Computer vision-based tank safety monitoring method and device, and storage medium

Abstract

The present application relates to the technical field of storage tank monitoring, and particularly relates to a storage tank safety monitoring method and device based on computer vision and a storage medium. The method comprises: performing color region segmentation processing on a sample image to divide sample color regions; extracting and quantifying color features of the sample color regions to obtain sample color features; constructing a color oil and gas mapping model based on the sample color features and sample environment data to obtain sample fitting parameters; constructing a floating plate analysis model based on the sample color features to determine a floating plate state; constructing a color compensation coefficient based on the sample environment data and dynamically correcting the division of the sample color regions based on the color compensation coefficient; collecting monitoring data of the storage tank, and determining oil and gas concentration and the floating plate state based on the monitoring data, the sample fitting parameters and the floating plate analysis model. The present application realizes accurate analysis of oil and gas concentration in the storage tank and accurate monitoring of floating plate abnormalities.

Description

Technical Field

[0001] This invention relates to the field of storage tank monitoring technology, and in particular to a storage tank safety monitoring method, device and storage medium based on computer vision. Background Technology

[0002] With the rapid development of the petrochemical industry, storage tanks, as core facilities for oil and gas storage and transfer, are of paramount importance for safe operation. In recent years, the integration of computer vision and artificial intelligence technologies has provided new solutions for industrial safety monitoring. Vision-based monitoring systems, with advantages such as non-contact operation, high precision, and real-time analysis, are gradually becoming an important development direction in the field of storage tank safety monitoring, helping to improve the intelligence and safety level of storage tank operation.

[0003] Existing tank safety monitoring technologies mainly rely on manual inspections and fixed-point sensors, which suffer from problems such as large monitoring blind spots, slow response, and susceptibility to environmental interference. Monitoring of floating roof status and oil and gas concentration lacks real-time performance and accuracy, and cannot effectively warn of potential hazards such as floating roof tilting, leakage, or oil and gas accumulation. Furthermore, existing methods have poor stability in complex environments and are difficult to meet the needs of modern tank safety management. Summary of the Invention

[0004] The purpose of this invention is to provide a method, device, and storage medium for monitoring the safety of storage tanks based on computer vision, so as to solve at least one of the problems existing in the prior art.

[0005] To achieve the above objectives, the present invention adopts the following technical solution:

[0006] A computer vision-based method for monitoring the safety of storage tanks, comprising:

[0007] Perform color region segmentation on the sample image to divide the sample color regions;

[0008] Color features are extracted and quantized from the sample color regions to obtain the sample color features;

[0009] A color oil and gas mapping model is constructed based on sample color features and sample environmental data to obtain sample fitting parameters;

[0010] A floating disk analysis model is constructed based on the color features of the samples to determine the status of the floating disk;

[0011] A color compensation coefficient is constructed based on the sample environment data, and the division of the sample color region is dynamically corrected based on the color compensation coefficient.

[0012] Collect monitoring data from the storage tanks, and determine the oil and gas concentration and floating roof status based on the monitoring data, sample fitting parameters, and floating roof analysis model.

[0013] Furthermore, sample data of the storage tank is collected, and the image change area is extracted based on the sample image and the preset background image. Pixels in the sample image whose gray value is greater than that in the preset background image are taken as changed pixels, and the area formed by the changed pixels is taken as the image change area.

[0014] The sample image is converted from the RGB color space to the HSV color space, and the image variation area in the sample image is divided into sample red area and sample yellow area. The area with red hue in the image variation area is divided into sample red area, and the area with yellow hue in the image variation area is divided into sample yellow area.

[0015] Furthermore, the number of pixels in the red region of the sample is counted and used as the red feature area; the number of pixels in the yellow region of the sample is counted and used as the yellow feature area; the coordinates of the center pixel of the red region of the sample are used as the red coloring position; and the coordinates of the center pixel of the yellow region of the sample are used as the yellow coloring position.

[0016] Furthermore, an area-based hydrocarbon correlation function is constructed based on the red and yellow feature areas and the sample hydrocarbon concentration. The area-based hydrocarbon correlation function is set as: C=k1×(A1+A2)+k2×(A1+A2) 2 +b, where C represents the sample oil and gas concentration, A1 represents the red feature area, A2 represents the yellow feature area, k1 represents the first fitting coefficient, k2 represents the second fitting coefficient, and b represents the fitting constant term. The sample fitting parameters in the area oil and gas correlation function are determined using the least squares method.

[0017] Furthermore, the red and yellow color-developing positions within the last 30 seconds are extracted and numbered and sorted in chronological order, and the sequence of red and yellow color-developing positions is taken as the color development time sequence;

[0018] Extract the red and yellow feature areas within the last 15 seconds and number and sort them in chronological order. Use the sequence of red and yellow feature areas as the area time series.

[0019] Furthermore, the timing change parameters are determined based on the area timing sequence, and the timing offset parameters are determined based on the color development timing sequence. The floating roof status is judged based on the timing change parameters, timing offset parameters, and area timing sequence. When F1 > 0.2, the floating roof status is determined to be floating hull leakage, and a floating hull leakage alarm is triggered. When F2 ≥ 0.8 and -0.2 ≤ F1 ≤ 0.2, the floating roof status is determined to be floating roof tilt, and a floating roof tilt alarm is triggered. In the remaining cases, the floating roof status is determined to be normal.

[0020] Furthermore, a color compensation coefficient is constructed based on the sample temperature and sample humidity. The expression for the color compensation coefficient is: Q=1+α1×(T-T0)+α2×(H-H0), where Q represents the color compensation coefficient, α1 represents the temperature weight, α2 represents the humidity weight, T represents the sample temperature, T0 represents the preset storage tank temperature, H represents the sample humidity, and H0 represents the preset storage tank humidity.

[0021] The saturation of the sample image is corrected based on the color compensation coefficient. The saturation of the corrected sample image is set as P, where P = p × Q, and p represents the saturation of the sample image before correction. The saturation of the corrected sample image is used to divide the sample red area and the sample yellow area to dynamically correct the division of the sample color area.

[0022] Furthermore, the monitoring image is segmented by color regions to divide the monitoring color regions, and color features are extracted and quantified from the monitoring color regions to obtain monitoring color features. The red and yellow feature areas in the monitoring color features are input into the area oil and gas correlation function to obtain the oil and gas concentration, and the monitoring color features are input into the floating disk analysis model to obtain the floating disk status.

[0023] On the other hand, the present invention also provides a computer vision-based tank safety monitoring device, comprising:

[0024] The sample acquisition module is used to collect sample data from the storage tank.

[0025] The region segmentation module is used to perform color region segmentation processing on the sample image to divide the sample color regions;

[0026] The feature extraction module is used to extract and quantify the color features of the sample color region to obtain the sample color features;

[0027] The fitting analysis module is used to construct a color oil and gas mapping model based on sample color features and sample environmental data in order to obtain sample fitting parameters;

[0028] The floating disk analysis module is used to build a floating disk analysis model based on the color features of the samples in order to determine the status of the floating disk.

[0029] The segmentation and correction module is used to construct color compensation coefficients based on sample environmental data and to dynamically correct the segmentation of sample color regions based on the color compensation coefficients.

[0030] The monitoring and analysis module is used to collect monitoring data from the storage tank and determine the oil and gas concentration and floating roof status based on the monitoring data, sample fitting parameters, and floating roof analysis model.

[0031] On the other hand, the present invention also provides a storage medium characterized in that it stores instructions that, when run on a computer, cause the computer to execute the computer vision-based tank safety monitoring method as described above.

[0032] The beneficial effects of this invention are as follows: By integrating computer vision, environmental sensing and intelligent modeling technologies, a complete tank safety monitoring system is constructed, which realizes real-time, accurate and automated monitoring of oil and gas concentration and floating roof status. It overcomes the shortcomings of traditional methods in terms of real-time performance, accuracy and environmental adaptability. It has strong engineering applicability and scalability, and provides reliable technical support for tank safety management. Attached Figure Description

[0033] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0034] Figure 1 This is a flowchart of the computer vision-based tank safety monitoring method in this embodiment.

[0035] Figure 2 This is a flowchart of the method for constructing the floating disk analysis model in this embodiment.

[0036] Figure 3 This is a flowchart of the correction method for dividing sample color regions in this embodiment.

[0037] Figure 4 This is a schematic diagram of the structure of the computer vision-based tank safety monitoring device in this embodiment. Detailed Implementation

[0038] The following detailed description, in conjunction with the accompanying drawings and specific embodiments, provides a more comprehensive overview of the computer vision-based tank safety monitoring method, apparatus, and storage medium disclosed in this invention. It should be noted that the technical features or combinations of technical features described in the following embodiments should not be considered isolated; they can be combined to achieve better technical results. In the accompanying drawings of the following embodiments, the same reference numerals in each drawing represent the same features or components, which can be applied to different embodiments. Therefore, once an item is defined in one drawing, it does not need to be further discussed in subsequent drawings.

[0039] It should be noted that the structures, proportions, sizes, etc., illustrated in the accompanying drawings are merely for illustrative purposes and to aid those skilled in the art in understanding and reading the invention. They are not intended to limit the conditions under which the invention can be implemented. Any modifications to the structure, changes in proportions, or adjustments to size, provided they do not affect the effectiveness or purpose of the invention, should fall within the scope of the technical content disclosed in the invention. The scope of the preferred embodiments of the present invention includes other implementations, wherein functions may be performed not in the order stated or discussed, including substantially simultaneously or in reverse order, depending on the functions involved. This should be understood by those skilled in the art to which the embodiments of the present invention pertain.

[0040] Techniques, methods, and apparatus known to those skilled in the art may not be discussed in detail, but where appropriate, such techniques, methods, and apparatus should be considered part of the specification. In all examples shown and discussed herein, any specific values ​​should be interpreted as merely exemplary and not as limitations. Therefore, other examples of exemplary embodiments may have different values.

[0041] In the description of the embodiments of this application, " / " means "or", and "and / or" is used to describe the relationship between related objects, indicating that there can be three relationships. For example, "A and / or B" means: A and B exist alone, B exists alone, and A and B exist simultaneously. In the description of the embodiments of this application, "multiple" refers to two or more embodiments.

[0042] Please see Figure 1 As shown, this embodiment illustrates a computer vision-based tank safety monitoring method, which includes:

[0043] Step S1: Collect sample data from the storage tank. The sample data includes sample images and sample environmental data. The sample images are images of the monitoring positions on the floating roof inside the tank, captured by an explosion-proof industrial camera installed inside the tank. The acquisition frequency is 25 frames per second. The monitoring positions on the floating roof are the locations where the pigment plate is set. The pigment plate adopts a red and yellow dual-color segmented design. It can be pushed to develop color by expanding an expansion material. The expansion material will deform based on the changes in oil and gas concentration inside the storage tank. The sample environmental data includes sample oil and gas concentration, sample temperature, and sample humidity. The unit of sample oil and gas concentration is VOL%, the unit of sample temperature is ℃, and the sample humidity is relative humidity, which is a percentage value. The sample environmental data is collected by sensors installed inside the storage tank.

[0044] Specifically, in step S1 of this embodiment, images and environmental data are collected synchronously by explosion-proof industrial cameras and sensors to ensure the comprehensiveness and real-time nature of the data source, avoid subjective errors and delays in manual inspections, provide highly reliable and multi-dimensional basic data support for subsequent model construction, and improve the automation level and data consistency of the monitoring system.

[0045] Please continue reading. Figure 1 As shown, the computer vision-based tank safety monitoring method further includes:

[0046] Step S2: Perform color region segmentation processing on the sample image to divide the sample color regions, which include sample red regions and sample yellow regions.

[0047] Specifically, in step S2 of this embodiment, the bottom left corner pixel of the sample image is taken as the origin, and the two sides adjacent to the origin are taken as the x-axis and y-axis respectively. The x-axis increases from left to right and the y-axis increases from bottom to top, with a unit length of 1 pixel. A Cartesian coordinate system is established, and the coordinate points represent the position of each pixel in the sample image.

[0048] Specifically, in step S2 of this embodiment, the image change region is extracted based on the sample image and the preset background image. Pixels in the sample image whose grayscale value is greater than that in the preset background image are considered as changed pixels, and the region formed by these changed pixels is considered as the image change region. The preset background image is the image of the monitoring position on the floating plate inside the tank when no color is displayed.

[0049] Specifically, in step S2 of this embodiment, the sample image is converted from the RGB color space to the HSV color space, and the image change area in the sample image is divided into a sample red area and a sample yellow area. The area in the image change area with a red hue is divided into a sample red area, and the area in the image change area with a yellow hue is divided into a sample yellow area.

[0050] Specifically, in step S2 of this embodiment, the color regions in the sample image are precisely divided using image processing technology to identify the red and yellow color regions, thereby effectively distinguishing the color states of different colors, providing a clear and stable visual basis for subsequent feature extraction and state judgment, and enhancing the system's sensitivity to color changes and recognition accuracy.

[0051] Please continue reading. Figure 1 As shown, the computer vision-based tank safety monitoring method further includes:

[0052] Step S3: Extract and quantify the color features of the sample color region to obtain the sample color features, which include the red feature area, the yellow feature area, the red color position, and the yellow color position.

[0053] Specifically, in step S3 of this embodiment, the number of pixels in the red sample area is counted and used as the red feature area; the number of pixels in the yellow sample area is counted and used as the yellow feature area; the coordinates of the center pixel in the red sample area are used as the red color display position, and the coordinates of the center pixel in the yellow sample area are used as the yellow color display position. The x-coordinate of the center pixel is the average of the x-coordinates of the pixels in the calculation area, and the y-coordinate of the center pixel is the average of the y-coordinates of the pixels in the calculation area, rounded to the nearest integer. When there is no corresponding red or yellow sample area, the color display position of the corresponding color is set to (-1, -1).

[0054] Specifically, in step S3 of this embodiment, by extracting and quantifying the area and position features of the color region, visual information is transformed into calculable numerical data, making color changes measurable. This lays the foundation for establishing a mapping relationship between oil and gas concentration and color features, and improves the analyzability of monitoring data and the trainability of the model.

[0055] Please continue reading. Figure 1 As shown, the computer vision-based tank safety monitoring method further includes:

[0056] Step S4: Construct a color oil and gas mapping model based on sample color features and sample environmental data to obtain sample fitting parameters, which include a first fitting coefficient, a second fitting coefficient, and a fitting constant term.

[0057] Specifically, in step S4 of this embodiment, an area-based oil-gas correlation function is constructed based on the red feature area, the yellow feature area, and the sample oil and gas concentration. The area-based oil-gas correlation function is set as: C=k1×(A1+A2)+k2×(A1+A2) 2 +b, where C represents the sample oil and gas concentration, A1 represents the red feature area, A2 represents the yellow feature area, k1 represents the first fitting coefficient, k2 represents the second fitting coefficient, and b represents the fitting constant term. The sample fitting parameters in the area oil and gas correlation function are determined using the least squares method.

[0058] Specifically, in step S4 of this embodiment, a correlation function between area and oil and gas concentration is established to achieve a mathematical mapping from visual features to concentration values, thereby accurately reflecting changes in oil and gas concentration under different operating conditions and providing a stable and reliable estimation method for concentration monitoring.

[0059] Please continue reading. Figure 1 As shown, the computer vision-based tank safety monitoring method further includes:

[0060] Step S5: Construct a floating disk analysis model based on the sample color features to determine the floating disk status.

[0061] Please see Figure 2 As shown, this is a method for constructing a floating roof analysis model, including:

[0062] Step S51: Construct a color development time sequence based on the red and yellow color development positions, and construct an area time sequence based on the red and yellow feature areas.

[0063] Specifically, in step S51 of this embodiment, the red and yellow color-developing positions within the last 30 seconds are extracted and numbered and sorted in chronological order, and the sequence of the red and yellow color-developing positions is used as the color-developing time sequence.

[0064] Specifically, in step S51 of this embodiment, the red and yellow feature areas within the most recent 15 seconds are extracted and numbered and sorted in chronological order, and the sequence of red and yellow feature areas is used as the area time sequence. The numbering used in the color development time sequence and the area time sequence is the same.

[0065] Please continue reading. Figure 2 As shown, the method for constructing the floating roof analysis model further includes:

[0066] Step S52: Determine the timing variation parameters and timing offset parameters based on the color development timing sequence and area timing sequence.

[0067] Specifically, in step S52 of this embodiment, the time series variation parameters are determined based on the area time series, and the expression for the time series variation parameters is: In the formula, F1 represents the time series variation parameter, A1(t) represents the red characteristic area in the area time series, A2(t) represents the yellow characteristic area in the area time series, t represents the time series number, and σ() represents the standard deviation of the data in parentheses.

[0068] Specifically, in step S52 of this embodiment, the time series offset parameter is determined based on the color development time series. The expression of the time series offset parameter is: F2=max(|R1|,|R2|), where F2 represents the time series offset parameter, R1 represents the red correlation coefficient, and R2 represents the yellow correlation coefficient. The red correlation coefficient and the yellow correlation coefficient are obtained by performing linear regression analysis on the horizontal and vertical coordinates of the red and yellow color development positions in the color development time series, respectively, and using the Pearson correlation coefficient as the red correlation coefficient and yellow correlation coefficient corresponding to each color.

[0069] Please continue reading. Figure 2 As shown, the method for constructing the floating roof analysis model further includes:

[0070] Step S53: Determine the floating disk status based on the timing change parameters, timing offset parameters, and area timing.

[0071] Specifically, in step S53 of this embodiment, the floating roof status is determined based on the timing change parameters, timing offset parameters, and area timing. When F1 > 0.2, the floating roof status is determined to be floating hull leakage, and a floating hull leakage alarm is triggered. When F2 ≥ 0.8 and -0.2 ≤ F1 ≤ 0.2, the floating roof status is determined to be floating roof tilt, and a floating roof tilt alarm is triggered. In the remaining cases, the floating roof status is determined to be normal.

[0072] Specifically, in step S5 of this embodiment, the floating table status is determined by the temporal changes of color features to identify abnormal situations such as floating table tilting and leakage. Through the analysis of dynamic data, intelligent diagnosis and early warning of the floating table's operating status are achieved, improving the timeliness and accuracy of floating table safety monitoring.

[0073] Please continue reading. Figure 1 As shown, the computer vision-based tank safety monitoring method further includes:

[0074] Step S6: Construct color compensation coefficients based on sample environment data, and dynamically correct the division of sample color regions based on color compensation coefficients.

[0075] Please see Figure 3 As shown, this is a correction method for dividing sample color regions, including:

[0076] Step S61: Construct color compensation coefficients based on sample environment data.

[0077] Specifically, in step S61 of this embodiment, a color compensation coefficient is constructed based on the sample temperature and sample humidity. The expression for the color compensation coefficient is: Q=1+α1×(T-T0)+α2×(H-H0), where Q represents the color compensation coefficient, α1 represents the temperature weight, α2 represents the humidity weight, T represents the sample temperature, T0 represents the preset storage tank temperature, H represents the sample humidity, and H0 represents the preset storage tank humidity.

[0078] Specifically, in step S61 of this embodiment, T0 = 20℃ and H0 = 0.5 are set as the reference environment. Under the reference environment, oil and gas samples of different concentrations are introduced into the storage tank, and the true value of the unsaturated colorimetric area corresponding to each oil and gas concentration point is recorded. Then, the temperature-changing colorimetric area and the humidity-changing colorimetric area are measured by changing a single variable. The temperature-changing colorimetric area is obtained by keeping the reference environment humidity constant, changing the temperature within the storage tank's operating range (-20℃ to 60℃) by a certain percentage, such as 10℃, and introducing a fixed concentration of oil and gas into the storage tank at each different temperature point. The colorimetric area measured by the system is recorded as the temperature-changing colorimetric area. The humidity-changing colorimetric area is obtained by keeping the reference environment temperature constant, changing the humidity within the possible humidity range (0.3 to 0.9) within the storage tank by a certain percentage, such as 0.05, and introducing a fixed concentration of oil and gas into the storage tank at each different humidity point. The colorimetric area measured by the system is recorded as the humidity-changing colorimetric area.

[0079] Specifically, in step S61 of this embodiment, temperature weight and humidity weight are determined based on the temperature-change color development area and the humidity-change color development area. The temperature weight is the relative change in color development area caused by a unit temperature change. It is calculated by taking the average of the percentage deviation between the temperature-change color development area and the true value of the color development area and the temperature change. The humidity weight is the relative change in the measured value of the color development area caused by a unit humidity change. It is calculated in the same way as the temperature weight.

[0080] Please continue reading. Figure 3 As shown, the correction method for dividing sample color regions further includes:

[0081] Step S62: Correct the saturation of the sample image based on the color compensation coefficient to dynamically correct the division of the sample color region.

[0082] Specifically, in step S62 of this embodiment, the saturation of the sample image is corrected according to the color compensation coefficient. The saturation of the corrected sample image is set to P, P = p × Q, where p represents the saturation of the sample image before correction. The saturation of the corrected sample image is used to divide the sample red area and the sample yellow area to dynamically correct the division of the sample color area.

[0083] Specifically, in step S6 of this embodiment, environmental data is used to dynamically compensate for color recognition, thereby eliminating the interference of temperature and humidity changes on color judgment. This enhances the stability and robustness of the system in complex environments, ensuring the consistency and reliability of color recognition results.

[0084] Please continue reading. Figure 1 As shown, the computer vision-based tank safety monitoring method further includes:

[0085] Step S7: Collect monitoring data from the storage tank, and determine the oil and gas concentration and floating roof status based on the monitoring data, sample fitting parameters, and floating roof analysis model. The monitoring data includes monitoring images and monitoring environmental data. The monitoring environmental data includes monitoring temperature and monitoring humidity. The unit of the monitoring temperature is ℃, and the monitoring humidity is relative humidity, which is a percentage value. The monitoring images are the same as the sample images, and the method of collecting the monitoring data is the same as the method of collecting the sample data.

[0086] Specifically, in step S7 of this embodiment, the monitoring image is segmented by color region to divide the monitoring color region, and the color features of the monitoring color region are extracted and quantified to obtain the monitoring color features. The red feature area and yellow feature area in the monitoring color features are input into the area oil and gas correlation function to obtain the oil and gas concentration. The monitoring color features are input into the floating disk analysis model to obtain the floating disk status.

[0087] Specifically, in step S7 of this embodiment, the monitoring image is used as a sample image, and the monitoring image is analyzed using the process of steps S1-S6 above, so as to determine the oil and gas concentration and the floating roof status.

[0088] Specifically, in this embodiment, when performing analysis and calculation, the unit is limited only to the order of magnitude of the data analysis, and the analysis process is numerical analysis without considering the dimensions between the data.

[0089] Please see Figure 4 As shown, this is a computer vision-based tank safety monitoring device according to this embodiment, including:

[0090] The sample acquisition module is used to collect sample data from the storage tank.

[0091] The region segmentation module is used to perform color region segmentation processing on the sample image to divide the sample color regions;

[0092] The feature extraction module is used to extract and quantify the color features of the sample color region to obtain the sample color features;

[0093] The fitting analysis module is used to construct a color oil and gas mapping model based on sample color features and sample environmental data in order to obtain sample fitting parameters;

[0094] The floating disk analysis module is used to build a floating disk analysis model based on the color features of the samples in order to determine the status of the floating disk.

[0095] The segmentation and correction module is used to construct color compensation coefficients based on sample environmental data and to dynamically correct the segmentation of sample color regions based on the color compensation coefficients.

[0096] The monitoring and analysis module is used to collect monitoring data from the storage tank and determine the oil and gas concentration and floating roof status based on the monitoring data, sample fitting parameters, and floating roof analysis model.

[0097] It will be understood by those skilled in the art that all or some of the steps and systems in the methods disclosed above can be implemented as software, firmware, hardware, and suitable combinations thereof. Some or all of the physical components can be implemented as software executed by a processor, such as a central processing unit, digital signal processor, or microprocessor, or as hardware, or as an integrated circuit, such as an application-specific integrated circuit. Such software can be distributed on a computer-readable medium, which can include computer storage media (or non-transitory media) and communication media (or transient media). As is known to those skilled in the art, the term computer storage media includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information (such as computer-readable programs, data structures, program modules, or other data). Computer storage media includes, but is not limited to, RAM, ROM, EEPROM, flash memory or other memory technologies, CD-ROM, digital versatile disc (DVD) or other optical disc storage, magnetic cartridges, magnetic tape, disk storage or other magnetic storage devices, or any other medium that can be used to store desired information and is accessible to a computer. Furthermore, as is known to those skilled in the art, communication media typically contain computer-readable programs, data structures, program modules, or other data in modulated data signals such as carrier waves or other transmission mechanisms, and may include any information delivery medium.

[0098] In the above description, the disclosure of this invention is not intended to limit itself to these aspects. Rather, within the scope of the objectives of this disclosure, components can be selectively and operationally combined in any number. Furthermore, terms such as “comprising,” “encompassing,” and “having” should be interpreted by default as inclusive or open-ended, rather than exclusive or closed, unless explicitly defined as such. All technical, scientific, or other terms are to be understood by those skilled in the art, unless defined as such. Public terms found in dictionaries should not be interpreted in the context of the relevant technical documents in an overly idealistic or impractical manner, unless explicitly defined as such in this disclosure. Any modifications or alterations made by those skilled in the art based on the foregoing disclosure are within the scope of the claims.

Claims

1. A computer vision-based method for monitoring the safety of storage tanks, characterized in that, include: Perform color region segmentation on the sample image to divide the sample color regions; The sample image is an image of the monitoring position on the floating roof inside the tank, taken by an explosion-proof industrial camera device installed inside the tank. The monitoring position on the floating roof is the setting position of the pigment plate. The pigment plate adopts a red and yellow two-color segmented design. The pigment plate can be pushed to develop color by expanding the expansion material. The expansion material will deform according to the change of oil and gas concentration inside the tank. Color features are extracted and quantized from the sample color regions to obtain the sample color features; A color oil and gas mapping model is constructed based on sample color features and sample environmental data to obtain sample fitting parameters; A floating disk analysis model is constructed based on the color features of the samples to determine the status of the floating disk; A color compensation coefficient is constructed based on the sample environment data, and the division of the sample color region is dynamically corrected based on the color compensation coefficient. Collect monitoring data from the storage tanks, and determine the oil and gas concentration and floating roof status based on the monitoring data, sample fitting parameters, and floating roof analysis model; According to the red characteristic area, the yellow characteristic area and the sample oil and gas concentration, an area oil and gas correlation function is constructed, and the area oil and gas correlation function is set as: C=k1×(A1+A2)+k2×(A1+A2) 2 +b, wherein C represents the sample oil and gas concentration, A1 represents the red characteristic area, A2 represents the yellow characteristic area, k1 represents a first fitting coefficient, k2 represents a second fitting coefficient, and b represents a fitting constant term; the sample fitting parameters in the area oil and gas correlation function are determined by using a least square method; Extract the red and yellow color-changing positions within the last 30 seconds and number and sort them in chronological order. Use the sequence of red and yellow color-changing positions as the color-changing time sequence. Extract the red and yellow feature areas within the last 15 seconds and number and sort them in chronological order. Use the sequence of red and yellow feature areas as the area time series. The time series variation parameters are determined based on the area time series, and the expressions for the time series variation parameters are as follows: In the formula, F1 represents the time-series variation parameter, A1(t) represents the red characteristic area in the area time-series, A2(t) represents the yellow characteristic area in the area time-series, t represents the time-series number, and the time-series offset parameter is determined based on the color development time-series. The expression of the time-series offset parameter is: F2=max(|R1|,|R2|), where F2 represents the time-series offset parameter, R1 represents the red correlation coefficient, and R2 represents the yellow correlation coefficient. The floating platform status is determined based on the time-series variation parameter, the time-series offset parameter, and the area time-series. When F1>0.2, the floating platform status is determined to be floating hull leakage, and a floating hull leakage alarm is triggered. When F2≥0.8 and -0.2≤F1≤0.2, the floating platform status is determined to be floating platform tilt, and a floating platform tilt alarm is triggered. In other cases, the floating platform status is determined to be normal.

2. The computer vision-based tank safety monitoring method according to claim 1, characterized in that, Collect sample data from the storage tank, extract the image change area based on the sample image and the preset background image, take the pixels in the sample image whose gray value is greater than the pixels in the preset background image as the change pixels, and take the area formed by the change pixels as the image change area. The sample image is converted from the RGB color space to the HSV color space, and the image variation area in the sample image is divided into sample red area and sample yellow area. The area with red hue in the image variation area is divided into sample red area, and the area with yellow hue in the image variation area is divided into sample yellow area.

3. The computer vision-based tank safety monitoring method according to claim 2, characterized in that, The number of pixels in the red region of the sample is counted and used as the red feature area. The number of pixels in the yellow region of the sample is counted and used as the yellow feature area. The coordinates of the center pixel of the red region of the sample are used as the red coloring position, and the coordinates of the center pixel of the yellow region of the sample are used as the yellow coloring position.

4. The computer vision-based tank safety monitoring method according to claim 3, characterized in that, A color compensation coefficient is constructed based on the sample temperature and sample humidity. The expression for the color compensation coefficient is: Q=1+α1×(T-T0)+α2×(H-H0), where Q represents the color compensation coefficient, α1 represents the temperature weight, α2 represents the humidity weight, T represents the sample temperature, T0 represents the preset storage tank temperature, H represents the sample humidity, and H0 represents the preset storage tank humidity. The saturation of the sample image is corrected based on the color compensation coefficient. The saturation of the corrected sample image is set as P, where P = p × Q, and p represents the saturation of the sample image before correction. The saturation of the corrected sample image is used to divide the sample red area and the sample yellow area to dynamically correct the division of the sample color area.

5. The computer vision-based tank safety monitoring method according to claim 4, characterized in that, The monitoring image is segmented by color region to divide the monitoring color region. Color features are extracted and quantified from the monitoring color region to obtain the monitoring color features. The red and yellow feature areas in the monitoring color features are input into the area oil and gas correlation function to obtain the oil and gas concentration. The monitoring color features are input into the floating disk analysis model to obtain the floating disk status.

6. A computer vision-based tank safety monitoring device, applied to the computer vision-based tank safety monitoring method as described in any one of claims 1-5, characterized in that, include: The sample acquisition module is used to collect sample data from the storage tank. The region segmentation module is used to perform color region segmentation processing on the sample image to divide the sample color regions; The feature extraction module is used to extract and quantify the color features of the sample color region to obtain the sample color features; The fitting analysis module is used to construct a color oil and gas mapping model based on sample color features and sample environmental data in order to obtain sample fitting parameters; The floating disk analysis module is used to build a floating disk analysis model based on the color features of the samples in order to determine the status of the floating disk. The segmentation and correction module is used to construct color compensation coefficients based on sample environmental data and to dynamically correct the segmentation of sample color regions based on the color compensation coefficients. The monitoring and analysis module is used to collect monitoring data from the storage tank and determine the oil and gas concentration and floating roof status based on the monitoring data, sample fitting parameters, and floating roof analysis model.

7. A storage medium, characterized in that, The device stores instructions that, when executed on a computer, cause the computer to perform the computer vision-based tank safety monitoring method as described in any one of claims 1-5.

Images