Non-contact type refueling bolt offset measuring device

Abstract

The invention discloses a non-contact type refueling bolt offset measuring device, which comprises a shell, and an image acquisition module, a control module and a communication module which are arranged in the shell, and is characterized in that the image acquisition module is used for acquiring images of a detected refueling bolt and a refueling bolt code; the control module is used for carrying out image processing calculation on the image acquired by the image acquisition module and then identifying the number of the refueling bolt and the offset of the refueling bolt; and the communication module is used for sending the number of the refueling bolt and the offset of the refueling bolt to a terminal system. The method has the advantages that various pain points of traditional airport well plug offset measurement are solved in a targeted mode, and remarkable breakthrough is achieved in the aspects of measurement precision, operation efficiency, data value, operation flexibility, safety guarantee and the like.

Description

Technical Field

[0001] This invention relates to the field of fuel management technology, specifically to a non-contact fuel filler nozzle offset measuring device. Background Technology

[0002] As core comprehensive transportation hubs at the national and regional levels, airports are key nodes supporting the efficient operation of air transport networks and ensuring the smooth flow of people and goods across regions. With the continuous development of the social economy and the significant improvement of people's living standards, the demand for air travel has experienced explosive growth, and the passenger and cargo throughput of airports has been increasing year by year. Their hub value, strategic position, and role in safeguarding people's livelihoods in the comprehensive transportation system are becoming increasingly prominent.

[0003] In the airport operation and support system, safety is the core bottom line. Besides continuously strengthening the airport's own basic operational safety (such as runway takeoff and landing safety, terminal personnel security, and aircraft ground dispatch safety), the safe and stable operation of the airport's fuel supply pipeline network is equally crucial. This pipeline network, as the "lifeline" of aircraft fuel replenishment, is a key energy link ensuring on-time flight takeoffs and landings and efficient aviation operations. The detection of "minor leaks" is a key concern in current aviation safety management: on the one hand, minor fuel leaks are easily concealed and accumulate over time, not only wasting valuable fuel resources but also potentially creating flammable and explosive hazard areas around the pipeline network; on the other hand, once the leak increases over time or is triggered by open flames, static electricity, or other conditions, it can easily lead to major safety accidents such as fires and explosions, directly threatening the operational safety of core areas such as airport runways and aprons, and even endangering the lives and property of aircraft, crew members, and ground staff. Therefore, how to accurately and efficiently detect minor leaks in the airport fuel supply pipeline network has become a key technical issue that urgently needs to be addressed in the aviation industry's safety management system.

[0004] The detection of minor leaks in airport fuel supply pipelines can be achieved by measuring the offset of refueling plugs. Traditional airport refueling plug offset measurement uses mechanical measuring devices. The core process is as follows: the target refueling plug is manually located and positioned, the measuring device is fixed to the refueling plug using a special clamp structure, the offset data is read with the help of mechanical measuring components such as a ruler, and finally the operator manually records the measurement results and enters them into the relevant system to complete the offset detection of a single refueling plug.

[0005] However, traditional detection methods have the following drawbacks:

[0006] Measurement accuracy is significantly affected by human factors. Operational errors in manual positioning and clamp fixing, as well as visual errors in scale readings, result in large data deviations and low measurement reliability.

[0007] The measurement efficiency is low. The manual positioning, fixing, reading and data entry process is cumbersome and requires testing each refueling hydrant in turn. It cannot meet the needs of airport scenarios with a large number of refueling hydrants and wide distribution, and it is difficult to meet the requirements of efficient batch measurement.

[0008] The data management capabilities are weak, and manual data entry is prone to errors or omissions. It can only save simple offset values ​​and cannot record detailed information such as measurement location and time, and it lacks data traceability capabilities.

[0009] Lacking trend analysis capabilities, traditional measurements can only obtain offset data at a single point in time. Without the support of big data analysis, it is impossible to predict the evolution trend of the fuel plug offset, making it difficult to assist management in formulating targeted maintenance plans.

[0010] The lack of operational flexibility and the insufficient portability of mechanical devices make the transportation and installation of refueling plugs scattered in different areas time-consuming and labor-intensive, further reducing the overall measurement efficiency.

[0011] Currently, computer vision-based non-contact inspection technology is widely used in industrial measurement due to its advantages of high efficiency, accuracy, and non-destructiveness. Combined with big data analytics and intelligent positioning technology, it enables real-time processing of measurement data, trend prediction, and end-to-end traceability, providing a scientific basis for equipment maintenance. Therefore, developing a non-contact intelligent measurement device suitable for airport refueling hydrant measurement scenarios could effectively address many limitations of traditional measurement methods and meet the operational support needs of modern airports. Summary of the Invention

[0012] In view of the shortcomings of the prior art, the purpose of this invention is to provide a non-contact refueling plug offset measuring device to overcome the defects of existing mechanical airport refueling plug offset measuring technology.

[0013] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0014] A non-contact fuel filler nozzle offset measuring device, the key feature of which includes a housing and an image acquisition module, a control module, and a communication module disposed within the housing, wherein:

[0015] The image acquisition module is used to acquire images of the tested fuel plug and its code.

[0016] The control module is used to perform image processing and calculation on the images acquired by the image acquisition module to identify the fuel plug number and fuel plug offset.

[0017] The communication module is used to send the fuel filler number and fuel filler offset to the terminal system.

[0018] Furthermore, the outer casing is composed of a front cover and a rear cover, and the image acquisition module, control module and communication module are located in the mounting cavity formed between the front cover and the rear cover, with the lens of the image acquisition module exposed.

[0019] Furthermore, an installation window is provided on the front cover, in which a display screen is installed and fixed. The display screen is electrically connected to the control module. An installation hole is provided on the front cover on the upper right side of the installation window, in which a button module electrically connected to the control module is installed.

[0020] Furthermore, a power supply module for supplying power to the image acquisition module, control module, and communication module is also installed in the mounting cavity formed between the front cover and the rear cover.

[0021] Furthermore, a carrying handle is formed on the top of the rear cover.

[0022] Furthermore, the step of the control module identifying the fuel filler number and offset after performing image processing calculations on the image acquired by the image acquisition module includes:

[0023] Step 1: Preprocess the images acquired by the image acquisition module and identify the fuel filler number;

[0024] Step 2: Use the Newton-Raphson iterative optimization algorithm for dual-circle cooperative localization to locate the fuel plug in the preprocessed image;

[0025] Step 3: Calculate the center-to-center distance between the inner and outer circles of the filler plug, and convert the offset of the inner valve core relative to the standard position to obtain the filler plug offset.

[0026] Furthermore, the preprocessing step in step 1 for the image acquired by the image acquisition module includes:

[0027] Step 1.1: Analyze the image grayscale histogram to determine the noise type of the current scene, and select different filtering algorithms to perform scene-specific noise filtering based on the noise type.

[0028] Step 1.2: Based on the rectangular arrangement features of the fuel injector codes in the image, perform dynamic angle correction and background separation on the filtered image;

[0029] Step 1.3: Initially identify the radial edge direction of the fuel filler plug using Canny edge detection, then tilt the filtering weights of the Laplacian operator toward the radial edge direction to sharpen and enhance the edge region of the circular outline of the fuel filler plug.

[0030] Furthermore, the strategy for scene-specific noise filtering based on different filtering algorithms in step 1.1 is as follows:

[0031] For strong light overexposure noise, an adaptive weighted average filtering algorithm is used;

[0032] For oil stain noise, first use guided filtering to preserve the edge contour, and then use median filtering to eliminate dark spot particle noise;

[0033] For random particle noise, a 3×3 weighted average filtering algorithm is used.

[0034] Furthermore, step 2, which uses the Newton-Raphson iterative optimization algorithm with dual-circle cooperative localization to locate the fuel plug in the preprocessed image, includes the following steps:

[0035] Step 2.1: Extract circular features from the preprocessed image using the Hough gradient method;

[0036] Step 2.2: Based on the standard size parameters of the airport refueling hydrant, select candidate double-circle regions that meet the size constraints, and mark the relative positional relationship of the double circles;

[0037] Step 2.3: Determine the constraints of the Newton-Raphson iterative optimization algorithm, identify the inner and outer circles of the fuel filler valve, and realize the positioning of the fuel filler valve.

[0038] Furthermore, the constraints of the Newton iterative optimization algorithm are as follows:

[0039] Using the concentricity of the two circles as a hard constraint, the distance between the centers of the inner and outer circles is checked in real time during the iteration process. If the distance exceeds the preset threshold, the iteration direction is automatically corrected.

[0040] The significant effects of this invention are:

[0041] This invention addresses many pain points of traditional airport well plug offset measurement through an integrated technical solution combining non-contact measurement, intelligent algorithms, portable design, and big data linkage. It achieves significant breakthroughs in measurement accuracy, operational efficiency, data value, operational flexibility, and safety assurance. Specific benefits are as follows:

[0042] 1. Significantly improves measurement accuracy and ensures data reliability: This invention abandons the traditional mechanical contact measurement and manual reading mode. It utilizes a combination of high-definition industrial camera image recognition, image preprocessing, and Newton's iteration method for center positioning, eliminating manual positioning deviations, bayonet fixing errors, and scale visual reading errors at the source. Specifically, image preprocessing effectively reduces the interference of brightness noise, complex backgrounds, and shooting angles on the detection. Newton's iteration method approximates the optimal solution through multiple iterations, accurately identifying the circular outline and center coordinates of the well plug. Ultimately, this significantly reduces the measurement error of the well plug offset, significantly improving data reliability and consistency, and providing a precise data foundation for subsequent offset trend analysis and maintenance decisions.

[0043] 2. Significantly Improved Measurement Efficiency, Adapting to Complex Airport Scenarios: Addressing the core scenario requirement of "numerous and widely distributed" well plugs in airports, this invention optimizes efficiency through two major design improvements: Firstly, the device integrates automatic positioning and well plug number recognition functions, eliminating the need for manual searching and marking of well plug numbers, and quickly matching target measurement objects, saving the tedious process of traditional manual positioning. Secondly, the device features a portable structure with a carrying handle, facilitating flexible transport by operators in different areas such as the apron and refueling hydrant clusters. Simultaneously, standardized industrial heat dissipation holes ensure the stability of the device during long-term continuous operation, preventing overheating and shutdown. Combined with the "on-site measurement + one-click upload from the control center" process, the measurement time for a single well plug is reduced by more than 80% compared to traditional methods, enabling efficient mobile detection of batch well plugs and completely solving the problem of low efficiency in traditional measurements.

[0044] 3. Achieve end-to-end data traceability and improve management standardization: This invention breaks through the limitations of traditional manual data entry. During the measurement process, it can automatically and synchronously record complete data dimensions, including well plug number, measurement time, real-time location, offset value, and original image data. The data can be uploaded to the control center with one click without manual intervention, forming a structured database for storage. This design not only avoids the risks of errors, omissions, and tampering associated with manual recording, but also achieves a closed-loop process of "measurement-storage-traceability." When it is necessary to verify the historical measurement status of a well plug, the complete data chain can be quickly retrieved through the well plug number, providing strong support for tracing the source of quality problems and verifying the compliance of the measurement process, significantly improving the management standardization of airport well plug measurement data.

[0045] 4. Supports deviation trend prediction and builds a preventive maintenance system: Traditional measurement can only obtain deviation data at a single point in time, and cannot determine the evolution pattern of well plug deviation. However, this invention, through the linkage mode of "device acquisition + big data analysis at the control center", can perform time-series analysis on the historical deviation data of the same well plug, generating deviation change trend curves such as monthly deviation growth rate and quarterly deviation accumulation. Based on this trend, management can accurately identify well plugs with accelerated deviation and deviation exceeding the threshold risk, predict potential safety hazards in advance, replace the traditional passive mode of "repair after failure", and build a preventive protection system of "trend prediction - early maintenance", effectively reducing safety risks such as abnormal refueling docking and fuel leakage caused by well plug deviation.

[0046] 5. Lowering operational barriers and operating costs, and optimizing resource allocation: From a practical and cost perspective, the beneficial effects of this invention are also reflected in the following aspects: On the one hand, the device has a high degree of automation, and operators do not need to have complex mechanical measurement skills or image algorithm knowledge. They only need simple training to start working, which greatly reduces the barriers to manual operation; on the other hand, based on accurate trend analysis and preventive maintenance, blind maintenance and delayed maintenance can be avoided, significantly reducing unnecessary human and material investment, optimizing airport operation and maintenance resource allocation, and achieving a dual balance between safety assurance and cost control. Attached Figure Description

[0047] Figure 1 This is a schematic diagram of the structure of the present invention;

[0048] Figure 2 This is a rear view of the present invention;

[0049] Figure 3 This is an exploded view of the present invention;

[0050] Figure 4 This is the working state of the present invention;

[0051] Figure 5 This is a flowchart of image processing calculations in this invention. Detailed Implementation

[0052] The specific embodiments and working principles of the present invention will be further described in detail below with reference to the accompanying drawings.

[0053] Based on the background technology of this application, it is clear that the core objective of this invention is to overcome the shortcomings of existing mechanical airport well bolt offset measurement technology, specifically addressing the following technical problems:

[0054] Solving the problems of low accuracy and high human interference in traditional measurements: By using a non-contact measurement method combined with image recognition technology from a high-definition industrial camera, the mechanical measurement process of manual positioning, bayonet fixing, and scale reading is replaced, eliminating positioning errors, fixing deviations, and visual reading errors, thereby improving the accuracy and reliability of well plug offset data.

[0055] To address the problem of low efficiency in traditional measurement methods: For scenarios where airport well plugs are widely distributed and need to be inspected one by one, the device integrates automatic positioning and well plug number recognition functions, eliminating the need for manual location searching; at the same time, the portability of the device optimizes the transportation and operation process, reducing the measurement time of a single well plug and meeting the needs of efficient inspection of batch well plugs.

[0056] Solve the problems of chaotic and untraceable traditional data management: realize the automated recording of measurement data (including complete information such as well plug number, measurement time, location, and offset), and support one-click data upload after returning to the control center, avoid the risk of errors and omissions from manual entry, ensure the traceability of data throughout the entire process, and improve the standardization of data management.

[0057] To address the issues of traditional measurement methods lacking trend analysis and failing to predict risks: By linking the device with the control center, big data analytics is used to process historical measurement data, generating the evolution trend of wellbore offset. This fills the gap in traditional technology, which "can only acquire data at a single point in time," and provides data support for predicting wellbore offset risks.

[0058] To address the problem that traditional measurements cannot support the development of scientific maintenance plans: By relying on accurate real-time data and trend analysis results, we provide management with visualized well hydrant status reports to assist in the development of targeted maintenance plans, avoid blind or delayed maintenance, and ensure the long-term safe and stable operation of airport refueling hydrants.

[0059] Therefore, the technical solution adopted by the present invention is as follows:

[0060] Example:

[0061] like Figures 1-4 As shown, a non-contact fuel filler nozzle offset measuring device includes a housing and an image acquisition module 1.4, a control module 1.7, and a communication module disposed within the housing, wherein:

[0062] The image acquisition module 1.4 is used to acquire images of the tested fuel plug and its code;

[0063] The control module 1.7 is used to perform image processing calculations on the images acquired by the image acquisition module 1.4 to identify the fuel filler number and fuel filler offset.

[0064] The communication module is integrated into the control module 1.7 and is used to send the fuel filler number and fuel filler offset to the terminal system.

[0065] The device is in the following state during use: Figure 4 As shown, after the images of the tested well plug 2.2 and well plug code 2.3 are captured by the image acquisition module 1.4, they are transmitted to the control module for storage. After image processing and calculation, the well plug number and well plug offset are identified and stored in a local CSV document. After arriving at the control room, they can be wirelessly sent to the terminal system for big data analysis with one click.

[0066] The outer casing is composed of a front cover 1.1 and a rear cover 1.6. The image acquisition module 1.4, the control module 1.7 and the communication module are located in the mounting cavity formed between the front cover 1.1 and the rear cover 1.6, and the lens of the image acquisition module 1.4 is exposed.

[0067] from Figures 1-4 It can also be seen that a mounting window is provided on the front cover 1.1, and a display screen 1.2 is installed and fixed in the mounting window. The display screen 1.2 is electrically connected to the control module 1.7. A mounting hole is provided on the front cover 1.1 on the upper right side of the mounting window, and a button module 1.8 electrically connected to the control module 1.7 is installed in the mounting hole. A power module 1.3 for supplying power to the image acquisition module 1.4, the control module 1.7 and the communication module is also installed in the mounting cavity formed between the front cover 1.1 and the rear cover 1.6. A carrying ring 1.5 is also formed on the top of the rear cover 1.6.

[0068] In implementation, the casing is made of PC material, the display screen 1.2 is a 7-inch IPS LCD screen, the power module uses a large-capacity explosion-proof battery pack and a voltage regulator module; the image acquisition module 1.4 uses an industrial camera, and the control module 1.7 uses the Orangepi AIPRO development board with a 5G communication module.

[0069] Based on the above structure, it can be seen that the overall structural design of this device strictly meets the requirements of industrial use, and its core is adapted to the mobile detection scenario at the airport: the device integrates a dedicated 1.5-inch carrying handle, which makes it easy for operators to flexibly move the device in different areas of the airport, such as the apron and the refueling hydrant cluster area, fully meeting the "portable" operation requirements.

[0070] In the specific implementation process, there are many noise points in the images collected at the airport. In order to address the problems of brightness noise interference, complex background environment and shooting angle deviation in the image background, this embodiment first performs preprocessing operations on the collected original images: through noise reduction, background separation and angle correction, the influence of irrelevant interference factors on subsequent detection is effectively reduced; on this basis, an edge detection algorithm is used to extract the edge features in the image, clearly delineating the contour boundary of the target area, laying the foundation for subsequent circular recognition.

[0071] In actual airport well plug offset detection scenarios, the above design can significantly reduce the difficulty of transportation and operation for operators, while ensuring the continuous and reliable operation of the device in complex field environments. Ultimately, it can achieve efficient and accurate acquisition of well plug offset data, perfectly meeting the detection needs of airport well plugs that are "widely distributed and require mobile operation".

[0072] To accurately locate the target circle and its center, Newton's iteration method is introduced: based on preprocessed edge features, multiple iterations are performed, and the results of each iteration are continuously compared and corrected with preset target parameters such as the circular contour feature threshold, gradually approaching the optimal solution. Ultimately, this method not only accurately detects the target circular region but also simultaneously determines the specific coordinates of its center, ensuring the accuracy of subsequent wellbore offset calculations. The technical process is as follows: Figure 5 As shown:

[0073] That is, the steps of the control module 1.7 in identifying the fuel filler number and offset after performing image processing calculations on the image acquired by the image acquisition module 1.4 include:

[0074] Step 1: Preprocess the image acquired by the image acquisition module 1.4 and identify the fuel filler number;

[0075] In conventional filtering processes, smoothing filters can weaken image features and blur image edges. To eliminate this effect, the sharpening filter used in this invention is the opposite of smoothing filters. It weakens or eliminates low-frequency components in the image without affecting high-frequency components, thus increasing image contrast and making edges more defined. In practical applications, it can be used to enhance blurred details or the edges of targets.

[0076] First, some definitions are used in this embodiment as follows:

[0077] Laplace operator definition:

[0078]

[0079] Discrete mask representation:

[0080]

[0081] Discrete expressions in digital image processing:

[0082]

[0083] As can be seen from the above, the filtering expressions with f(x,y) as the filtering center and the x-axis are as follows:

[0084]

[0085] Performing second derivative operations on image pixels can efficiently extract feature information about image edge changes. However, it should be noted that the Laplacian operator is extremely sensitive to noise, which can cause edge detection results to be distorted by noise interference. Therefore, noise filtering must be performed before using the Laplacian operator for edge feature extraction to reduce the impact of noise on the subsequent edge detection accuracy.

[0086] Therefore, considering the brightness noise, complex background, shooting angle deviation, and oil / wear interference unique to airport scenes in the original image, this embodiment abandons the traditional fixed parameter preprocessing process and innovatively designs a scene-aware adaptive preprocessing logic. The specific steps are as follows:

[0087] Step 1.1, Dynamic Angle Correction and Background Separation:

[0088] Traditional smoothing filters weaken image edge features, while single sharpening filters are insufficient for suppressing strong light noise. This embodiment first uses image grayscale histogram analysis to automatically determine the noise type of the current scene, and then selects different filtering algorithms based on the noise type to perform scene-specific noise filtering. Specifically:

[0089] For strong light overexposure noise, an adaptive weighted average filtering algorithm is used to increase the neighborhood weight ratio of overexposed pixel areas while preserving the details of non-overexposed areas.

[0090] For oil stain dark spot noise, a hybrid filtering strategy of guided filtering + median filtering is introduced. First, guided filtering is used to preserve the edge contour, and then median filtering is used to eliminate dark spot particle noise, avoiding edge blurring or noise residue caused by single filtering.

[0091] For random particle noise, a 3×3 weighted average filtering algorithm is used to eliminate interference for subsequent edge detection.

[0092] Step 1.2, Dynamic Angle Correction and Background Separation: Based on the rectangular arrangement features of the fuel injector codes in the image, dynamic angle correction and background separation are performed on the filtered image;

[0093] During on-site measurements at airports, operators often use handheld devices, resulting in tilted shooting angles, and apron markings and ground support equipment can easily create background interference. This embodiment innovatively introduces a well plug code feature anchoring method to achieve angle correction: first, the rectangular arrangement feature of well plug codes 2.3 in the image is identified; then, using the horizontal / vertical boundaries of the codes as a reference, the image tilt angle is calculated and real-time correction is completed. Compared with traditional straight-line correction based on Hough transform, anchoring well plug codes can avoid interference from ground markings, improving correction accuracy by more than 40%.

[0094] In the background separation stage, an adaptive threshold segmentation based on color clustering is adopted. By extracting the inherent grayscale range of the hydrant's metal material and combining it with the color features of the coded region, a background segmentation model is established. Irrelevant backgrounds such as the tarmac ground and surrounding obstacles are automatically removed, leaving only the hydrant body and the coded region, which greatly reduces the computational load of subsequent feature extraction.

[0095] Step 1.3: Initially identify the radial edge direction of the fuel filler plug using Canny edge detection, then tilt the filtering weights of the Laplacian operator toward the radial edge direction to sharpen and enhance the edge region of the circular outline of the fuel filler plug.

[0096] The specific implementation process is as follows: In response to the edge blurring problem caused by wear of the well plug, this embodiment improves the traditional Laplacian operator and designs an edge-oriented sharpening operator: First, the radial edge direction of the well plug is initially identified by Canny edge detection, and then the filtering weight of the Laplacian operator is tilted towards the radial edge direction. Only the edge area of ​​the circular outline of the well plug is sharpened and enhanced, while the original pixel features are preserved in the non-edge area. This not only solves the problem of overall image distortion caused by traditional sharpening, but also enhances the recognition of key edges.

[0097] Step 2: Use the Newton-Raphson iterative optimization algorithm for dual-circle cooperative localization to locate the fuel plug in the preprocessed image;

[0098] Airport refueling hydrants typically feature a double-circle structure with an inner valve core and an outer flange. Traditional single-circle positioning algorithms are prone to failure due to wear or obstruction of the single circle. This embodiment innovatively proposes a Newton-style iterative positioning method with dual-circle cooperative constraints to achieve precise locking of the circle center coordinates. The specific process is as follows:

[0099] Step 2.1: Extract circular features from the preprocessed image using the Hough gradient method;

[0100] Step 2.2: Based on the standard size parameters of the airport refueling plug (preset inner / outer circle diameter range), select double-circle candidate areas that meet the size constraints, and mark the relative positional relationship of the double circles, such as concentricity, spacing threshold, etc., and remove pseudo-circular features caused by oil stains and wear.

[0101] Step 2.3: Determine the constraints of the Newton-Raphson iterative optimization algorithm, identify the inner and outer circles of the fuel filler valve, and realize the positioning of the fuel filler valve.

[0102] Traditional Newton's iterative method relies solely on the edge pixels of a single circle for iteration, making it susceptible to convergence deviations due to local disturbances. This embodiment introduces a dual-circle cooperative constraint condition into the iterative calculation:

[0103] Using the concentricity of the two circles as a hard constraint, the distance between the centers of the inner and outer circles is checked in real time during the iteration process. If the distance exceeds the preset threshold, the iteration direction is automatically corrected.

[0104] A weighted iterative weight allocation method is adopted. Since the wear degree of the outer ring is lower than that of the inner ring and the contour is more stable, the edge pixel weight of the outer ring flange is set to 1.2 times that of the inner ring valve core. The outer ring center is fitted first, and then the inner ring center coordinates are optimized based on the outer ring center to ensure the stability of the iterative results.

[0105] Step 3: Calculate the center-to-center distance between the inner and outer circles of the filler plug, and convert the offset of the inner valve core relative to the standard position to obtain the filler plug offset.

[0106] After completing the double-circle center positioning, the offset of the inner ring valve core relative to the standard position is calculated. At the same time, the concentricity deviation between the inner and outer rings is included in the offset evaluation system. That is, if the concentricity deviation exceeds the threshold, it is simultaneously marked as a well plug structure abnormality, providing additional structural status data for subsequent maintenance, thus breaking through the limitation of traditional algorithms that only calculate the offset of a single circle center.

[0107] As can be seen from the above steps, this embodiment of the invention innovatively designs a multimodal adaptive preprocessing module to address the special interferences such as strong light, oil stains, and wear in airport scenarios. Through a combination strategy of scene-specific filtering, encoding anchor angle correction, and directional sharpening enhancement, it solves the problem of insufficient robustness of traditional algorithms in complex airport environments. At the same time, based on the double-circle structure characteristics of the refueling hydrant, it proposes a Newton iterative positioning method with double-circle collaborative constraints, introduces concentricity hard constraints and weighted iterative weights, and incorporates concentricity deviation into the offset evaluation, thus overcoming the limitations of single-circle positioning being prone to failure and having a single data dimension.

[0108] Finally, the Orangepi AIPRO development board on this device has limited computing power, and directly running complex algorithms would lead to insufficient real-time performance. Therefore, this embodiment innovatively designs a lightweight inference logic with hierarchical computing power allocation to achieve a balance between accuracy and real-time performance:

[0109] Front-end rapid screening: A lightweight feature screening model is first run locally on the device. Only images containing well plug codes and double circle features are processed in the entire process. Invalid images are directly removed to reduce unnecessary computation.

[0110] Hierarchical scheduling of computing power: Low-computing-power tasks such as preprocessing and initial screening are completed locally on the device. The high-precision calculation task of dual-circle collaborative iteration is split into "local coarse fitting + control center fine fitting". The local device first completes the coarse positioning of the circle center coordinates (error ≤ 0.5mm) to meet the needs of rapid reading on site. After returning to the control center, the fine fitting is completed through the high-performance server (error ≤ 0.1mm). This ensures both on-site operation efficiency and high accuracy of the final data.

[0111] Data compression and storage: The original image and intermediate calculation data are lightly compressed, retaining only key data such as the double circle outline of the well plug, the center coordinates, and the encoding information. This data is stored in a structured format in a local CSV document, which saves device storage resources and facilitates one-click uploading to the terminal system later.

[0112] Based on the above reasoning logic, this device adapts to the hardware computing power of the device and innovatively constructs a lightweight inference architecture with hierarchical computing power allocation to achieve hierarchical calculation of "local coarse fitting + control center fine fitting", which takes into account the real-time performance of on-site measurements and the high accuracy of the final data, and solves the contradiction between the computing power of embedded devices and the accuracy of algorithms.

[0113] In summary, this invention addresses many pain points of traditional airport well plug offset measurement through an integrated technical solution combining non-contact measurement, intelligent algorithms, portable design, and big data linkage, achieving significant breakthroughs in measurement accuracy, operational efficiency, data value, operational flexibility, and safety assurance.

[0114] The technical solution provided by this invention has been described in detail above. Specific examples have been used to illustrate the principles and implementation methods of this invention. The descriptions of the embodiments above are only for the purpose of helping to understand the method and core ideas of this invention. It should be noted that those skilled in the art can make several improvements and modifications to this invention without departing from the principles of this invention, and these improvements and modifications also fall within the protection scope of the claims of this invention.

Claims

1. A non-contact fuel filler nozzle offset measuring device, characterized in that: It includes a housing and an image acquisition module (1.4), a control module (1.7), and a communication module housed within the housing, wherein: The image acquisition module (1.4) is used to acquire images of the tested fuel plug and its code; The control module (1.7) is used to perform image processing calculations on the image acquired by the image acquisition module (1.4) to identify the fuel plug number and fuel plug offset; The communication module is used to send the fuel filler number and fuel filler offset to the terminal system.

2. The non-contact fuel filler nozzle offset measuring device according to claim 1, characterized in that: The outer shell is composed of a front cover (1.1) and a rear cover (1.6). The image acquisition module (1.4), the control module (1.7) and the communication module are located in the mounting cavity formed between the front cover (1.1) and the rear cover (1.6), and the lens of the image acquisition module (1.4) is exposed.

3. The non-contact fuel filler nozzle offset measuring device according to claim 2, characterized in that: An installation window is provided on the front cover (1.1), and a display screen (1.2) is installed and fixed in the installation window. The display screen (1.2) is electrically connected to the control module (1.7). An installation hole is provided on the front cover (1.1) on the upper right side of the installation window, and a button module (1.8) electrically connected to the control module (1.7) is installed in the installation hole.

4. The non-contact fuel filler nozzle offset measuring device according to claim 2, characterized in that: A power supply module (1.3) for supplying power to the image acquisition module (1.4), control module (1.7) and communication module is also installed in the mounting cavity formed between the front cover (1.1) and the rear cover (1.6).

5. The non-contact fuel filler nozzle offset measuring device according to claim 2, characterized in that: A carrying handle (1.5) is also formed on the top of the rear cover (1.6).

6. The non-contact fuel filler nozzle offset measuring device according to any one of claims 1-5, characterized in that: The steps by which the control module (1.7) identifies the fuel filler number and offset after performing image processing calculations on the image acquired by the image acquisition module (1.4) include: Step 1: Preprocess the image acquired by the image acquisition module (1.4) and identify the fuel filler number; Step 2: Use the Newton-Raphson iterative optimization algorithm for dual-circle cooperative localization to locate the fuel plug in the preprocessed image; Step 3: Calculate the center-to-center distance between the inner and outer circles of the filler plug, and convert the offset of the inner valve core relative to the standard position to obtain the filler plug offset.

7. The non-contact fuel filler nozzle offset measuring device according to claim 6, characterized in that: Step 1, which involves preprocessing the image acquired by the image acquisition module (1.4), includes the following steps: Step 1.1: Analyze the image grayscale histogram to determine the noise type of the current scene, and select different filtering algorithms to perform scene-specific noise filtering based on the noise type. Step 1.2: Based on the rectangular arrangement features of the fuel injector codes in the image, perform dynamic angle correction and background separation on the filtered image; Step 1.3: Initially identify the radial edge direction of the fuel filler plug using Canny edge detection, then tilt the filtering weights of the Laplacian operator toward the radial edge direction to sharpen and enhance the edge region of the circular outline of the fuel filler plug.

8. The non-contact fuel filler nozzle offset measuring device according to claim 7, characterized in that: The strategy for scene-specific noise filtering based on noise type in step 1.1 is as follows: For strong light overexposure noise, an adaptive weighted average filtering algorithm is used; For oil stain noise, first use guided filtering to preserve the edge contour, and then use median filtering to eliminate dark spot particle noise; For random particle noise, a 3×3 weighted average filtering algorithm is used.

9. The non-contact fuel filler nozzle offset measuring device according to claim 6, characterized in that: Step 2, which uses the Newton-Raphson iterative optimization algorithm with dual-circle cooperative localization to locate the fuel plug in the preprocessed image, includes the following steps: Step 2.1: Extract circular features from the preprocessed image using the Hough gradient method; Step 2.2: Based on the standard size parameters of the airport refueling hydrant, select candidate double-circle regions that meet the size constraints, and mark the relative positional relationship of the double circles; Step 2.3: Determine the constraints of the Newton-Raphson iterative optimization algorithm, identify the inner and outer circles of the fuel filler valve, and realize the positioning of the fuel filler valve.

10. The non-contact fuel filler nozzle offset measuring device according to claim 9, characterized in that: The constraints of the Newton-Raphson iterative optimization algorithm are: Using the concentricity of the two circles as a hard constraint, the distance between the centers of the inner and outer circles is checked in real time during the iteration process. If the distance exceeds the preset threshold, the iteration direction is automatically corrected.

Images