Oil tank layout information generation method, device, electronic device, and medium

Abstract

The embodiments of the present disclosure disclose a method, device, electronic device and medium for generating oil tank layout information. A specific implementation of the method includes: performing image cropping on the target synthetic aperture radar image to obtain a sub-image set; for each area information in the area information set corresponding to the sub-image set, obtaining an optical image set; for each optical image set in the optical image set group, inputting each optical image in the optical image set into the oil tank feature information extraction network to obtain a feature atlas set; for each feature atlas set in the feature atlas set group, performing feature map fusion on each feature map in the feature atlas set to obtain a fused feature map; for each fused feature map in the obtained fused feature atlas set, generating the position information of each first oil tank information in the first oil tank information set and the radius information of each first oil tank information; generating oil tank layout information. This implementation can generate oil tank layout information accurately and efficiently.

Description

Technical Field

[0001] Embodiments of the present disclosure relate to the field of computer technology, and in particular to a method, device, electronic device, and medium for generating oil tank layout information. Background Art

[0002] At present, in daily life, oil tanks are often large containers with regular shapes (e.g., cylindrical) used to store oil products. Target detection and parameter extraction of oil tanks are of great significance to application fields such as oil tank monitoring and oil storage analysis. For target detection of oil tanks, the usual method is: first, obtain the synthetic aperture radar image of the target to be detected. Then, the target synthetic aperture radar image is input into the pre-trained convolutional neural network (CNN) to perform target detection on the oil tank.

[0003] However, when the above method is used to detect oil tanks, the following technical problems often occur:

[0004] First, although synthetic aperture radar images can effectively reflect the phase information of each oil tank in the target area, using synthetic aperture radar images as input to the neural network for oil tank detection often has the problem of low accuracy. In addition, using high-resolution target synthetic aperture radar images as input to the neural network has the problem of high computational complexity in training the neural network and subsequent oil tank detection in the target area.

[0005] Third, using convolutional neural networks alone to detect oil tanks in target areas often results in low accuracy. For example, convolutional neural networks often fail to consider some of the main feature information in oil tank images (e.g., structure type, shadow area, etc.), resulting in low accuracy in oil tank detection. Summary of the invention

[0006] The content of this disclosure is used to introduce concepts in a brief form, which will be described in detail in the detailed implementation section below. The content of this disclosure is not intended to identify the key features or essential features of the technical solution claimed for protection, nor is it intended to limit the scope of the technical solution claimed for protection.

[0007] Some embodiments of the present disclosure propose a method, device, electronic device and computer-readable medium for generating oil tank layout information to solve one or more of the technical problems mentioned in the above background technology section.

[0008] In a first aspect, some embodiments of the present disclosure provide a method for generating oil tank layout information, comprising: performing image cropping on a pre-acquired target synthetic aperture radar image to obtain a sub-image set, wherein the target synthetic aperture radar image is an image taken by a synthetic aperture radar sensor for a target area and whose resolution size satisfies a first condition; for each area information in an area information set corresponding to the sub-image set, obtaining an optical image set taken by an optical image sensor for the area corresponding to the area information and whose resolution size satisfies different second conditions; for each optical image set in the obtained optical image set group, respectively inputting each optical image in the optical image set into a pre-trained oil tank feature information extraction network to obtain a feature atlas. ; For each feature atlas in the obtained feature atlas group, perform feature map fusion on each feature map in the above feature atlas to obtain a fused feature map; for each fused feature map in the obtained fused feature map, generate the position information of each first oil tank information in the first oil tank information set corresponding to the above fused feature map and the radius information of each first oil tank information in the above first oil tank information set according to the above fused feature map and the oil tank recognition model set; generate the oil tank layout information of the above target area according to the above target synthetic aperture radar image, the above oil tank recognition model set, the position information of each first oil tank information in the obtained first oil tank information set group and the radius information of each first oil tank information in the above first oil tank information set group.

[0009] In a second aspect, some embodiments of the present disclosure provide an oil tank layout information generating device, comprising: a cropping unit, configured to perform image cropping on a pre-acquired target synthetic aperture radar image to obtain a sub-image set, wherein the target synthetic aperture radar image is an image shot by a synthetic aperture radar sensor for a target area, and the resolution size satisfies a first condition; an acquisition unit, configured to acquire, for each area information in an area information set corresponding to the sub-image set, an optical image set shot by an optical image sensor for an area corresponding to the area information, and the resolution size satisfies different second conditions; an input unit, configured to input, for each optical image set in the obtained optical image set group, each optical image in the optical image set into a pre-trained oil tank feature information extraction network, to obtain a feature atlas; and a fusion unit. The combining unit is configured to fuse the feature maps in each feature map set in the obtained feature map set group to obtain a fused feature map; the first generating unit is configured to generate, for each fused feature map in the obtained fused feature map set, the position information of each first oil tank information in the first oil tank information set corresponding to the fused feature map and the radius information of each first oil tank information in the first oil tank information set according to the fused feature map and the oil tank recognition model set; the second generating unit is configured to generate the oil tank layout information of the above-mentioned target area according to the above-mentioned target synthetic aperture radar image, the above-mentioned oil tank recognition model set, the position information of each first oil tank information in the obtained first oil tank information set group and the radius information of each first oil tank information in the above-mentioned first oil tank information set group.

[0010] In a third aspect, some embodiments of the present disclosure provide an electronic device comprising: one or more processors; a storage device on which one or more programs are stored, and when the one or more programs are executed by the one or more processors, the one or more processors implement the method described in any implementation manner in the first aspect.

[0011] In a fourth aspect, some embodiments of the present disclosure provide a computer-readable medium having a computer program stored thereon, wherein when the program is executed by a processor, the method described in any implementation manner in the first aspect is implemented.

[0012] The above-mentioned embodiments of the present disclosure have the following beneficial effects: the oil tank layout information generation method of some embodiments of the present disclosure accurately and efficiently generates the oil tank layout information of the target area. Specifically, the reasons for the inaccurate and ineffective generation of the oil tank layout information are: first, the synthetic aperture radar image is used as the input of the neural network to detect the oil tank, which often has the problem of low accuracy. In addition, the high-resolution target synthetic aperture radar image is used as the input of the neural network, and there is a problem of large amount of calculation for training the convolutional neural network and the subsequent oil tank detection of the target area. Second, the use of the convolutional neural network alone to detect the oil tank for the target area often has the problem of low accuracy. For example, the convolutional neural network often does not consider some main feature information in the oil tank image (for example, structure type, shadow area, etc.), resulting in low accuracy of oil tank detection. Based on this, the oil tank layout information generation method of some embodiments of the present disclosure can first perform image clipping on the above-mentioned target synthetic aperture radar image to obtain a sub-image set. Here, the target aperture radar image is clipped for subsequent acquisition of the corresponding optical image set. Among them, the above-mentioned target synthetic aperture radar image is an image taken by a synthetic aperture radar sensor for the target area, and the resolution size satisfies the first condition. Then, for each area information in the area information set corresponding to the sub-image set, an optical image set is obtained, which is taken by an optical image sensor for the area corresponding to the above-mentioned area information and the resolution size satisfies different second conditions. Here, the purpose of obtaining the optical image set is that the corresponding optical image is easier to detect the target object than the target synthetic aperture radar image. Although the accuracy of oil tank detection by optical images is higher, no target synthetic aperture radar image can better reflect the phase information of the oil tank. Based on this, oil tank detection from both the target synthetic aperture radar image and the optical image can greatly improve the detection accuracy of the oil tank. Then, for each optical image set in the obtained optical image set, each optical image in the above optical image set is respectively input into the pre-trained oil tank feature information extraction network to obtain a feature atlas. Here, since the resolution of each optical image in the optical image set is different, the feature information extracted from the corresponding area is also different. The difference in feature information makes it possible to determine the oil tank information of each oil tank in the corresponding area from various perspectives such as the global perspective and the local perspective. Then, for each feature atlas in the obtained feature atlas group, each feature map in the feature atlas is fused to obtain a fused feature map. Here, the obtained fused feature maps are used as the input of the model to more accurately determine the oil tank information of the target area.Furthermore, for each fused feature map in the obtained fused feature map set, based on the fused feature map and the oil tank recognition model set, the position information of each first oil tank information in the first oil tank information set corresponding to the fused feature map and the radius information of each first oil tank information in the first oil tank information set can be efficiently and accurately generated. Finally, based on the target synthetic aperture radar image, the oil tank recognition model set, the position information of each first oil tank information in the obtained first oil tank information set group and the radius information of each first oil tank information in the first oil tank information set group, the oil tank layout information of the target area can be efficiently and accurately generated. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] The above and other features, advantages and aspects of the embodiments of the present disclosure will become more apparent with reference to the following detailed description in conjunction with the accompanying drawings. Throughout the accompanying drawings, the same or similar reference numerals represent the same or similar elements. It should be understood that the drawings are schematic and the originals and elements are not necessarily drawn to scale.

[0014] Figure 1 is a schematic diagram of an application scenario of a method for generating oil tank layout information according to some embodiments of the present disclosure;

[0015] Figure 2 is a flow chart of some embodiments of the method for generating oil tank layout information according to the present disclosure;

[0016] Figure 3 is a flow chart of other embodiments of the method for generating oil tank layout information according to the present disclosure;

[0017] Figure 4 is a schematic structural diagram of some embodiments of the oil tank layout information generating device according to the present disclosure;

[0018] Figure 5 It is a schematic diagram of the structure of an electronic device suitable for implementing some embodiments of the present disclosure. DETAILED DESCRIPTION

[0019] Embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although certain embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be construed as being limited to the embodiments set forth herein. On the contrary, these embodiments are provided to provide a more thorough and complete understanding of the present disclosure. It should be understood that the drawings and embodiments of the present disclosure are only for exemplary purposes and are not intended to limit the scope of protection of the present disclosure.

[0020] It should also be noted that, for ease of description, only the parts related to the invention are shown in the drawings. In the absence of conflict, the embodiments and features in the embodiments of the present disclosure can be combined with each other.

[0021] It should be noted that the concepts such as "first" and "second" mentioned in the present disclosure are only used to distinguish different devices, modules or units, and are not used to limit the order or interdependence of the functions performed by these devices, modules or units.

[0022] It should be noted that the modifications of "one" and "plurality" mentioned in the present disclosure are illustrative rather than restrictive, and those skilled in the art should understand that unless otherwise clearly indicated in the context, it should be understood as "one or more".

[0023] The names of the messages or information exchanged between multiple devices in the embodiments of the present disclosure are only used for illustrative purposes and are not used to limit the scope of these messages or information.

[0024] The present disclosure will be described in detail below with reference to the accompanying drawings and in conjunction with embodiments.

[0025] Figure 1 It is a schematic diagram of an application scenario of the method for generating oil tank layout information according to some embodiments of the present disclosure.

[0026] exist Figure 1In the application scenario, the electronic device 101 may first perform image cropping on the pre-acquired target synthetic aperture radar image 102 to obtain a sub-image set 103. In this application scenario, the sub-image set 103 includes: sub-image 1031, sub-image 1032, and sub-image 1033. The target synthetic aperture radar image 102 is an image captured by a synthetic aperture radar sensor for the target area, and the resolution size meets the first condition. Then, for each area information in the area information set 104 corresponding to the sub-image set 103, the electronic device 101 may obtain an optical image set captured by an optical image sensor for the area corresponding to the above area information, and the resolution size meets different second conditions. In this application scenario, the area information set 104 includes: area information 1041 corresponding to the sub-image 1031, area information 1042 corresponding to the sub-image 1032, and area information 1043 corresponding to the sub-image 1033. The optical image set 105 corresponding to the region information 1041 includes: optical image 1051, optical image 1052, and optical image 1053. The optical image set 106 corresponding to the region information 1042 includes: optical image 1061, optical image 1062, and optical image 1063. The optical image set 107 corresponding to the region information 1043 includes: optical image 1071, optical image 1072, and optical image 1073. Then, for each optical image set in the obtained optical image set group, the electronic device 101 can input each optical image in the above optical image set into a pre-trained oil tank feature information extraction network to obtain a feature atlas. In this application scenario, each optical image in the optical image set 105 is input into a pre-trained oil tank feature information extraction network to obtain a feature atlas 111. The above feature atlas 111 includes: feature map 1111, feature map 1112, and feature map 1113. Each optical image in the optical image set 106 is input into the pre-trained oil tank feature information extraction network to obtain a feature atlas set 112. The feature atlas set 112 includes: feature map 1121, feature map 1122, and feature map 1123. Each optical image in the optical image set 107 is input into the pre-trained oil tank feature information extraction network to obtain a feature atlas set 113. The feature atlas set 113 includes: feature map 1131, feature map 1132, and feature map 1133. Furthermore, for each feature atlas set in the obtained feature atlas set group, the electronic device 101 can perform feature map fusion on each feature map in the feature atlas set to obtain a fused feature map. In this application scenario, feature map 1111, feature map 1112, and feature map 1113 are fused to obtain a fused feature map 1141 in the fused feature atlas set 114. Feature maps 1121 , 1122 , and 1123 are fused to obtain a fused feature map 1142 in the fused feature map set 114 .Feature graphs 1131, 1132, and 1133 are fused to obtain a fused feature graph 1143 in the fused feature graph set 114. Further, for each fused feature graph in the obtained fused feature graph set 114, the electronic device 101 can generate, based on the fused feature graph and the oil tank recognition model set 108, the position information 109 of each first oil tank information in the first oil tank information set corresponding to the fused feature graph and the radius information 110 of each first oil tank information in the first oil tank information set. In this application scenario, the oil tank recognition model set 108 may include: oil tank recognition model 1081, oil tank recognition model 1082, and oil tank recognition model 1083. Finally, the electronic device 101 can generate the oil tank layout information 115 of the above-mentioned target area based on the above-mentioned target synthetic aperture radar image 102, the above-mentioned oil tank identification model set 108, the location information of each first oil tank information in the obtained first oil tank information set group 109, and the radius information 110 of each first oil tank information in the above-mentioned first oil tank information set group.

[0027] It should be noted that the above-mentioned electronic device 101 can be hardware or software. When the electronic device is hardware, it can be implemented as a distributed cluster consisting of multiple server terminals or terminal devices, or it can be implemented as a single server terminal or a single terminal device. When the electronic device is embodied as software, it can be installed in the hardware devices listed above. It can be implemented as multiple software or software modules for providing distributed services, for example, or it can be implemented as a single software or software module. No specific limitation is made here.

[0028] It should be understood that Figure 1 The number of electronic devices in the embodiment is only for illustration. Any number of electronic devices may be provided according to implementation requirements.

[0029] Continue to refer Figure 2 , shows a process 200 of some embodiments of the method for generating oil tank layout information according to the present disclosure. The method for generating oil tank layout information comprises the following steps:

[0030] Step 201 , cropping a pre-acquired target synthetic aperture radar image to obtain a sub-image set.

[0031] In some embodiments, the execution subject may perform image cropping on the target synthetic aperture radar image (Synthetic Aperture Radar, SAR) to obtain a sub-image set. Each sub-image in the sub-image set has a high probability of being an image corresponding to each oil tank. The target synthetic aperture radar image is an image taken by a synthetic aperture radar sensor for a target area, and the resolution size satisfies the first condition. The target area may be an area where oil tank detection is to be performed. The first condition may be pre-set. The first condition may be that the resolution size of the image is within a certain interval, for example, the resolution size of the target synthetic aperture radar image may be 1000 pixels × 800 pixels. The interval corresponding to the first condition may be [800, 1500] pixels × [400, 1000] pixels.

[0032] As an example, the execution subject may perform image cropping on the target synthetic aperture radar image to obtain a sub-image set, including the following steps:

[0033] In the first step, the execution subject extracts color feature information of the target synthetic aperture radar image.

[0034] In practice, the execution subject may extract the color feature information of the target synthetic aperture radar image by using a color moment method.

[0035] In a second step, the execution subject may crop the target synthetic aperture radar image according to the color feature information to obtain a cropped image set.

[0036] In the third step, a cropped image satisfying the fifth condition is selected from the cropped image set as a sub-image to obtain the sub-image set. The fifth condition may be preset. The fifth condition may be a cropped image having a fixed shape (e.g., a rectangle) and an area greater than the first threshold.

[0037] In some optional implementations of some embodiments, the step of cropping the target synthetic aperture radar image to obtain a sub-image set may include the following steps:

[0038] In the first step, the execution entity may extract texture feature information of the target synthetic aperture radar image.

[0039] In practice, the execution entity may use LBP (Local Binary Pattern) to extract texture feature information of the target synthetic aperture radar image.

[0040] In a second step, the execution subject may crop the target synthetic aperture radar image according to the texture feature information to obtain a cropped image set.

[0041] In practice, the execution subject may determine the texture feature information including multiple circles from the texture feature information. Then, the execution subject may crop the image corresponding to the texture feature information including multiple circles from the target synthetic aperture radar image to obtain the cropped image set.

[0042] In the third step, the execution subject may select the cropped images satisfying the third condition from the cropped image set as sub-images to obtain the sub-image set. The third condition may be preset. The third condition may be a cropped image whose area is greater than the second threshold.

[0043] Step 202: for each area information in the area information set corresponding to the sub-image set, obtain an optical image set which is taken by an optical image sensor for the area corresponding to the above area information and has a resolution that satisfies different second conditions.

[0044] In some embodiments, the execution subject may obtain, for each area information in the obtained area information set, an optical image set captured by an optical image sensor for the area corresponding to the area information, whose resolution size satisfies different second conditions. Each sub-image in the sub-image set has corresponding area information. The second condition may be pre-set. The second condition may be that the resolution size of the optical image is within a certain resolution interval. The area information may represent the geographic location information of the area. The area information corresponding to the sub-image may be the geographic location information corresponding to the sub-image. That is, the geographic location information of the image content corresponding to the sub-image.

[0045] For example, the region information may include various coordinates to represent the geographical location information of the corresponding region.

[0046] It should be noted that the optical image set corresponds to each second condition. The resolutions corresponding to each optical image in the optical image set are different. That is, each second condition corresponds to a different interval.

[0047] As an example, the optical image set includes: a first optical image, a second optical image, and a third optical image. The interval corresponding to the first optical image is: [100, 200] pixels × [50, 100] pixels. The interval corresponding to the second optical image is: [400, 600] pixels × [200, 300] pixels. The interval corresponding to the second optical image is: [800, 1500] pixels × [400, 1000] pixels. The resolution of the first optical image is: 150 pixels × 75 pixels. The resolution of the second optical image is: 500 pixels × 250 pixels. The resolution of the third optical image is: 1000 pixels × 800 pixels.

[0048] Step 205: generating, based on the target synthetic aperture radar image and the pre-trained oil tank recognition model set, position information of each first oil tank information in the first oil tank information set corresponding to the target synthetic aperture radar image and radius information of each first oil tank information in the first oil tank information set.

[0049] In some embodiments, the execution subject may generate the position information of each first oil tank information in the first oil tank information set corresponding to the target synthetic aperture radar image and the radius information of each first oil tank information in the first oil tank information set according to the target synthetic aperture radar image and the pre-trained oil tank recognition model set. The position information of the first oil tank information may be the coordinate information of the oil tank on the target synthetic aperture radar image.

[0050] As an example, the execution subject may perform the following steps to generate the position information of each first oil tank information in the first oil tank information set corresponding to the target synthetic aperture radar image and the radius information of each first oil tank information in the first oil tank information set according to the target synthetic aperture radar image and the pre-trained oil tank recognition model set:

[0051] In the first step, the above-mentioned execution entity can input the above-mentioned target synthetic aperture radar image into each oil tank recognition model in the pre-trained oil tank recognition model set to output the position information of each third oil tank information in the third oil tank information set corresponding to the above-mentioned target synthetic aperture radar image and the radius information of each third oil tank information in the above-mentioned third oil tank information set, and obtain the position information of each third oil tank information in the third oil tank information set group and the radius information of each third oil tank information in the above-mentioned third oil tank information set group.

[0052] In a second step, the execution subject may perform location information deduplication processing to remove duplicate location information in the location information of each third oil tank information in the third oil tank information set, and obtain deduplicated location information of each third oil tank information.

[0053] The third step is to determine the position information of each third oil tank information after deduplication as the position information of each first oil tank information in the first oil tank information set corresponding to the target synthetic aperture radar image, and obtain the radius information of each first oil tank information in the first oil tank information set.

[0054] Step 203: for each optical image set in the obtained optical image set group, each optical image in the optical image set is input into a pre-trained oil tank feature information extraction network to obtain a feature atlas.

[0055] In some embodiments, for each optical image set in the obtained optical image set group, the execution subject may input each optical image in the optical image set into a pre-trained oil tank feature information extraction network to obtain a feature atlas. The oil tank feature information extraction network may be, but is not limited to, at least one of the following: Convolutional Neural Networks (CNN), Residual Network (ResNet).

[0056] Step 204: for each feature atlas in the obtained feature atlas group, feature maps are fused on each feature map in the feature atlas to obtain a fused feature map.

[0057] In some embodiments, for each feature atlas in the obtained feature atlas group, the execution entity may perform feature map fusion on each feature map in the feature atlas to obtain a fused feature map.

[0058] As an example, the above execution entity may concatenate the feature maps in the feature set to obtain a fused feature map.

[0059] As another example, the execution entity may add the feature maps in the feature map set to obtain the added feature map as the fused feature map.

[0060] Step 205, for each fused feature map in the obtained fused feature map set, based on the fused feature map and the oil tank recognition model set, generate the position information of each first oil tank information in the first oil tank information set corresponding to the fused feature map and the radius information of each first oil tank information in the first oil tank information set.

[0061] In some embodiments, for each fused feature map in the obtained fused feature map set, the execution entity may generate the position information of each first oil tank information in the first oil tank information set corresponding to the fused feature map and the radius information of each first oil tank information in the first oil tank information set according to the fused feature map and the oil tank recognition model set. The position information of the first oil tank information may be the coordinate information of the oil tank. The oil tank recognition model set may include: SSD (Single Shot MultiBox Detector) algorithm, R-CNN (Region-Convolutional Neural Networks) algorithm, Fast R-CNN (Fast Region- Convolutional Neural Networks) algorithm, Faster R-CNN (Faster Region- Convolutional Neural Networks) algorithm, SPP-NET (Spatial Pyramid Pooling Network) algorithm, YOLO (You Only LookOnce) algorithm, FPN (Feature Pyramid Networks) algorithm, DCN (Deformable ConvNets) algorithm, RetinaNet target detection algorithm.

[0062] Among them, the training data set of each oil tank recognition model in the oil tank recognition model set is generated by the following steps:

[0063] In the first step, 8 times sinc interpolation is performed on the benchmark synthetic aperture radar image.

[0064] In the second step, the interpolated image is divided into image blocks of size 512x512, and more than 500 image blocks are selected.

[0065] The third step is to label the oil tank target in each image block, marking two types of information: radius and location.

[0066] As an example, the above execution entity can find two strong scattering points parallel to the boundary in the distance direction for each oil tank area, point A and point B, and draw a circle for the arc near point B, where point P is the center of the circle, adjust the radius R, move the circle to pass through point A, and mark the circumscribed rectangle of the two circles as the position box, and mark the position box of each oil tank. The marking format is: x, y, w, h. Among them, x, y represent the pixel position of the upper left corner of the oil tank target in the image, w, h represent the number of pixels of the length and width of the circumscribed box, and record the radius R of the marked oil tank at the same time.

[0067] In the fourth step, the above execution entity can generate a corresponding xml file for each image block, and the file contains the number of oil tanks and the x, y, w, h, and R parameter information of all oil tanks.

[0068] The fifth step is to divide the image blocks into training set and test set in a ratio of 3:1.

[0069] As an example, the execution subject may perform the following steps to generate the position information of each first oil tank information in the first oil tank information set corresponding to the fused feature graph and the radius information of each first oil tank information in the first oil tank information set according to the fused feature graph and the oil tank identification model set:

[0070] In the first step, the above-mentioned execution entity can input the above-mentioned fused feature map into each oil tank recognition model in the pre-trained oil tank recognition model set to output the position information of each fifth oil tank information in the fifth oil tank information set corresponding to the above-mentioned fused feature map and the radius information of each fifth oil tank information in the above-mentioned fifth oil tank information set, and obtain the position information of each fifth oil tank information in the fifth oil tank information set group and the radius information of each fifth oil tank information in the above-mentioned fifth oil tank information set group.

[0071] In the second step, the above-mentioned execution entity can perform location information deduplication processing to remove the repeated location information in the location information of each fifth oil tank information in the fifth oil tank information set group, and obtain the location information of each first oil tank information in the first oil tank information set corresponding to the above-mentioned fused feature map and the radius information of each first oil tank information in the above-mentioned first oil tank information set.

[0072] In some optional implementations of some embodiments, the generating, according to the fused feature graph and the oil tank recognition model set, the position information of each first oil tank information in the first oil tank information set corresponding to the fused feature graph and the radius information of each first oil tank information in the first oil tank information set may include the following steps:

[0073] In the first step, the execution entity may input the fused feature map into a pre-trained target oil tank recognition model to obtain the position information of each fifth oil tank information in the fifth oil tank information set and the radius information of each fifth oil tank information in the fifth oil tank information set.

[0074] In the second step, the above-mentioned execution entity can input the above-mentioned fused feature map into each oil tank recognition model in the above-mentioned at least one oil tank recognition model to output the position information of each sixth oil tank information in the sixth oil tank information set and the radius information of each sixth oil tank information in the above-mentioned sixth oil tank information set, and obtain the position information of each sixth oil tank information in the sixth oil tank information set group and the radius information of each sixth oil tank information in the above-mentioned sixth oil tank information set group.

[0075] In the third step, for each piece of location information of the fifth oil tank information, the following determination steps are performed:

[0076] In a first sub-step, the execution entity may determine whether there is location information in the sixth oil tank information that is the same as the location information of the fifth oil tank information.

[0077] In the second sub-step, in response to determining existence, the execution subject may filter out the location information that is the same as the location information of each of the fifth oil tank information from the location information of each of the sixth oil tank information as the second target location information, and obtain at least one second target location information.

[0078] In a third sub-step, the execution entity may determine a second number of each second target location information corresponding to the at least one second target location information.

[0079] In step 4, in response to determining that the second number is greater than or equal to the predetermined threshold, the execution entity may determine the location information of the fifth oil tank information as the location information of the first oil tank information, and determine the radius information of the fifth oil tank information as the radius information corresponding to the first oil tank information. The predetermined threshold may be preset.

[0080] Step 206, generating the oil tank layout information of the target area according to the target synthetic aperture radar image, the oil tank identification model set, the position information of each first oil tank information in the obtained first oil tank information set group, and the radius information of each first oil tank information in the first oil tank information set group.

[0081] In some embodiments, the execution subject may generate the oil tank layout information of the target area in various ways according to the target synthetic aperture radar image, the oil tank identification model set, the position information of each first oil tank information in the obtained first oil tank information set group, and the radius information of each first oil tank information in the first oil tank information set group. The oil tank layout information may be the position information and radius information corresponding to each oil tank in the target area. The oil tank layout information may be information in various forms. For example, it may be information in a tabular form.

[0082] In some optional implementations of some embodiments, the generating of the oil tank layout information of the target area according to the target synthetic aperture radar image, the oil tank identification model set, the position information of each first oil tank information in the obtained first oil tank information set group, and the radius information of each first oil tank information in the first oil tank information set group may include the following steps:

[0083] The first step is to generate, based on the target synthetic aperture radar image and the oil tank recognition model set, the position information of each second oil tank information in the second oil tank information set corresponding to the target synthetic aperture radar image and the radius information of each second oil tank information in the second oil tank information set. The position information of the second oil tank information may be the coordinate information of the oil tank on the target synthetic aperture radar image.

[0084] As an example, the execution subject may perform the following steps to generate the position information of each second oil tank information in the second oil tank information set corresponding to the target synthetic aperture radar image and the radius information of each second oil tank information in the second oil tank information set according to the target synthetic aperture radar image and the pre-trained oil tank recognition model set:

[0085] In the first step, the above-mentioned execution entity can input the above-mentioned target synthetic aperture radar image into each oil tank recognition model in the pre-trained oil tank recognition model set to output the position information of each third oil tank information in the third oil tank information set corresponding to the above-mentioned target synthetic aperture radar image and the radius information of each third oil tank information in the above-mentioned third oil tank information set, and obtain the position information of each third oil tank information in the third oil tank information set group and the radius information of each third oil tank information in the above-mentioned third oil tank information set group.

[0086] In a second step, the execution subject may perform location information deduplication processing to remove duplicate location information in the location information of each third oil tank information in the third oil tank information set, and obtain deduplicated location information of each third oil tank information.

[0087] The third step is to determine the position information of each third oil tank information after deduplication as the position information of each second oil tank information in the second oil tank information set corresponding to the target synthetic aperture radar image, and obtain the radius information of each second oil tank information in the second oil tank information set.

[0088] In some optional implementations of some embodiments, the generating, based on the target synthetic aperture radar image and the pre-trained oil tank recognition model set, the position information of each second oil tank information in the second oil tank information set corresponding to the target synthetic aperture radar image and the radius information of each second oil tank information in the second oil tank information set may include the following steps:

[0089] In the first step, the execution subject may obtain the priority of each oil tank identification model in the oil tank identification model set, wherein the priority of each oil tank model may be preset.

[0090] In the second step, the execution subject may select an oil tank recognition model whose priority satisfies the fourth condition from the oil tank recognition model set as the target oil tank recognition model. The fourth condition may be pre-set. The fourth condition may be the oil tank recognition model with the highest priority. As an example, the target oil tank recognition model may be the Faster R-CNN algorithm based on Vgg (Visual Geometry Group) 16.

[0091] Among them, the parameters of the Faster R-CNN algorithm can be set as follows:

[0092] Vgg16 network parameter settings: convolution kernel kernel_size=3, padding=1, stride=1. Pooling layer parameters: convolution kernel kernel_size=2, padding=0, stride=2. RPN network parameter settings: input channel number 256, feature layer channel number 256, Anchor generator, scales=[8], ratios=[0.5, 1.0, 2.0], strides=[4, 8, 16, 32, 64]). Number of input channels in_channels=256; number of fully connected layer output channels, fc_out_channels=1024. ROI feature layer size roi_feat_size=7. Classification parameter settings: num_classes=2. After bounding box regression, expand 1 parameter vector to store radius R. The result outputs the detection box parameters: x, y, w, h and radius R. Training parameter settings: the maximum number of target boxes is set to 100; the learning rate of the network is 2e-4; the batch_size is 8; the total number of iterations is 300 epochs. Loss function settings, the category loss item selects cross entropy, and the bounding box loss function selects L1 norm.

[0093] In the third step, the execution entity may input the target synthetic aperture radar image into a pre-trained target oil tank recognition model to obtain the position information of each third oil tank information in the third oil tank information set and the radius information of each third oil tank information in the third oil tank information set.

[0094] In the fourth step, the execution subject may input the target synthetic aperture radar image into each oil tank recognition model in at least one oil tank recognition model in the oil tank recognition model set to output the position information of each fourth oil tank information in the fourth oil tank information set and the radius information of each fourth oil tank information in the fourth oil tank information set, and obtain the position information of each fourth oil tank information in the fourth oil tank information set group and the radius information of each fourth oil tank information in the fourth oil tank information set group. The at least one oil tank recognition model is each oil tank recognition model in the oil tank recognition model set except for the target oil tank recognition model.

[0095] In step 5, for each piece of location information of the third oil tank information, the execution subject may perform the following determination steps:

[0096] In a first sub-step, the execution entity may determine whether there is location information in the fourth oil tank information that is the same as the location information of the third oil tank information.

[0097] In the second sub-step, in response to determining existence, the execution entity may filter out location information identical to the location information of the third oil tank information from the location information of each fourth oil tank information as first target location information, and obtain at least one first target location information.

[0098] In a third sub-step, the execution entity may determine a first number of each first target location information corresponding to the at least one first target location information.

[0099] In the sixth step, in response to determining that the above-mentioned first number is greater than or equal to a predetermined threshold, the above-mentioned execution entity can determine the location information of the above-mentioned third oil tank information as the location information of the second oil tank information, and determine the radius information of the above-mentioned third oil tank information as the radius information corresponding to the second oil tank information.

[0100] As an inventive point of an embodiment of the present disclosure, the second technical problem mentioned in the background technology is solved: "The use of convolutional neural networks alone to detect oil tanks in target areas often has the problem of low accuracy". The factors that lead to the relatively low accuracy of oil tank detection are as follows: convolutional neural networks often do not consider some main feature information in oil tank images (for example, structural types, shadow areas, etc.), resulting in low accuracy of oil tank detection. In order to achieve this effect, the present disclosure adopts multiple oil tank recognition models to perform oil tank detection. The comprehensive detection of multiple oil tank recognition models can greatly improve the accuracy of oil tank detection in target areas. In addition, corresponding priorities are set for each of the multiple oil tank recognition models. The purpose of setting priorities is that each oil tank recognition model has its own model recognition characteristics. For each characteristic of the oil tank, there is often a model with particularly high accuracy in identifying the oil tank among the multiple oil tank recognition models (that is, the model is sensitive to each characteristic of the oil tank and is easy to identify). In this way, setting priorities for each oil tank recognition model can make the detection of each oil tank in the target area more accurate.

[0101] In the second step, the oil tank layout information of the above-mentioned target area is generated according to the position information of each second oil tank information in the above-mentioned second oil tank information set, the radius information of each second oil tank information in the above-mentioned second oil tank information set, the position information of each first oil tank information in the above-mentioned first oil tank information set group and the radius information of each first oil tank information in the above-mentioned first oil tank information set group.

[0102] As an example, the above-mentioned execution entity can summarize and deduplicate the location information of the above-mentioned each second oil tank information, the radius information of the above-mentioned each second oil tank information, the location information of each first oil tank information in the obtained first oil tank information set group, and the radius information of each first oil tank information in the above-mentioned first oil tank information set group according to a predetermined data format to obtain the oil tank layout information of the above-mentioned target area.

[0103] In some optional implementations of some embodiments, the generating of the oil tank layout information of the target area according to the location information of each of the second oil tank information, the radius information of each of the second oil tank information, the location information of each of the first oil tank information in the obtained first oil tank information set group, and the radius information of each of the first oil tank information in the first oil tank information set group may include the following steps:

[0104] In the first step, for each of the second oil tank information, the following generation steps are performed to generate the oil tank layout information:

[0105] In the first sub-step, the execution entity may determine the location information of the second oil tank information.

[0106] In a second sub-step, the execution entity may determine the area information corresponding to the location information.

[0107] In a third sub-step, the execution entity may determine the first oil tank information set in the first oil tank information set group corresponding to the area information as the target oil tank information set.

[0108] In a fourth sub-step, the execution entity may determine whether there is first oil tank information having the same geographical location as the second oil tank information in the target oil tank information set.

[0109] In a fifth sub-step, the execution subject may generate layout information indicating that an oil tank corresponding to the second oil tank information exists at the position information in the target area in response to the existence determination.

[0110] The second step is to generate the oil tank layout information of the target area according to the obtained layout information.

[0111] As an example, the execution subject may first obtain the format information of the oil tank layout information, and then, according to the format information, the execution subject may convert each obtained layout information into the oil tank layout information of the target area.

[0112] The above-mentioned embodiments of the present disclosure have the following beneficial effects: the oil tank layout information generation method of some embodiments of the present disclosure accurately and efficiently generates the oil tank layout information of the target area. Specifically, the reasons for the inaccurate and ineffective generation of the oil tank layout information are: first, the synthetic aperture radar image is used as the input of the neural network to detect the oil tank, which often has the problem of low accuracy. In addition, the high-resolution target synthetic aperture radar image is used as the input of the neural network, and there is a problem of large amount of calculation for training the convolutional neural network and the subsequent oil tank detection of the target area. Second, the use of the convolutional neural network alone to detect the oil tank for the target area often has the problem of low accuracy. For example, the convolutional neural network often does not consider some main feature information in the oil tank image (for example, structure type, shadow area, etc.), resulting in low accuracy of oil tank detection. Based on this, the oil tank layout information generation method of some embodiments of the present disclosure can first perform image clipping on the above-mentioned target synthetic aperture radar image to obtain a sub-image set. Here, the target aperture radar image is clipped for subsequent acquisition of the corresponding optical image set. Among them, the above-mentioned target synthetic aperture radar image is an image taken by a synthetic aperture radar sensor for the target area, and the resolution size satisfies the first condition. Then, for each area information in the area information set corresponding to the sub-image set, an optical image set is obtained, which is taken by an optical image sensor for the area corresponding to the above-mentioned area information and the resolution size satisfies different second conditions. Here, the purpose of obtaining the optical image set is that the corresponding optical image is easier to detect the target object than the target synthetic aperture radar image. Although the accuracy of oil tank detection by optical images is higher, no target synthetic aperture radar image can better reflect the phase information of the oil tank. Based on this, oil tank detection from both the target synthetic aperture radar image and the optical image can greatly improve the detection accuracy of the oil tank. Then, for each optical image set in the obtained optical image set, each optical image in the above optical image set is respectively input into the pre-trained oil tank feature information extraction network to obtain a feature atlas. Here, since the resolution of each optical image in the optical image set is different, the feature information extracted from the corresponding area is also different. The difference in feature information makes it possible to determine the oil tank information of each oil tank in the corresponding area from various perspectives such as the global perspective and the local perspective. Then, for each feature atlas in the obtained feature atlas group, each feature map in the feature atlas is fused to obtain a fused feature map. Here, the obtained fused feature maps are used as the input of the model to more accurately determine the oil tank information of the target area.Furthermore, for each fused feature map in the obtained fused feature map set, based on the fused feature map and the oil tank recognition model set, the position information of each first oil tank information in the first oil tank information set corresponding to the fused feature map and the radius information of each first oil tank information in the first oil tank information set can be efficiently and accurately generated. Finally, based on the target synthetic aperture radar image, the oil tank recognition model set, the position information of each first oil tank information in the obtained first oil tank information set group and the radius information of each first oil tank information in the first oil tank information set group, the oil tank layout information of the target area can be efficiently and accurately generated.

[0113] Further references Figure 3 , shows a process 300 of another embodiment of the method for generating oil tank layout information according to the present disclosure. The method for generating oil tank layout information comprises the following steps:

[0114] Step 301 , cropping a pre-acquired target synthetic aperture radar image to obtain a sub-image set.

[0115] Step 302: for each area information in the area information set corresponding to the sub-image set, obtain an optical image set captured by an optical image sensor for the area corresponding to the above area information and having a resolution satisfying different second conditions.

[0116] Step 303: for each optical image set in the obtained optical image set group, each optical image in the optical image set is input into a pre-trained oil tank feature information extraction network to obtain a feature atlas.

[0117] Step 304: for each feature atlas in the obtained feature atlas group, feature maps are fused on each feature map in the feature atlas to obtain a fused feature map.

[0118] Step 305, for each fused feature map in the obtained fused feature map set, based on the fused feature map and the oil tank recognition model set, generate the position information of each first oil tank information in the first oil tank information set corresponding to the fused feature map and the radius information of each first oil tank information in the first oil tank information set.

[0119] Step 306, generating the oil tank layout information of the target area according to the target synthetic aperture radar image, the oil tank identification model set, the position information of each first oil tank information in the obtained first oil tank information set group, and the radius information of each first oil tank information in the first oil tank information set group.

[0120] In some embodiments, the specific implementation of steps 301-306 and the technical effects thereof can be referred to in Figure 2The steps 201-206 in the corresponding embodiment are not described in detail here.

[0121] Step 307, obtaining the location information of each oil tank corresponding to the above oil tank layout information.

[0122] In some embodiments, an execution entity (e.g. Figure 1 The electronic device shown in the figure) can obtain the location information of each oil tank corresponding to the above-mentioned oil tank layout information through a wired or wireless method.

[0123] Step 308: Generate the center coordinates corresponding to each of the oil tanks according to the position information of each of the oil tanks.

[0124] In some embodiments, the execution entity may generate the center coordinates corresponding to the oil tanks according to the location information of the oil tanks, wherein the center coordinates may represent the location information of the oil tanks in the target area.

[0125] As an example, for each of the oil tanks, the execution subject may perform the following generation steps to generate the center coordinates corresponding to the oil tank:

[0126] The first step is to determine the coordinate information of the above oil tanks.

[0127] The second step is to obtain the average coordinate information corresponding to each coordinate information.

[0128] As an example, the individual coordinate information includes: (2, 3, 4), (4, 6, 8). The average coordinate information may be (3, 4.5, 6).

[0129] The third step is to determine the above average coordinate information as the center coordinate corresponding to the above oil tank.

[0130] Step 309: Determine the distance between every two oil tanks in each oil tank according to the central coordinates corresponding to each of the oil tanks, and obtain a distance set.

[0131] In some embodiments, the execution entity may determine the distance between every two oil tanks in each oil tank according to the central coordinates corresponding to each oil tank, and obtain a distance set.

[0132] As an example, the execution entity may determine the Euclidean distance between every two oil tanks in each oil tank according to the central coordinates corresponding to each oil tank, and obtain a distance set.

[0133] Step 310: verify the oil tank layout information according to the distance set.

[0134] In some embodiments, the execution entity may verify the oil tank layout information according to the distance set.

[0135] As an example, the execution entity may first filter out distances whose values ​​are less than a predetermined threshold from the distance set to obtain at least one distance. Then, the execution entity may determine the oil tank set corresponding to each distance in the at least one distance to obtain an oil tank set group. Furthermore, the execution entity may perform deduplication processing on the oil tank set group to obtain the deduplicated oil tanks. Finally, the execution entity may compare the deduplicated oil tanks with the oil tank layout information to determine whether the deduplicated oil tanks are consistent with the oil tanks corresponding to the oil tank layout information. If they are consistent, the verification is passed. If they are inconsistent, the verification fails.

[0136] from Figure 3 It can be seen that Figure 2 Compared with the description of some corresponding embodiments, Figure 3 The process 300 of the method for generating the oil tank layout information in some corresponding embodiments further highlights the specific steps of verifying the oil tank layout information. Therefore, the solutions described in these embodiments more accurately determine the oil tank information of each oil tank in the target area.

[0137] Further references Figure 4 As an implementation of the methods shown in the above figures, the present disclosure provides some embodiments of a device for generating oil tank layout information. Figure 2 Corresponding to the method embodiments shown, the device can be specifically applied to various electronic devices.

[0138] like Figure 4As shown, an oil tank layout information generating device 400 includes: a cropping unit 401, an acquisition unit 402, an input unit 403, a fusion unit 404, a first generating unit 405 and a second generating unit 406. The cropping unit 401 is configured to perform image cropping on a pre-acquired target synthetic aperture radar image to obtain a sub-image set, wherein the target synthetic aperture radar image is an image taken by a synthetic aperture radar sensor for a target area and whose resolution size satisfies a first condition; the acquisition unit 402 is configured to acquire, for each area information in a regional information set corresponding to the sub-image set, an optical image set taken by an optical image sensor for an area corresponding to the above regional information and whose resolution size satisfies different second conditions; the input unit 403 is configured to, for each optical image set in the obtained optical image set group, input each optical image in the above optical image set into a pre-trained oil tank feature information extraction network to obtain a feature atlas; the fusion unit 404 is configured to, for the obtained For each feature atlas in the obtained feature atlas group, feature maps are fused on each feature atlas in the above feature atlas to obtain a fused feature map; the first generating unit 405 is configured to generate, for each fused feature map in the obtained fused feature atlas, the position information of each first oil tank information in the first oil tank information set corresponding to the fused feature map and the radius information of each first oil tank information in the above first oil tank information set according to the above fused feature map and the oil tank recognition model set; the second generating unit 406 is configured to generate the oil tank layout information of the above target area according to the above target synthetic aperture radar image, the above oil tank recognition model set, the position information of each first oil tank information in the obtained first oil tank information set group and the radius information of each first oil tank information in the above first oil tank information set group.

[0139] It is understood that the units described in the device 400 are similar to those described in the reference Figure 2 Therefore, the operations, features and beneficial effects described above for the method are also applicable to the device 400 and the units included therein, and will not be described in detail here.

[0140] Reference below Figure 5 , which shows an electronic device (eg, Figure 1 A schematic diagram of the structure of an electronic device (electronic device in FIG. 500). Figure 5 The electronic device shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present disclosure.

[0141] like Figure 5As shown, the electronic device 500 may include a processing device (e.g., a central processing unit, a graphics processor, etc.) 501, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 502 or a program loaded from a storage device 508 into a random access memory (RAM) 503. In the RAM 503, various programs and data required for the operation of the electronic device 500 are also stored. The processing device 501, the ROM 502, and the RAM 503 are connected to each other via a bus 504. An input / output (I / O) interface 505 is also connected to the bus 504.

[0142] Typically, the following devices may be connected to the I / O interface 505: an input device 506 including, for example, a touch screen, a touch pad, a keyboard, a mouse, a camera, a microphone, an accelerometer, a gyroscope, etc.; an output device 507 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; a storage device 508 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 509. The communication device 509 may allow the electronic device 500 to communicate with other devices wirelessly or by wire to exchange data. Although Figure 5 The electronic device 500 is shown with various devices, but it should be understood that it is not required to implement or possess all the devices shown. More or fewer devices may be implemented or possessed instead. Figure 5 Each block shown in the figure may represent one device, or may represent multiple devices as required.

[0143] In particular, according to some embodiments of the present disclosure, the process described above with reference to the flowchart can be implemented as a computer software program. For example, some embodiments of the present disclosure include a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program includes a program code for executing the method shown in the flowchart. In some such embodiments, the computer program can be downloaded and installed from the network through the communication device 609, or installed from the storage device 608, or installed from the ROM 602. When the computer program is executed by the processing device 601, the above-mentioned functions defined in the method of some embodiments of the present disclosure are executed.

[0144] It should be noted that the computer-readable medium mentioned above in some embodiments of the present disclosure may be a computer-readable signal medium or a computer-readable storage medium or any combination of the above two. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In some embodiments of the present disclosure, the computer-readable storage medium may be any tangible medium containing or storing a program, which may be used by or in combination with an instruction execution system, device or device. In some embodiments of the present disclosure, the computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, which carries a computer-readable program code. This propagated data signal may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. Computer readable signal media may also be any computer readable medium other than computer readable storage media, which may send, propagate or transmit a program for use by or in conjunction with an instruction execution system, apparatus or device. The program code contained on the computer readable medium may be transmitted using any appropriate medium, including but not limited to: wires, optical cables, RF (radio frequency), etc., or any suitable combination of the above.

[0145] In some embodiments, the client and the server may communicate using any currently known or future developed network protocol such as HTTP (HyperText Transfer Protocol), and may be interconnected with any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include a local area network ("LAN"), a wide area network ("WAN"), an internet (e.g., the Internet), and a peer-to-peer network (e.g., an ad hoc peer-to-peer network), as well as any currently known or future developed network.

[0146] The computer-readable medium may be included in the electronic device; or it may exist independently without being installed in the electronic device. The computer-readable medium carries one or more programs. When the one or more programs are executed by the electronic device, the electronic device: performs image cropping on the pre-acquired target synthetic aperture radar image to obtain a sub-image set, wherein the target synthetic aperture radar image is an image taken by a synthetic aperture radar sensor for the target area, and the resolution size satisfies a first condition; for each area information in the area information set corresponding to the sub-image set, obtain an optical image set taken by an optical image sensor for the area corresponding to the area information, and the resolution size satisfies different second conditions; for each optical image set in the obtained optical image set group, input each optical image in the optical image set into a pre-trained oil tank feature information extraction system. network, and obtain a feature atlas; for each feature atlas in the obtained feature atlas group, perform feature map fusion on each feature map in the above feature atlas to obtain a fused feature map; for each fused feature map in the obtained fused feature map, generate the position information of each first oil tank information in the first oil tank information set corresponding to the fused feature map and the radius information of each first oil tank information in the above first oil tank information set according to the above fused feature map and the oil tank recognition model set; generate the oil tank layout information of the above target area according to the above target synthetic aperture radar image, the above oil tank recognition model set, the position information of each first oil tank information in the obtained first oil tank information set group and the radius information of each first oil tank information in the above first oil tank information set group.

[0147] Computer program code for performing the operations of some embodiments of the present disclosure may be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, C++, and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer may be connected to the user's computer via any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0148] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present disclosure. In this regard, each square box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and the module, the program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some implementations as replacements, the functions marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two square boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0149] The units described in some embodiments of the present disclosure may be implemented by software or by hardware. The described units may also be provided in a processor, for example, may be described as: a processor including a cropping unit, an acquisition unit, an input unit, a fusion unit, a first generation unit, and a second generation unit. The names of these units do not, in some cases, constitute limitations on the units themselves, for example, the cropping unit may also be described as a "unit for cropping a pre-acquired target synthetic aperture radar image to obtain a sub-image set".

[0150] The functions described above herein may be performed at least in part by one or more hardware logic components. For example, without limitation, exemplary types of hardware logic components that may be used include: field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), complex programmable logic devices (CPLDs), and the like.

[0151] The above descriptions are only some preferred embodiments of the present disclosure and an explanation of the technical principles used. Those skilled in the art should understand that the scope of the invention involved in the embodiments of the present disclosure is not limited to the technical solutions formed by a specific combination of the above technical features, but should also cover other technical solutions formed by any combination of the above technical features or their equivalent features without departing from the above inventive concept. For example, the above features are replaced with (but not limited to) technical features with similar functions disclosed in the embodiments of the present disclosure.

Claims

1. A method for generating oil tank layout information, comprising: Performing image cropping on a pre-acquired target synthetic aperture radar image to obtain a sub-image set, wherein the target synthetic aperture radar image is an image captured by a synthetic aperture radar sensor for a target area and having a resolution that satisfies a first condition; For each area information in the area information set corresponding to the sub-image set, obtaining an optical image set captured by an optical image sensor for the area corresponding to the area information and having a resolution satisfying different second conditions; For each optical image set in the obtained optical image set group, inputting each optical image in the optical image set into a pre-trained oil tank feature information extraction network to obtain a feature atlas; For each feature atlas in the obtained feature atlas group, performing feature map fusion on each feature map in the feature atlas to obtain a fused feature map; For each fused feature graph in the obtained fused feature graph set, based on the fused feature graph and the oil tank recognition model set, the position information of each first oil tank information in the first oil tank information set corresponding to the fused feature graph and the radius information of each first oil tank information in the first oil tank information set are generated, wherein: The step of generating, according to the fused feature graph and the oil tank identification model set, the position information of each first oil tank information in the first oil tank information set corresponding to the fused feature graph and the radius information of each first oil tank information in the first oil tank information set comprises: Inputting the fused feature map into a pre-trained target oil tank recognition model to obtain position information of each fifth oil tank information in the fifth oil tank information set and radius information of each fifth oil tank information in the fifth oil tank information set; Input the fused feature map to each oil tank recognition model in at least one oil tank recognition model to output the position information of each sixth oil tank information in the sixth oil tank information set and the radius information of each sixth oil tank information in the sixth oil tank information set, and obtain the position information of each sixth oil tank information in the sixth oil tank information set group and the radius information of each sixth oil tank information in the sixth oil tank information set group; For each piece of position information of the fifth oil tank information, the following determination steps are performed: Determine whether there is location information of each of the sixth oil tank information that is the same as the location information of the fifth oil tank information; In response to determining that there is, selecting the same location information as the location information of each fifth oil tank information from the location information of each sixth oil tank information as the second target location information, and obtaining at least one second target location information; Determine a second number of each second target position information corresponding to the at least one second target position information; In response to determining that the second number is greater than or equal to a predetermined threshold, determining the location information of the fifth oil tank information as the location information of the first oil tank information, and determining the radius information of the fifth oil tank information as the radius information corresponding to the first oil tank information; The oil tank layout information of the target area is generated according to the target synthetic aperture radar image, the oil tank identification model set, the position information of each first oil tank information in the obtained first oil tank information set group, and the radius information of each first oil tank information in the first oil tank information set group, wherein: The step of generating the oil tank layout information of the target area according to the target synthetic aperture radar image, the oil tank identification model set, the position information of each first oil tank information in the obtained first oil tank information set group, and the radius information of each first oil tank information in the first oil tank information set group comprises: According to the target synthetic aperture radar image and the oil tank recognition model set, the position information of each second oil tank information in the second oil tank information set corresponding to the target synthetic aperture radar image and the radius information of each second oil tank information in the second oil tank information set are generated, wherein, according to the target synthetic aperture radar image and the oil tank recognition model set, the position information of each second oil tank information in the second oil tank information set corresponding to the target synthetic aperture radar image and the radius information of each second oil tank information in the second oil tank information set are generated, including: Obtaining the priority of each oil tank identification model in the oil tank identification model set; Selecting an oil tank identification model whose priority satisfies a fourth condition from the oil tank identification model set as a target oil tank identification model; Inputting the target synthetic aperture radar image into a pre-trained target oil tank recognition model to obtain position information of each third oil tank information in the third oil tank information set and radius information of each third oil tank information in the third oil tank information set; Input the target synthetic aperture radar image to each oil tank recognition model in at least one oil tank recognition model in the oil tank recognition model set to output the position information of each fourth oil tank information in the fourth oil tank information set and the radius information of each fourth oil tank information in the fourth oil tank information set, and obtain the position information of each fourth oil tank information in the fourth oil tank information set group and the radius information of each fourth oil tank information in the fourth oil tank information set group, wherein the at least one oil tank recognition model is each oil tank recognition model in the oil tank recognition model set excluding the target oil tank recognition model; For each piece of location information of the third oil tank information, the following determination steps are performed: Determine whether there is location information of each fourth oil tank information that is the same as the location information of the third oil tank information; In response to determining that there is, selecting the same location information as the location information of the third oil tank information from the location information of each fourth oil tank information as the first target location information, and obtaining at least one first target location information; Determine a first number of each first target position information corresponding to the at least one first target position information; In response to determining that the first number is greater than or equal to a predetermined threshold, determining the location information of the third oil tank information as the location information of the second oil tank information, and determining the radius information of the third oil tank information as the radius information corresponding to the second oil tank information; The oil tank layout information of the target area is generated according to the position information of each second oil tank information in the second oil tank information set, the radius information of each second oil tank information in the second oil tank information set, the position information of each first oil tank information in the first oil tank information set group, and the radius information of each first oil tank information in the first oil tank information set group, wherein the oil tank layout information of the target area is generated according to the position information of each second oil tank information in the second oil tank information set, the radius information of each second oil tank information in the second oil tank information set, the position information of each first oil tank information in the first oil tank information set group, and the radius information of each first oil tank information in the first oil tank information set group, including: For each piece of second oil tank information, the following generation steps are performed to generate layout information of the oil tank corresponding to the second oil tank information: Determine the location information of the second oil tank information; Determining the area information corresponding to the location information; Determine a first oil tank information set in the first oil tank information set group corresponding to the area information as a target oil tank information set; Determine whether there is first oil tank information having the same geographical location as the second oil tank information in the target oil tank information set; In response to determining that the oil tank exists, generating layout information indicating that the oil tank corresponding to the second oil tank information exists at the position information in the target area; The oil tank layout information of the target area is generated according to the obtained layout information.

2. The method according to claim 1, wherein: The step of performing image cropping on the pre-acquired target synthetic aperture radar image to obtain a sub-image set includes: Extracting texture feature information of the target synthetic aperture radar image; According to the texture feature information, the target synthetic aperture radar image is cropped to obtain a cropped image set; The cropped images satisfying the third condition are screened out from the cropped image set as sub-images to obtain the sub-image set.

3. The method according to claim 1, wherein: The method further comprises: Obtaining location information of each oil tank corresponding to the oil tank layout information; Generate the center coordinates corresponding to each oil tank according to the position information of each oil tank; According to the central coordinates corresponding to the oil tanks, the distance between every two oil tanks in the oil tanks is determined to obtain a distance set; The oil tank layout information is verified according to the distance set.

4. A device for generating oil tank layout information, comprising: a cropping unit configured to crop a pre-acquired target synthetic aperture radar image to obtain a sub-image set, wherein the target synthetic aperture radar image is an image captured by a synthetic aperture radar sensor for a target area and has a resolution that satisfies a first condition; An acquisition unit is configured to acquire, for each area information in the area information set corresponding to the sub-image set, an optical image set captured by an optical image sensor for the area corresponding to the area information and having a resolution satisfying different second conditions; An input unit is configured to input, for each optical image set in the obtained optical image set group, each optical image in the optical image set into a pre-trained oil tank feature information extraction network to obtain a feature atlas; A fusion unit is configured to perform feature map fusion on each feature map in each feature map set in the obtained feature map set group to obtain a fused feature map; The first generating unit is configured to generate, for each fused feature map in the obtained fused feature map set, the position information of each first oil tank information in the first oil tank information set corresponding to the fused feature map and the radius information of each first oil tank information in the first oil tank information set according to the fused feature map and the oil tank recognition model set, wherein the generating the position information of each first oil tank information in the first oil tank information set corresponding to the fused feature map and the radius information of each first oil tank information in the first oil tank information set according to the fused feature map and the oil tank recognition model set includes: inputting the fused feature map into a pre-trained target oil tank recognition model to obtain the position information of each fifth oil tank information in the fifth oil tank information set and the radius information of each fifth oil tank information in the fifth oil tank information set; inputting the fused feature map into each oil tank recognition model in at least one oil tank recognition model to output the position information of each sixth oil tank information in the sixth oil tank information set and the radius information of each sixth oil tank information in the sixth oil tank information set. The radius information of the sixth oil tank information is used to obtain the position information of each sixth oil tank information in the sixth oil tank information set and the radius information of each sixth oil tank information in the sixth oil tank information set; for each fifth oil tank information in the position information of each fifth oil tank information, the following determination steps are performed: determine whether there is position information identical to the position information of the fifth oil tank information in the position information of each sixth oil tank information; in response to determining that there is, filter out the position information identical to the position information of each fifth oil tank information from the position information of each sixth oil tank information as the second target position information to obtain at least one second target position information; determine the second number of each second target position information corresponding to the at least one second target position information; in response to determining that the second number is greater than or equal to a predetermined threshold, determine the position information of the fifth oil tank information as the position information of the first oil tank information, and determine the radius information of the fifth oil tank information as the radius information corresponding to the first oil tank information; The second generating unit is configured to generate the oil tank layout information of the target area according to the target synthetic aperture radar image, the oil tank recognition model set, the position information of each first oil tank information in the first oil tank information set group, and the radius information of each first oil tank information in the first oil tank information set group, wherein the generating the oil tank layout information of the target area according to the target synthetic aperture radar image, the oil tank recognition model set, the position information of each first oil tank information in the first oil tank information set group, and the radius information of each first oil tank information in the first oil tank information set group comprises: generating a second oil tank layout information corresponding to the target synthetic aperture radar image according to the target synthetic aperture radar image and the oil tank recognition model set; The method comprises the following steps: obtaining the position information of each second oil tank information in the information set and the radius information of each second oil tank information in the second oil tank information set, wherein the position information of each second oil tank information in the second oil tank information set and the radius information of each second oil tank information in the second oil tank information set corresponding to the target synthetic aperture radar image are generated according to the target synthetic aperture radar image and the oil tank recognition model set, including: obtaining the priority of each oil tank recognition model in the oil tank recognition model set; selecting an oil tank recognition model whose priority meets the fourth condition from the oil tank recognition model set as the target oil tank recognition model; inputting the target synthetic aperture radar image into the pre-trained target oil tank recognition model to obtain the position information of each third oil tank information in the third oil tank information set information and the radius information of each third oil tank information in the third oil tank information set; input the target synthetic aperture radar image to each oil tank recognition model in at least one oil tank recognition model in the oil tank recognition model set to output the position information of each fourth oil tank information in the fourth oil tank information set and the radius information of each fourth oil tank information in the fourth oil tank information set, and obtain the position information of each fourth oil tank information in the fourth oil tank information set group and the radius information of each fourth oil tank information in the fourth oil tank information set group, wherein the at least one oil tank recognition model is each oil tank recognition model in the oil tank recognition model set without the target oil tank recognition model; for each position information of the third oil tank information in the position information information, performing the following determination steps: determining whether there is location information identical to the location information of the third oil tank information in the location information of each of the fourth oil tank information; in response to determining that there is, screening out the location information identical to the location information of the third oil tank information from the location information of each of the fourth oil tank information as the first target location information, and obtaining at least one first target location information; determining a first number of each of the first target location information corresponding to the at least one first target location information; in response to determining that the first number is greater than or equal to a predetermined threshold, determining the location information of the third oil tank information as the location information of the second oil tank information, and determining the radius information of the third oil tank information as the radius information corresponding to the second oil tank information;The oil tank layout information of the target area is generated according to the position information of each second oil tank information in the second oil tank information set, the radius information of each second oil tank information in the second oil tank information set, the position information of each first oil tank information in the first oil tank information set group, and the radius information of each first oil tank information in the first oil tank information set group, wherein the oil tank layout information of the target area is generated according to the position information of each second oil tank information in the second oil tank information set, the radius information of each second oil tank information in the second oil tank information set, the position information of each first oil tank information in the first oil tank information set group, and the radius information of each first oil tank information in the first oil tank information set group, including: Each of the second oil tank information in the second oil tank information performs the following generation steps to generate the layout information of the oil tank corresponding to the second oil tank information: determine the location information of the second oil tank information; determine the area information corresponding to the location information; determine the first oil tank information set in the first oil tank information set group corresponding to the area information as the target oil tank information set; determine whether there is a first oil tank information with the same geographical location as the second oil tank information in the target oil tank information set; in response to the determination of existence, generate the layout information indicating that the oil tank corresponding to the second oil tank information exists at the location information in the target area; generate the oil tank layout information of the target area according to the obtained layout information. ; 5. An electronic device comprising: one or more processors; a storage device having one or more programs stored thereon, When the one or more programs are executed by the one or more processors, the one or more processors implement the method according to any one of claims 1 to 3.

6. A computer readable medium having a computer program stored thereon, wherein: When the program is executed by a processor, the method according to any one of claims 1 to 3 is implemented.

Images