Method for detecting external floating roof oil tank and calculating oil reserve based on SAR (Synthetic Aperture Radar) image

By analyzing SAR images and detecting deep convolutional neural networks, combining satellite imaging parameter information, the geometric scale and oil storage volume of the oil tank target are automatically calculated, which solves the problems of low detection efficiency and insufficient accuracy in the existing technology, and realizes efficient and automatic oil tank target detection and reserve calculation.

CN120065158APending Publication Date: 2025-05-30PLA PEOPLES LIBERATION ARMY OF CHINA STRATEGIC SUPPORT FORCE AEROSPACE ENG UNIV
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202510116309.5
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2025-01-24
Publication Date
2025-05-30

AI Technical Summary

Technical Problem

The existing oil tank target reserve calculation method based on high-resolution SAR images has problems such as low detection efficiency, manpower dependence, and difficulty in processing massive remote sensing data. The method based on image threshold segmentation has insufficient applicability and accuracy.

Method used

Through the analysis of the imaging relationship of SAR images and the scattering structure of the oil tank target, the geometric scale of the oil tank target is determined, and the deep convolutional neural network is used to detect the oil tank target, and the oil storage volume of the oil tank target is automatically calculated based on the satellite imaging parameter information.

Benefits of technology

The oil tank target detection positioning and capacity estimation based on single SAR images is realized, with the advantages of simplicity of calculation, high accuracy and good adaptability, and can automatically detect and calculate the reserves of the oil tank target.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN120065158A_ABST
    Figure CN120065158A_ABST
Patent Text Reader

Abstract

The invention discloses an external floating-roof oil tank detection and oil storage quantity calculation method based on an SAR image. The method comprises the following steps that the SAR image containing a target area of a cylindrical external floating-roof oil tank to be detected and meta-file parameter information during satellite imaging are input; preprocessing the image; detecting a single oil tank target by using the oil tank target detection network; determining image coordinates of key scattering points of the oil tank target based on the image slice of the single oil tank target; calculating the oil storage quantity of the single external floating roof oil tank target by utilizing the geometric parameters of the oil tank target based on the imaging geometric relationship of the image and the image coordinates of the key scattering points; and calculating the oil storage quantity of all N oil tank targets in the region of interest. According to the invention, oil tank target detection positioning and capacity estimation based on a single SAR image are realized; in combination with parameter information of satellite imaging, the electromagnetic scattering characteristics of the oil tank target are fully utilized, and parameter information related to the geometric scale of the target is extracted from key scattering points of the oil tank target.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of remote sensing image processing, and in particular, to a method for detecting an external floating roof oil tank based on SAR images and calculating the oil storage volume. Background Art

[0002] An oil tank is a container for storing petroleum and its products. As an indispensable facility for strategic petroleum reserves, it has an important position in the fields of safety monitoring, disaster prevention, intelligence analysis, etc. An external floating roof oil tank is composed of a floating roof floating on the surface of the medium and the side wall of the tank body. The floating roof can rise and fall with the increase and decrease of the storage medium in the tank. An annular sealing device is provided between its outer edge and the side wall. In this way, the medium in the tank is in direct contact with the floating roof, and no air layer will be generated between them, which can greatly reduce the evaporation loss of oil products. It has the characteristics of small storage oil product loss and small environmental pollution of oil and gas to the environment. It is a commonly used device for storing petroleum products and is widely used in various petroleum depots.

[0003] Synthetic Aperture Radar (SAR) uses active microwave imaging and has unique advantages of all-weather and all-time. It is an important means to realize target reconnaissance and surveillance. In recent decades, SAR data acquisition equipment and imaging technology have developed rapidly, and the quantity and quality of the acquired images have been greatly improved. Especially at present, the image resolution has reached the sub-meter level, providing a data basis for SAR target detection and information extraction. This makes it a feasible method to extract oil tank targets from SAR images and analyze the oil reserves.

[0004] Most of the existing methods for calculating the oil storage volume of oil tank targets based on high-resolution SAR images adopt two methods: the method based on visual interpretation. This method has high accuracy in detecting oil tank targets and estimating the oil storage volume, but the detection efficiency is low, highly dependent on manpower, difficult to be applied on a large scale, difficult to process a large amount of remote sensing data, and unable to be popularized and applied on a large scale. The method based on image threshold segmentation can realize the automatic detection and recognition of oil tank targets. However, for different images, the imaging angle and the scattering intensity of the target change during imaging, resulting in different applicable segmentation thresholds. This makes it difficult to select a unified segmentation threshold for the detection and segmentation of oil tank targets, and the accuracy of estimating the oil storage volume of the oil tank is not high. Summary of the Invention

[0005] The purpose of the present invention is to provide a method for detecting an external floating roof oil tank based on SAR images and calculating the oil storage volume, so as to solve the foregoing problems existing in the prior art.

[0006] By analyzing the imaging relationship of the SAR image and the scattering structure of the oil tank target in the image, the geometric scale of the oil tank target is determined, and finally the goal of automatically calculating the oil storage volume of the oil tank target is achieved.

[0007] To achieve the above object, the technical solution adopted by the present invention is as follows:

[0008] A method for detecting an external floating roof oil tank and calculating the oil storage capacity based on SAR images, comprising the following steps:

[0009] S1. Input the SAR image containing the target area of the cylindrical external floating roof oil tank to be detected and the metadata parameter information during satellite imaging;

[0010] S2. Preprocess the image: mainly weaken the influence of speckle by image filtering; rotate the image so that from left to right corresponds to the range direction of radar imaging, and from top to bottom corresponds to the azimuth direction of imaging;

[0011] S3. Detect a single oil tank target using an oil tank target detection network;

[0012] S4. Based on the image slices of a single oil tank target, determine the image coordinates of the key scattering points of the oil tank target;

[0013] S5.

[0014] Based on the imaging geometric relationship of the image and the image coordinates of the key scattering points, determine the geometric parameters of the oil tank radius R, the height H of the external floating roof oil tank, and the height difference h between the floating roof plane of the external floating roof oil tank and the rear side wall of the oil tank;

[0015] S6. According to the volume calculation formula of a cylinder, use the geometric parameters of the oil tank target to calculate the oil storage capacity of a single external floating roof oil tank target;

[0016] S7. Calculate the oil storage capacity of all N oil tank targets in the area of interest.

[0017] Preferably, the metadata parameter information during satellite imaging specifically includes the orbital height L of the satellite, the imaging pitch angle θ, the incident angle β relative to the target during radar imaging obtained by conversion, and the range direction resolution δ r m / pixel, and the azimuth direction resolution δ a m / pixel.

[0018] Preferably, step S3 is specifically: detecting the oil tank target based on an oil tank target detection model based on a deep convolutional neural network to determine all oil tank targets in the image, and using a rectangular bounding box to locate the position and image range of each single oil tank in the image.

[0019] Preferably, in step S4, the key scattering points of the oil tank target specifically include: the front top P1, the front bottom P2, the floating roof P3, the rear top P4, the left side wall P5, and the right side wall P6 of the external floating roof oil tank.

[0020] Preferably,

[0021] The radius R of the oil tank = 0.5 * (P6 - P5) * δ a ;

[0022] The height H of the external floating roof oil tank = tanβ * (P2 - P1) * δ r ;

[0023] The height difference h between the floating roof plane of the external floating roof oil tank and the rear side wall of the oil tank = tanβ * (P4 - P3) * δ r ;

[0024] Subtract the height H of the external floating roof oil tank and the height difference h between the floating roof plane of the external floating roof oil tank and the rear side wall of the oil tank to obtain the height of the oil plane when the oil tank stores oil.

[0025] Preferably, the volume calculation formula of the cylinder is: V = πR 2 (H - h).

[0026] Preferably, the calculation formula of the oil storage capacity is:

[0027] Among them, the oil storage capacity of the i-th oil tank target is V i .

[0028] The beneficial effects of the present invention are:

[0029] (1) The present invention realizes the detection, positioning and capacity estimation of oil tank targets based on a single SAR image;

[0030] (2) The present invention makes full use of the electromagnetic scattering characteristics of oil tank targets, and extracts parameter information related to the geometric scale of the targets from the key scattering points of the oil tank targets;

[0031] (3) The present invention designs a method for extracting the position parameters of key scattering points based on local maxima, and quickly locates and extracts the parameter information of the scattering points;

[0032] (4) The present invention can automatically detect oil tank targets and calculate the oil storage of the oil tank targets, and has the advantages of simple calculation, high accuracy and good adaptability, meeting the application requirements. Description of the Drawings

[0033] Figure 1 is the flow chart of the detection of cylindrical external floating roof oil tank targets and the calculation of oil storage capacity based on a single SAR image of the present invention;

[0034] Figure 2 is the SAR image of the area to be detected of the present invention;

[0035] Figure 3 is the relationship diagram between the satellite and the imaging viewing angle θ and the ground incident angle β when the radar platform of the present invention images the targets on the earth's surface;

[0036] Figure 4 is the SAR image after preprocessing of the present invention;

[0037] Figure 5 is the schematic diagram of radar electromagnetic scattering of the external floating roof oil tank of the present invention;

[0038] Figure 6 is the schematic diagram of radar electromagnetic scattering of another embodiment of the external floating roof oil tank of the present invention;

[0039] Figure 7 is the schematic diagram of the detection result of the key scattering points of the external floating roof oil tank target of the present invention.

[0040] Figure 8 is the flowchart of a method for detecting an external floating roof oil tank and calculating the oil storage capacity based on SAR images of the present invention. Detailed implementation manners

[0041] In order to make the objectives, technical solutions and advantages of the present invention clearer and more understandable, the present invention will be further described in detail below with reference to the accompanying drawings. It should be understood that the specific implementation manners described herein are only used to explain the present invention and are not used to limit the present invention.

[0042] For the detection of a cylindrical external floating roof oil tank and the calculation of its oil storage capacity, the present invention uses a deep convolutional network to detect the oil tank target, and then based on the parameter information of SAR satellite imaging, comprehensively considers the electromagnetic scattering mechanism of the oil tank target in the image and the electromagnetic scattering characteristics of the key points, obtains the geometric dimension information of the oil tank target, and further calculates the oil storage capacity of the oil tank target.

[0043] Refer to Figures 1 to 8 The method for detecting an external floating roof oil tank and calculating the oil storage capacity based on SAR images described above includes the following steps:

[0044] S1. Input the SAR image containing the target area of the cylindrical external floating roof oil tank to be detected and the meta-file parameter information during satellite imaging.

[0045] In this step, given the SAR image data of the area of interest (as Figure 2 shown), and the meta-file parameter information during satellite imaging, specifically including the orbital altitude L = 500 KM of the satellite, the radar viewing angle θ = 35.1181° during satellite imaging, the range resolution δ r = 0.5 m / pixel, and the azimuth resolution δ a = 0.5 m / pixel.

[0046] Appendix Figure 3For the geometric relationship of satellite imaging, where the satellite imaging angle θ is the angle between the irradiation direction of the satellite radar wave and the straight line connecting the satellite and the center of the earth. θ is the satellite imaging angle and can be read from the auxiliary file in the SAR image. The angle between the radar wave and the normal of the surface target during imaging is the incident angle β. From the imaging geometric relationship, in triangle △ABO, according to the sine theorem of the triangle, it can be known that the incident angle β and the radar angle θ satisfy the following relational expression:

[0047]

[0048] Therefore, there is

[0049]

[0050] Among them, L is the orbital height of the satellite, and the value here is L = 500 Km, r is the radius of the earth, r = 6371 Km.

[0051] Therefore, β = 38.3461, that is, the converted imaging incident angle is 38.3461°.

[0052] S2. Preprocess the image: mainly reduce the influence of speckle through image filtering; rotate the image so that the image corresponds to the range direction of radar imaging from left to right and the azimuth direction of imaging from top to bottom. The preprocessed image is as shown in Figure 4 shown.

[0053] S3. Detect a single oil tank target using an oil tank target detection network.

[0054] Perform oil tank target detection based on the oil tank target detection model of the deep convolutional neural network to determine all oil tank targets in the image, and use a rectangular bounding box to locate the position and image range of each single oil tank in the image.

[0055] When performing oil tank target detection, first divide the input image of any size into manageable slice images (600×600 pixels). Each slice image is obtained by partitioning through a sliding window with a certain overlap (default is 10%). Subsequently, use the oil tank target detector to detect each slice image, obtain the oil tank targets in it, and integrate the oil tank target detection results in each slice image into the original input image to determine the positions of the oil tank targets.

[0056] S4. Based on the image slice of a single oil tank target, determine the image coordinates of the key scattering points of the oil tank target.

[0057] The backscattering in the external floating roof oil tank target is as shown in Figure 6As shown in the figure, the light blue in the figure is the primary reflection imaging of the front edge of the oil tank top, and the imaging point is P1 in the image; the red is the secondary reflection imaging of the oil tank side wall and bottom integrated, and the imaging point is P2 in the image; the purple is the secondary reflection of the rear side wall of the oil tank and the floating roof plane, and the imaging point is P3 in the image; the green is the primary reflection imaging of the rear part of the tank top, and the imaging point is P4 in the image.

[0058] S5. Based on the imaging geometric relationship of the image and the image coordinates of the key scattering points, determine the geometric parameters of the oil tank radius R, the height H of the external floating roof oil tank, and the height difference h between the floating roof plane of the external floating roof oil tank and the rear side wall of the oil tank.

[0059] As Figure 5 shown, tanβ = H / ((P2 - P1) * δ r );

[0060] Calculate the height H of the external floating roof oil tank = tanβ * (P2 - P1) * δ r = 18.5108m;

[0061] When SAR images a cylindrical oil tank, due to the influence of multipath scattering, the curvature radius of the cylindrical tank body is greater than the oil tank radius when imaging the cylindrical tank body. Therefore, the cylindrical oil tank becomes an ellipse when imaging. So when measuring the diameter of the oil tank, it is necessary to measure the inner distance between the two walls of the oil tank in the azimuth direction, that is, characterized by the distance between two points P5 on the left side wall and P6 on the right side wall (see Figure 7 ).

[0062] The oil tank radius R = 0.5 * (P6 - P5) * δ a = 39.9m;

[0063] The height difference h between the floating roof plane of the external floating roof oil tank and the rear side wall of the oil tank = tanβ * (P4 - P3) * δ r = 4.9837m;

[0064] S6. According to the volume calculation formula of the cylinder, use the geometric parameters of the oil tank target to calculate the oil storage capacity of a single external floating roof oil tank target.

[0065] S7. Calculate the oil storage capacity of all N oil tank targets in the area of interest.

[0066] Preferably, the metadata parameter information during satellite imaging specifically includes the orbital height L of the satellite, the imaging pitch angle θ, the incident angle β relative to the target obtained by conversion during radar imaging, the range resolution δ r m / pixel, and the azimuth resolution δ a m / pixel.

[0067] Preferably, step S3 is specifically: performing oil tank target detection based on an oil tank target detection model of a deep convolutional neural network to determine all oil tank targets in the image, and using a rectangular bounding box to locate the position and image range of each individual oil tank in the image.

[0068] Preferably, in step S4, the key scattering points of the oil tank target specifically include: the front top P1, the front bottom P2, the floating roof P3, the rear top P4, the left side wall P5, and the right side wall P6 of the external floating roof oil tank.

[0069] Preferably, the geometric parameters of the outer side wall of the oil tank target specifically include: the height H, the radius R, and the oil storage height h of the outer side wall of the oil tank target;

[0070] Among them, the oil tank radius R = 0.5 * (P6 - P5) * δ a ;

[0071] The height H of the external floating roof oil tank = tanβ * (P2 - P1) * δ r ;

[0072] The height difference h between the floating roof plane of the external floating roof oil tank and the rear side wall of the oil tank = tanβ * (P4 - P3) * δ r ;

[0073] Subtract the height H of the external floating roof oil tank from the height difference h between the floating roof plane of the external floating roof oil tank and the rear side wall of the oil tank to obtain the height of the oil plane when the oil tank stores oil.

[0074] Preferably, the volume calculation formula of the cylinder is: V = πR 2 (H - h).

[0075] In this embodiment, subtracting (H - h) from both H and h is the height of the oil plane when the oil tank stores oil. Finally, the oil storage capacity of the external floating roof oil tank is calculated according to the volume of the cylinder.

[0076] V = πR 2 (H - h) = 67655m 3

[0077] If the floating plane of the external floating roof oil tank rises to the top, at this time the height difference h between the floating plane of the external floating roof oil tank and the rear side wall is 0, then the oil storage capacity of the oil tank at this time is V max = πR 2 H = 92581m 3

[0078] Taking the integer of the calculated oil tank volume above, it can be known that the maximum oil storage capacity of a single tank of this cylindrical external floating roof oil tank is about 100,000 cubic meters.

[0079] Step Seven: Calculate the oil storage capacity of all oil tank targets in the area of interest.

[0080] Preferably, the calculation formula for the oil storage capacity is as follows:

[0081] where the oil storage capacity of the i-th oil tank target is V i . Thus, the oil storage capacity of the oil tanks in the area of interest can be calculated and determined.

[0082] By adopting the above technical solutions disclosed in the present invention, the following beneficial effects are obtained:

[0083] (1) The present invention realizes the detection, positioning and capacity estimation of oil tank targets based on a single SAR image;

[0084] (2) The present invention makes full use of the electromagnetic scattering characteristics of oil tank targets, and extracts parameter information related to the geometric scale of the targets from the key scattering points of the oil tank targets;

[0085] (3) The present invention designs a method for extracting scattering point position parameters, and quickly locates and extracts the parameter information of the scattering points;

[0086] (4) The present invention can automatically detect oil tank targets and calculate the oil storage of the oil tank targets, and has the advantages of simple calculation, high accuracy and good adaptability, meeting the application requirements.

[0087] The above are only the preferred embodiments of the present invention. It should be noted that for those of ordinary skill in the art, without departing from the principle of the present invention, several improvements and refinements can be made, and these improvements and refinements should also be regarded as the protection scope of the present invention.

Claims

1. A method for detecting and calculating the oil storage capacity of an external floating roof tank based on SAR images, characterized in that: The following steps are involved: S1, inputting the SAR image of the target area of ​​the cylindrical external floating roof tank to be detected and the metafile parameter information of the satellite imaging; S2. Preprocess the image: mainly reduce the influence of coherent speckle by image filtering; rotate the image so that the image from left to right corresponds to the range direction of radar imaging, and from top to bottom corresponds to the azimuth direction of imaging; S3, using the oil tank target detection network to detect a single oil tank target; S4, based on the image slices of a single oil tank target, determining the image coordinates of key scattering points of the oil tank target; S5. Based on the imaging geometric relationship of the image and the image coordinates of the key scattering points, determine the geometric parameters of the oil tank radius R, the height H of the external floating roof oil tank, and the height difference h between the floating roof plane of the external floating roof oil tank and the rear side wall of the oil tank; S6. Calculate the oil storage capacity of a single external floating roof tank target according to a volume calculation formula of a cylinder and using geometric parameters of the tank target; S7. Calculate the oil storage capacity of all N oil tank targets in the area of ​​interest.

2. The method for detecting and calculating the oil storage capacity of an external floating roof tank based on SAR images according to claim 1 is characterized in that: The metafile parameter information during satellite imaging specifically includes the satellite's orbital height L, the imaging pitch angle θ, the incident angle β relative to the target during radar imaging, and the image range resolution δ r Meter / pixel, azimuth resolution δ a Meters / pixel.

3. The method for detecting and calculating the oil storage capacity of an external floating roof tank based on SAR images according to claim 2 is characterized in that: The step S3 specifically includes: performing oil tank target detection based on an oil tank target detection model of a deep convolutional neural network to determine all oil tank targets in the image, and using a rectangular envelope frame to locate the position and image range of each single oil tank in the image.

4. The method for detecting and calculating the oil storage capacity of an external floating roof tank based on SAR images according to claim 3 is characterized in that: In the step S4, the key scattering points of the oil tank target specifically include: the front top P1, the front bottom P2, the floating top P3, the rear top P4, the left side wall P5 and the right side wall P6 of the external floating roof oil tank.

5. The method for detecting and calculating the oil storage capacity of an external floating roof tank based on SAR images according to claim 4 is characterized in that: The oil tank radius R = 0.5*(P6-P5)*δ a ; The height of the external floating roof tank H = tanβ*(P2-P1)*δ r ; The height difference between the floating roof plane and the rear side wall of the external floating roof tank is h=tanβ*(P4-P3)*δ r ; The height H of the external floating roof oil tank and the height difference h between the floating roof plane of the external floating roof oil tank and the rear side wall of the oil tank are subtracted to obtain the height of the oil level when the oil tank is storing oil.

6. The method for detecting and calculating the oil storage capacity of an external floating roof tank based on SAR images according to claim 5 is characterized in that: The volume calculation formula of the cylinder is: V = πR 2 (Hh).

7. The method for detecting and calculating the oil storage capacity of an external floating roof tank based on SAR images according to claim 6 is characterized in that: The calculation formula of the oil storage capacity is: The oil storage capacity of the i-th oil tank target is V i .