land environment, replacing the central tube of the fine tube array with a hybrid array configuration of a local array + peripheral array with a smaller inner diameter, and the simplest structure of a single fine tube + optical observation module + main control module (connecting to the inner cavity of other liquid storage devices to directly measure the liquid level). All of the above changes and improvements must meet the core technical features of "physical suppression of liquid surface distortion + optical image recognition of liquid level". For dynamic environments (such as shipborne, vehicle-borne, and buoy-borne), a "attitude adaptation acquisition strategy" (including attitude synchronous return acquisition, image-based attitude replacement, etc.) is required. For static environments (such as fixed land or stationary containers), the attitude adaptation process can be simplified. Any modifications that only change the structural layout, quantity, or application scenario but do not depart from the core technical concept of this invention fall within the scope of this invention as claimed. The scope of protection of this invention is defined by the appended claims and their equivalents.

Claims

1. A shipborne high-precision rainfall measurement device based on video image and attitude synchronization, characterized in that, It includes a rain collection head (1), a main body (2), and a multi-tube array rain gauge (3); The rain collection head (1) has a funnel-shaped structure and is used to collect rainfall and guide it to the multi-tube array rain gauge (3). The rain collection head (1) is sealed to the main body (2). The multi-tube array rain gauge (3) is located inside the main body (2) and is composed of multiple thin tubes arranged symmetrically and connected at the bottom. The equivalent cross-sectional area of ​​the multi-tube array rain gauge (3) is... , and the cross-sectional area of ​​the rain collection head (1) Satisfy the equivalent diameter ratio formula Actual rainfall With the liquid level height inside the rain gauge Satisfy conversion formula Equivalent diameter ratio It can achieve the dual functions of amplifying rainfall and suppressing errors; The main body (2) is also equipped with an optical observation module (4) that is opposite to the multi-tube array rain gauge (3), a main control module (5) that is electrically connected to the optical observation module (4), and a base plate (6) for supporting the multi-tube array rain gauge (3).

2. The shipborne high-precision rainfall measurement device based on video-synchronized mode as described in claim 1, characterized in that, The multi-tube array rain gauge (3) includes a central observation main tube (O), a first water inlet tube (B) and a second water inlet tube (C) symmetrically distributed around the observation main tube (O), and a spare observation tube (D) with an integrated siphon structure. The standby observation tube (D) is equipped with a siphon stop conduit (3-2). The siphon stop conduit (3-2) is located at a preset height above the liquid seal tube (3-1). During the siphon process, the siphon stop conduit introduces air to break the negative pressure of the siphon tube, thereby achieving liquid level cutoff and ensuring the stability of the measurement benchmark. All thin tubes are sealed at the top and have air inlets to balance the air pressure moving up and down on the liquid surface.

3. The shipborne high-precision rainfall measurement device based on video image and attitude synchronization according to claim 1, characterized in that, The bottom of the multi-tube array rain gauge (3) is connected to a liquid seal tube (3-1). The observation main tube (O) and the spare observation tube (D) are inserted into the liquid seal tube (3-1) to a preset depth so that a liquid seal structure is formed when the liquid level rises above the bottom of the inserted tube. The bottom surface of the liquid seal tube (3-1) is a funnel shape inclined towards the center. The inner wall edge of the connection between the liquid seal tube (3) and the water inlet tube is chamfered to form an air trap so as to discharge air bubbles in time and avoid air bubbles from interfering with the liquid level measurement. The bottom of the thin tube of the multi-tube array rain gauge (3) is provided with a damping structure or filled with a damping medium to form a coupled damping system, which is used to dissipate the oscillation energy of the liquid column and suppress the resonance of the liquid column to suppress the oscillation of the liquid surface.

4. The shipborne high-precision rainfall measurement device based on video image and attitude synchronization according to claim 1, characterized in that, The equivalent inner diameter of the multi-tube array rain gauge (3) , and the inner diameter of the rain collection head (1) The equivalent radius ratio that forms d / D corresponds to the equivalent variable diameter ratio. ; The thin tube of the multi-tube array rain gauge (3) is made of transparent hydrophobic material.

5. The shipborne high-precision rainfall measurement device based on video image and attitude synchronization according to claim 2, characterized in that, The multi-tube array rain gauge (3) replaces the central observation tube (O) with multiple observation tubes with smaller inner diameters. The bottom is connected to the outer thin tubes, forming a hybrid array configuration with the original surrounding first water inlet tube (B), second water inlet tube (C), and spare observation tube (D). The central thin tube further reduces the diameter to constrain liquid surface distortion. The total cross-sectional area of ​​the hybrid array and the cross-sectional area of ​​the rain head (1) satisfy the preset equivalent diameter ratio formula K=(d / D)².

6. The shipborne high-precision rainfall measurement device based on video image and attitude synchronization according to claim 1, characterized in that, The optical observation module (4) is a wide-angle infrared camera whose field of view covers the entire height of the multi-tube array rain gauge (3) from the bottom to the top of the effective observation point; The non-mirror area of ​​the infrared camera is protected by a metal casing.

7. The shipborne high-precision rainfall measurement device based on video image and attitude synchronization according to claim 1, characterized in that, The optical observation module (4) is a wide-angle visible light camera, whose field of view covers the entire height of the multi-tube array rain gauge (3) from the bottom to the top of the effective observation; and visible light supplement lamps are added at appropriate positions on the inner wall and top of the main body (2) to realize visible light image shooting. The non-mirror area of ​​the visible light camera is encapsulated and protected by a metal shell.

8. The shipborne high-precision rainfall measurement device based on video image and attitude synchronization according to claim 1, characterized in that, The main control module (5) integrates a pitch and roll attitude sensor and supports receiving attitude data from an external platform from the communication interface; The main control module (5) is fully enclosed in a metal shell for protection. Its built-in algorithm supports image algorithms and is used to perform edge detection and water level line recognition on the collected liquid level images.

9. The shipborne high-precision rainfall measurement device based on video image and attitude synchronization according to claim 1, characterized in that, The rain collection head (1) integrates a diversion port (1-1), a one-way valve (1-2), and a heating wire (1-3). The heating wire (1-3) has a ring-shaped structure and is used for snow melting to adapt to the measurement of rain and snow in all scenarios. An electromagnetic / manual drain valve (7) is provided on the bottom surface of the base plate (6), and a temperature and humidity sensor (8) is provided on the inner side of the base plate (6). The base plate (6) is provided with a vent hole (6-1) and a positioning hole (6-2) for setting the temperature and humidity sensor (8). The electromagnetic / manual drain valve (7) is connected to the liquid seal pipe (3-1) through the through hole on the base plate (6) and is used for draining accumulated water after the last rainfall, assisting drainage during heavy rain, and emergency manual drainage.

10. The shipborne high-precision rainfall measurement device based on video image and attitude synchronization according to claim 1, characterized in that, The connection surfaces of the main body (2) with the rain collection head (1) and the base plate (6) are all provided with sealing rubber gaskets; the power supply end of the main control module (5) integrates a lightning protection module, and the metal shell of the main control module (5), the metal shell of the optical observation module (4), and the Faraday cage structure of the main body (2) together form a double electromagnetic shielding protection. The main body (2) has a sealed dark cavity inside to isolate external natural light interference and ensure the measurement stability of the optical observation module (4). The main body (2) is made of salt spray corrosion resistant metal or non-metallic material with conductive coating on the outer wall. Together with the rain head (1) and the base plate (6), it forms a closed Faraday cage structure to shield external electromagnetic interference.

11. The shipborne high-precision rainfall measurement device based on video image and attitude synchronization according to claim 3, characterized in that, Two conductive electrodes are inserted into the top of the liquid seal tube (3-1), and the electrode ends are inserted to the depth of the internal cavity of the liquid seal tube (3-1). When the rainwater level rises to the electrode position, the measurement process is triggered. The main control module (5) also supports receiving external forced measurement commands through the communication interface.

12. The shipborne high-precision rainfall measurement device based on video image and attitude synchronization according to claim 1, characterized in that, The optical observation module (4), main control module (5), and observation main tube (O) are integrated in a sealed shell to form an independent dynamic liquid level measurement module. The observation main tube (O) is connected to the liquid cavity inside the external container through a connecting pipe. This module is suitable for installation on the external container through a mechanical interface to measure the liquid level inside the container.

13. A measurement method for a shipborne high-precision rainfall measurement device based on video image and attitude synchronization as described in any one of claims 1-11, characterized in that, Includes the following steps: a. Calibration: Pre-calibrate the correspondence between the vertical coordinate of the image pixels and the actual liquid level height. Set the position where the liquid level is slightly higher than the preset height on the upper side of the liquid seal pipe (3-1) as the zero point of the first rainfall liquid level, and calibrate the rainfall corresponding to this position. ; b. Attitude acquisition: The pitch and roll data of the carrier are acquired through the attitude sensor integrated in the main control module (5) or an external platform; c. Image acquisition: When the pitch and roll angles approach 0° and the system returns to center, the optical observation module (4) is triggered to capture liquid level images including the observation main tube (O) and the backup observation tube (D); d. Liquid level calculation: The acquired image is cropped for the region of interest (ROI), converted to grayscale, binarized, and edge detected. The mean ordinate of multiple key points on the water level line is extracted, and then combined with the equivalent diameter ratio formula. Convert to actual rainfall; when the liquid level exceeds the zero point of the initial rainfall level, the cumulative rainfall is: ; e. Siphon circulation: When the liquid level rises to the siphon trigger height, the siphon structure starts to drain the water. The liquid level stops when it drops to the height of the siphon stop pipe (3-2). This height is used as the new reference zero point for subsequent measurements. f. Data traceability and redundancy assurance: All original liquid level images, attitude data, and temperature and humidity data are retained as the basis for data traceability; the ambient temperature is collected in real time through the temperature and humidity sensor (8), based on the formula Temperature compensation was applied to the converted rainfall, whereby... Rainfall after temperature compensation. For the initial calculation of rainfall, The actual ambient temperature is used, with 25℃ as the standard compensation benchmark, to correct for errors caused by water volume expansion or contraction due to temperature changes; when the attitude sensor fails, an image-based attitude substitution algorithm is activated to obtain the equivalent pitch and roll angles by fitting the water level line angles, ensuring measurement continuity. g. Drainage reset: When the liquid level does not change after more than 24 hours or 10 consecutive measurements, open the electromagnetic / manual drain valve (7) to drain the accumulated water and wait for the next rainfall measurement.

14. A liquid level measurement method for an independent dynamic liquid level measurement module as described in claim 12, characterized in that, Includes the following steps: a. Calibration: Pre-calibrate the correspondence between the vertical coordinates of image pixels and the actual liquid level height; b. Attitude acquisition: The pitch and roll data of the carrier are acquired through the attitude sensor integrated in the main control module (5) or an external platform; c. Image acquisition: When the pitch and roll angles approach 0° and the system returns to center, the optical observation module (4) is triggered to capture the liquid level image of the observation main pipe (O); d. Liquid level calculation: Perform ROI cropping, grayscale conversion, binarization, and edge detection on the acquired image, extract the mean ordinate of multiple key points on the water level line, and calculate the actual liquid level height.

Images