Method and device for detecting high-altitude cirrus cloud, electronic equipment and storage medium

Abstract

Embodiments of the present application relate to the technical field of image processing, in particular to a high-altitude cirrus cloud detection method and device, electronic equipment and storage medium. The method comprises: acquiring a first infrared image and a second infrared image, the first infrared image and the second infrared image being adjacent two frames of infrared images; constructing a first image pyramid of the first infrared image and a second image pyramid of the second infrared image, the number of layers of the first image pyramid and the second image pyramid and the size of each layer of image being the same; based on the correlation of the first image pyramid and the second image pyramid in space-time, obtaining the optical flow vector field of the first infrared image by using the optical flow method; performing threshold segmentation and morphological operation on the optical flow vector field of the first infrared image to obtain the detection result of the first infrared image. The scheme can improve the detection accuracy of high-altitude cirrus cloud and has low false detection rate.

Description

TECHNICAL FIELD

[0001] Embodiments of the present application relate to the technical field of image processing, and particularly relate to a high-altitude cirrus cloud detection method and device, electronic equipment and storage medium. BACKGROUND

[0002] An infrared detection system can realize remote detection and identification of a target, is not affected by obstacles such as smoke and fog, and can detect the target all day long. Therefore, the infrared detection system is widely used in military and civilian fields. For the infrared detection system, high-altitude cirrus cloud is a typical false alarm source. Due to the influence of factors such as light, wind, season, and air pressure, the high-altitude cirrus cloud has great uncertainty in time and space distribution, and detection is difficult.

[0003] In the prior art, high-altitude cirrus cloud cannot be accurately detected, and the false detection rate is high.

[0004] Therefore, there is an urgent need for a high-altitude cirrus cloud detection method, device, electronic equipment and storage medium to solve the above technical problems. SUMMARY

[0005] Based on the problem of poor detection accuracy of high-altitude cirrus cloud in related technologies, embodiments of the present application provide a high-altitude cirrus cloud detection method, device, electronic equipment and storage medium, which can improve the detection accuracy of high-altitude cirrus cloud and has a low false detection rate.

[0006] In a first aspect, the embodiments of the present application provide a high-altitude cirrus cloud detection method, comprising:

[0007] obtaining a first infrared image and a second infrared image, the first infrared image and the second infrared image being adjacent two frames of infrared images;

[0008] constructing a first image pyramid of the first infrared image and a second image pyramid of the second infrared image, the number of layers and the size of each layer of the first image pyramid and the second image pyramid being the same;

[0009] obtaining an optical flow vector field of the first infrared image by using an optical flow method based on the correlation of the first image pyramid and the second image pyramid in space-time;

[0010] performing threshold segmentation and morphological operation on the optical flow vector field of the first infrared image to obtain a detection result of the first infrared image.

[0011] In a possible design, the first image pyramid of the first infrared image and the second image pyramid of the second infrared image are constructed by the following method:

[0012] The first infrared image and the second infrared image are downsampled according to a preset rule until a preset number of layers are sampled, thus obtaining the first image pyramid and the second image pyramid.

[0013] In one possible design, obtaining the optical flow vector field of the first infrared image using optical flow based on the first image pyramid and the second image pyramid includes:

[0014] An optical flow field energy function is constructed for the first image pyramid. The energy function includes a data term and a smoothing term. The smoothing term is constructed based on the spatiotemporal correlation between the first image pyramid and the second image pyramid.

[0015] Solve for the optical flow field energy function, and then solve for the optical flow vector field of each layer of the first image pyramid in turn;

[0016] The optical flow vector field of the lowest layer image obtained by solving is used as the optical flow vector field of the first infrared image.

[0017] In one possible design, the data item is constructed based on the assumptions of constant brightness and spatial consistency, and the expression of the data item is as follows:

[0018]

[0019] In the formula, Let β = 0.001 be the penalty function, and V = [u(x,y,t),v(x,y,t),1] T Let ρ represent the optical flow vector of the pixel at coordinates (x, y) in the image at time t, and let J be the size of the local neighborhood centered on the current pixel. ρ It is a local spatiotemporal derivative smoothing term. J is the image brightness gradient vector. ρ and The expressions are as follows:

[0020]

[0021]

[0022] In the formula, f = f(x,y,t) is the gray value of the pixel at coordinates (x,y) in the image at time t, and G ρ The Gaussian kernel is ρ;

[0023] The expression for the smoothing term is as follows:

[0024]

[0025] In the formula, and respectively represent the horizontal and vertical velocity change rates of a pixel point at (x, y) coordinates in two adjacent frames of images, and is a regularization coefficient represents a smoothing factor of the optical flow field, and C and ε are fixed constants.

[0026] The optical flow field energy function of the first image pyramid is:

[0027]

[0028] In the formula, a is the size of a square region in the image.

[0029] In a possible design, the solving of the optical flow field energy function sequentially solves the optical flow vector field of each layer of image in the first image pyramid, and includes:

[0030] The each layer of image in the first image pyramid is sequentially traversed in the order from small to large according to the image feature scale, and the following is performed for each layer of image traversed:

[0031] The optical flow field energy function is solved to obtain the optical flow vector increment of the layer of image;

[0032] The sum of the optical flow vector increment of the layer of image and the initial value of the optical flow vector of the layer of image is taken as the optical flow vector field of the layer of image, and the optical flow vector field of the layer of image is taken as the initial value of the optical flow vector of the next layer of image, and the above is sequentially performed until the optical flow vector field of the bottom layer of image of the first image pyramid is solved.

[0033] In a possible design, the optical flow field energy function is solved by an Euler-Lagrange equation.

[0034] In a possible design, the threshold segmentation and morphological operation on the optical flow vector field of the first infrared image to obtain the detection result of the first infrared image include:

[0035] According to the optical flow vector field of the first infrared image, the gray value of each pixel point in the first infrared image is determined.

[0036] According to the relationship between the gray value of each pixel point and the segmentation threshold, the first infrared image is threshold segmented to obtain a binary image of the first infrared image.

[0037] The binary image is morphologically operated to obtain the detection result of the first infrared image.

[0038] In a second aspect, the embodiments of the present application further provide a high-altitude cirrus detection device, which includes:

[0039] The acquisition module is configured to acquire a first infrared image and a second infrared image, the first infrared image and the second infrared image being two adjacent infrared images;

[0040] The construction module is configured to construct a first image pyramid of the first infrared image and a second image pyramid of the second infrared image, the first image pyramid and the second image pyramid having the same number of layers and the same size of each layer image;

[0041] The solving module is configured to obtain an optical flow vector field of the first infrared image by using an optical flow method based on the correlation of the first image pyramid and the second image pyramid in space-time;

[0042] The segmentation module is configured to perform threshold segmentation and morphological operation on the optical flow vector field of the first infrared image to obtain a detection result of the first infrared image.

[0043] In a third aspect, an electronic device is provided, including a memory and a processor, the memory stores a computer program, and the processor executes the computer program to implement the method in any of the embodiments.

[0044] In a fourth aspect, a computer readable storage medium is provided, which stores a computer program, and the computer program, when executed in a computer, causes the computer to perform the method in any of the embodiments.

[0045] The embodiments of the present application provide a high-altitude cirrus detection method, device, electronic device and storage medium. The method first acquires a first infrared image and a second infrared image of two adjacent frames. Since the two infrared images have correlation in space-time, each layer image in the first image pyramid and the second image pyramid constructed has correlation in space-time, so that a plurality of scale images related to each other are formed. Then, based on the correlation of the first image pyramid and the second image pyramid in space-time, the optical flow (speed) field of each layer image in the first image pyramid is calculated from coarse to fine by using the optical flow method, which can avoid the problem that the target does not meet the brightness constant assumption of the optical flow method under large and discontinuous motion. Therefore, the optical flow vector field of the first infrared image obtained by using the method of the present application is more accurate, and the detection result obtained by using the optical flow vector field for segmentation is also more accurate. It can be seen that the method provided by the present application can improve the detection accuracy of high-altitude cirrus and has low false detection rate. BRIEF DESCRIPTION OF DRAWINGS

[0046] In order to make the technical solutions in the above-described embodiments or prior art clearer, the accompanying drawings needed to be used in the embodiments or prior art description will be briefly introduced. Obviously, the accompanying drawings in the following description are only some embodiments of the present application, and all other drawings obtained by those of ordinary skill in the art without creative work based on these drawings also belong to the protection scope of the present application.

[0047] Figure 1 is a flow chart of a high-altitude cirrus cloud detection method provided by an embodiment of the present application;

[0048] Figure 2 is a first infrared image provided by an embodiment of the present application;

[0049] Figure 3 is a second infrared image provided by an embodiment of the present application;

[0050] Figure 4 is a principle schematic diagram of a multi-scale algorithm provided by an embodiment of the present application;

[0051] Figure 5 is an optical flow vector diagram of the first infrared image shown in Figure 2

[0052] Figure 6 is a hardware architecture diagram of an electronic device provided by an embodiment of the present application;

[0053] Figure 7 is a structural diagram of a high-altitude cirrus cloud detection device provided by an embodiment of the present application. DETAILED DESCRIPTION

[0054] In order to make the technical solutions in the above-described embodiments or prior art clearer, the accompanying drawings needed to be used in the embodiments or prior art description will be briefly introduced. Obviously, the accompanying drawings in the following description are only some embodiments of the present application, and all other drawings obtained by those of ordinary skill in the art without creative work based on these drawings also belong to the protection scope of the present application.

[0055] The specific implementation of the above concept will be described below.

[0056] Please refer to Figure 1 The embodiment of the present application provides a high-altitude cirrus cloud detection method, which comprises the following steps.

[0057] In step 100, a first infrared image and a second infrared image are acquired, and the first infrared image and the second infrared image are adjacent two frames of infrared images.

[0058] ​Step 102, constructing a first image pyramid of the first infrared image and a second image pyramid of the second infrared image, the number of layers and the size of each layer image of the first image pyramid and the second image pyramid are the same;

[0059] Step 104, obtaining the optical flow vector field of the first infrared image by using the optical flow method based on the correlation of the first image pyramid and the second image pyramid in space-time;

[0060] Step 106, performing threshold segmentation and morphological operation on the optical flow vector field of the first infrared image to obtain the detection result of the first infrared image.

[0061] In the embodiment of the present application, first, the first infrared image and the second infrared image of the adjacent two frames are acquired, since the two frames of infrared images have correlation in space-time, therefore, each layer image in the first image pyramid and the second image pyramid constructed has correlation in space-time, thus, the multi-scale images which are related to each other are formed. Then, based on the correlation of the first image pyramid and the second image pyramid in space-time, the optical flow (velocity) field of each layer image in the first image pyramid is calculated in turn by using the optical flow method from coarse to fine, which can avoid the problem that the target does not meet the brightness constant assumption of the optical flow method under large and incoherent motion. Therefore, the optical flow vector field of the first infrared image obtained by using the method of the present application is more accurate, and the detection result obtained by using the optical flow vector field for segmentation is also more accurate. As can be seen, the method provided by the present application can improve the detection accuracy of high-altitude cirrus cloud, and the false detection rate is low.

[0062] Next, taking the first infrared image and the second infrared image shown in Figure 2 and Figure 3 as an example, the execution mode of each step shown in Figure 1 is described.

[0063] First, for step 100, the first infrared image and the second infrared image are acquired, the first infrared image and the second infrared image are adjacent two frames of infrared images.

[0064] In this step, since the two frames of infrared images are adjacent images obtained continuously, therefore, they have correlation in space-time. Thus, based on the rate change of the corresponding pixel points in the two frames of infrared images, the optical flow vector field of the first infrared image is calculated by using the optical flow method.

[0065] Then, for step 102, the first image pyramid of the first infrared image and the second image pyramid of the second infrared image are constructed, the number of layers and the size of each layer image of the first image pyramid and the second image pyramid are the same.

[0066] The principle of each image pyramid and multi-scale algorithm is as shown in Figure 4The neighborhood size of each pixel point in each layer image is equal to meet the spatial consistency assumption and the velocity vector field smoothness assumption. In addition, the number of layers of each image pyramid is at least greater than 3 layers, the more the number of layers, the higher the calculation accuracy, but the calculation speed also increases, and the number of layers of the image pyramid is not specifically limited in the present application.

[0067] In some embodiments, the first image pyramid of the first infrared image and the second image pyramid of the second infrared image are constructed by the following way:

[0068] The first infrared image and the second infrared image are respectively down-sampled according to a preset rule until a preset number of layers are sampled to obtain the first image pyramid and the second image pyramid.

[0069] In some embodiments, the preset rule can be row-by-row and column-by-column, that is, the first infrared image and the second infrared image are respectively down-sampled in a row-by-row and column-by-column manner. Assuming that the number of layers of the image pyramid is N layers, the original infrared image is denoted as L0, that is, the bottom layer image, and the image scale of this layer is the largest; the image at the top layer is denoted as LN, and the image scale of this layer is the smallest. N

[0070] For step 104, in some embodiments, based on the spatio-temporal correlation of the first image pyramid and the second image pyramid, the optical flow vector field of the first infrared image is obtained by using the optical flow method, including:

[0071] An optical flow field energy function of the first image pyramid is constructed, the energy function includes a data item and a smoothing item, and the smoothing item is constructed based on the spatio-temporal correlation of the first image pyramid and the second image pyramid;

[0072] The optical flow field energy function is solved to sequentially solve the optical flow vector field of each layer image in the first image pyramid;

[0073] The solved optical flow vector field of the bottom layer image is taken as the optical flow vector field of the first infrared image. The optical flow vector field of the first infrared image is shown in FIG. 4. Figure 5

[0074] In this embodiment, the optical flow field energy function is composed of a data item and a smoothing item. The data item meets the brightness constant assumption and the spatial consistency assumption, that is, the gray value of the same point in the adjacent two images is unchanged, and all points in the local neighborhood range centered on the point have the same optical flow loss. The data item reflects the local characteristics of the optical flow field, and the smoothing item reflects the global characteristics of the optical flow field.

[0075] In some embodiments, in order to better process the discontinuous part in the optical flow field, the data item adopts a penalty function ψ(x), and the expression of the data item is as follows: ​​

[0076]

[0077] wherein, is a penalty function, β = 0.001, V = [u(x, y, t), v(x, y, t), 1] T represents the optical flow vector of the pixel point at (x, y) coordinates of the image at time t, ρ is the size of the local neighborhood centered on the current pixel point, J ρ is a local spatio-temporal derivative smoothing term, is the image brightness gradient vector, J ρ and The expressions of and are respectively:

[0078]

[0079]

[0080] wherein, f = f(x, y, t) is the gray value of the pixel point at (x, y) coordinates of the image at time t, G ρ is the Gaussian kernel of ρ.

[0081] In some embodiments, adjacent pixels often have similar motion, therefore, in order to make the optical flow field have better noise robustness, and also more accurately describe the target in the place where the optical flow is discontinuous, and reduce the complexity of calculation, the smoothing term is designed as a weighted l1 norm, instead of the previous l2 norm. Therefore, the expression of the smoothing term is as follows:

[0082]

[0083] wherein, and respectively represent the velocity change rate of the pixel point at (x, y) coordinates in the horizontal direction and the vertical direction in the adjacent two frames of images, the regularization coefficient represents the smoothing factor of the optical flow field, C and ε are fixed constants;

[0084] The optical flow field energy function of the first image pyramid is:

[0085]

[0086] wherein, a is the size of the square region in the image.

[0087] This embodiment improves the ability of infrared high-altitude cirrus cloud detection, reduces the false alarm rate, and reduces the algorithm complexity by adopting the local and global combined manner, and designing the l1 norm smoothing term, while ensuring the optical flow density and improving the robustness to noise.

[0088] It can be understood that each layer of image in the image pyramid is composed of at least one square region with a size of a, and when the image contains multiple a regions, the optical flow field energy function is obtained by accumulating the energy functions of multiple regions.

[0089] In some embodiments, the optical flow field energy function is solved, and the optical flow vector field of each layer of image in the first image pyramid is sequentially solved, including:

[0090] Each layer of image in the first image pyramid is sequentially traversed in the order of the image feature scale from small to large, and for each layer of image traversed, the following is performed:

[0091] The optical flow field energy function is solved to obtain the optical flow vector increment of the layer of image; the optical flow vector increment of the layer of image is the minimum solution of the optical flow field energy function of the layer of image;

[0092] The sum of the optical flow vector increment of the layer of image and the initial value of the optical flow vector of the layer of image is taken as the optical flow vector field of the layer of image, and the optical flow vector field of the layer of image is taken as the initial value of the optical flow vector of the next layer of image, and the operation is sequentially repeated until the optical flow vector field of the bottom layer of image of the first image pyramid is solved.

[0093] In some embodiments, the optical flow field energy function is solved by Euler-Lagrange equation. The minimum value of each layer of optical flow field energy function is solved by Euler-Lagrange equation, that is, the increment of the optical flow vector of the layer is obtained, and the optical flow vector field of the original image is obtained by sequentially solving the optical flow vector field of each layer of image from the top layer to the bottom layer. N The increment of the optical flow vector field is calculated from the top layer (the top layer), and the sum of the optical flow vector field increment and the initial value of the layer is taken as the initial value of the optical flow vector field of the next layer. N-1 The operation is repeated until the resolution of the original image L0 layer is obtained, and the optical flow vector field d=[u,v] of the original image is obtained. T u and v are the motion velocities of the pixel point in the horizontal direction and the vertical direction, respectively.

[0094] Finally, for step 106, including:

[0095] According to the optical flow vector field of the first infrared image, the gray value of each pixel point in the first infrared image is determined;

[0096] According to the relationship between the gray value of each pixel point and the segmentation threshold, the first infrared image is threshold segmented to obtain a binary image of the first infrared image.

[0097] The binary image is subjected to morphological operation to obtain the detection result of the first infrared image.

[0098] In this step, the gray value of each pixel point in the first infrared image is calculated, that is, the optical flow loss size of each point in the first infrared image The segmentation threshold T is determined by the maximum inter-class variance method, the gray value of the pixel point is set to 1 when |d|>T, and the gray value of the pixel point is set to 0 when |d|<T, a binary image is obtained after segmentation, and the region with the gray value set to 1 represents the high-level cirrus cloud; finally, the high-level cirrus cloud target is extracted by using the opening and closing operation of morphology.

[0099] As shown in Figure 6 , Figure 7 , the embodiment of the application provides a high-level cirrus cloud detection device. The device embodiment can be realized by software, or realized by hardware or a combination of software and hardware. From the hardware layer, as shown in Figure 6 , it is a hardware architecture diagram of an electronic device where the high-level cirrus cloud detection device provided by the embodiment of the application is located. In addition to the processor, memory, network interface and non-volatile memory shown in Figure 6 , the electronic device where the device in the embodiment is usually also includes other hardware, such as a forwarding chip responsible for processing packets and the like. Taking software implementation as an example, as shown in Figure 7 , as a logically meaningful device, it is formed by the CPU of the electronic device where it is located to read the corresponding computer program in the non-volatile memory into the memory for running. The high-level cirrus cloud detection device provided by the embodiment includes:

[0100] The acquisition module 700 is configured to acquire a first infrared image and a second infrared image, the first infrared image and the second infrared image being adjacent two frames of infrared images.

[0101] The construction module 702 is configured to construct a first image pyramid of the first infrared image and a second image pyramid of the second infrared image, the number of layers of the first image pyramid and the second image pyramid and the size of each layer of image being the same.

[0102] The solving module 704 is configured to obtain a light flow vector field of the first infrared image based on the correlation of the first image pyramid and the second image pyramid in space-time by using a light flow method.

[0103] The segmentation module 706 is configured to perform threshold segmentation and morphological operation on the light flow vector field of the first infrared image to obtain a detection result of the first infrared image.

[0104] In the embodiment of the application, the acquisition module 700 can be used to execute step 100 in the above-mentioned method embodiment, the construction module 702 can be used to execute step 102 in the above-mentioned method embodiment, the solving module 704 can be used to execute step 104 in the above-mentioned method embodiment, and the segmentation module 706 can be used to execute step 106 in the above-mentioned method embodiment.

[0105] In some embodiments, the first image pyramid of the first infrared image and the second image pyramid of the second infrared image are constructed by:

[0106] down-sampling the first infrared image and the second infrared image respectively according to a preset rule until a preset number of layers is reached, to obtain the first image pyramid and the second image pyramid.

[0107] In some embodiments, the solving module 704 is configured to perform:

[0108] constructing an optical flow field energy function of the first image pyramid, the energy function comprising a data term and a smoothing term, the smoothing term being constructed based on the spatio-temporal correlation between the first image pyramid and the second image pyramid;

[0109] solving the optical flow field energy function to sequentially solve the optical flow vector field of each layer of image in the first image pyramid;

[0110] taking the optical flow vector field of the bottom layer of image as the optical flow vector field of the first infrared image.

[0111] In some embodiments, the data term is constructed based on the constant brightness assumption and the spatial consistency assumption, and the expression of the data term is as follows:

[0112]

[0113] wherein, is a penalty function, β = 0.001, V = [u(x, y, t), v(x, y, t), 1] T represents the optical flow vector of the pixel point at (x, y) coordinates of the image at t time, ρ is the size of the local neighborhood centered on the current pixel point, J ρ is the local spatio-temporal derivative smoothing term, is the image brightness gradient vector, J ρ and The expressions of and are as follows:

[0114]

[0115]

[0116] wherein, f = f(x, y, t) is the gray value of the pixel point at (x, y) coordinates of the image at t time, G ρ is the Gaussian kernel of ρ;

[0117] The expression of the smoothing term is as follows:

[0118]

[0119] wherein, and respectively represent the horizontal and vertical velocity change rates of the pixel point at (x, y) coordinates in two adjacent frames of images, and is a regularization coefficient represents a smoothing factor of the optical flow field, and C and ε are fixed constants.

[0120] The optical flow field energy function of the first image pyramid is:

[0121]

[0122] In the formula, a is the size of a square region in the image.

[0123] In some embodiments, the optical flow field energy function is solved, and the optical flow vector field of each image in the first image pyramid is sequentially solved, including:

[0124] Each image in the first image pyramid is sequentially traversed in order of the image feature scale from small to large, and the following is performed for each traversed image:

[0125] The optical flow field energy function is solved to obtain the optical flow vector increment of the image;

[0126] The sum of the optical flow vector increment of the image and the initial value of the optical flow vector of the image is taken as the optical flow vector field of the image, and the optical flow vector field of the image is taken as the initial value of the optical flow vector of the next image, and the same is sequentially applied until the optical flow vector field of the bottommost image of the first image pyramid is solved.

[0127] In some embodiments, the optical flow field energy function is solved by Euler-Lagrange equation.

[0128] In some embodiments, the segmentation module 706 is configured to perform:

[0129] According to the optical flow vector field of the first infrared image, the gray value of each pixel point in the first infrared image is determined;

[0130] According to the relationship between the gray value of each pixel point and the segmentation threshold, the first infrared image is threshold segmented to obtain a binary image of the first infrared image.

[0131] The binary image is subjected to morphological operation to obtain a detection result of the first infrared image.

[0132] It can be understood that the structure shown in the embodiments of the present application does not constitute a specific limitation on the high-altitude cirrus cloud detection device. In other embodiments of the present application, a high-altitude cirrus cloud detection device can include more or fewer components than the diagram, or combine certain components, or split certain components, or different component arrangements. The components shown can be implemented in hardware, software, or a combination of software and hardware.

[0133] The information interaction, execution process and the like between the modules in the device are based on the same concept as the method embodiments of the present application, and the specific content can be referred to the description in the method embodiments of the present application, which will not be repeated here.

[0134] The embodiment of the present application further provides an electronic device, including a memory and a processor, the memory stores a computer program, and the processor implements the detection method of the high-altitude cirrus cloud in any embodiment of the present application when executing the computer program.

[0135] The embodiment of the present application further provides a computer readable storage medium, the computer readable storage medium stores a computer program, and the computer program makes the processor execute the detection method of the high-altitude cirrus cloud in any embodiment of the present application when being executed by the processor.

[0136] Specifically, a system or device equipped with a storage medium can be provided, the storage medium stores software program codes for realizing the functions of any embodiment of the above embodiments, and the computer (or CPU or MPU) of the system or device reads and executes the program codes stored in the storage medium.

[0137] In this case, the program codes read from the storage medium can realize the functions of any embodiment of the above embodiments, and therefore the program codes and the storage medium storing the program codes constitute a part of the present application.

[0138] The storage medium embodiments for providing the program codes include floppy disks, hard disks, magneto-optical disks, optical disks (such as CD-ROM, CD-R, CD-RW, DVD-ROM, DVD-RAM, DVD-RW, DVD+RW), magnetic tapes, nonvolatile memory cards and ROMs. Alternatively, the program codes can be downloaded from a server computer through a communication network.

[0139] In addition, it should be clear that not only the program codes read by the computer can be executed, but also the operating system and the like operating on the computer can be made to perform part or all of the actual operations based on the instructions of the program codes, so as to realize the functions of any embodiment of the above embodiments.

[0140] In addition, it can be understood that the program codes read from the storage medium can be written into the memory provided in the expansion board inserted into the computer or the memory provided in the expansion module connected to the computer, and then the CPU and the like installed on the expansion board or the expansion module can be made to perform part or all of the actual operations based on the instructions of the program codes, so as to realize the functions of any embodiment of the above embodiments.

[0141] It should be noted that, in the present document, relational terms such as first and second and the like can be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Moreover, the terms "comprises", "comprising", or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but can include other elements not expressly listed or inherent to such process, method, article, or apparatus. An element proceeded by "comprises... a" does not, without more constraints, exclude the existence of additional identical elements in the process, method, article, or apparatus that comprises the element.

[0142] Those skilled in the art can understand that all or part of the steps of the above-mentioned method embodiments can be completed by program instruction related hardware, and the foregoing program can be stored in a computer readable storage medium, and the program performs the steps of the above-mentioned method embodiments when executed; and the foregoing storage medium includes various storage media that can store program codes, such as ROM, RAM, magnetic disk or optical disk.

[0143] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacement for part of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for detecting high-altitude cirrus clouds, characterized in that, include: Acquire a first infrared image and a second infrared image, wherein the first infrared image and the second infrared image are two adjacent frames of infrared images; A first image pyramid of the first infrared image and a second image pyramid of the second infrared image are constructed, wherein the number of layers and the size of each layer of the first image pyramid and the second image pyramid are the same; Based on the spatiotemporal correlation between the first image pyramid and the second image pyramid, the optical flow vector field of the first infrared image is obtained using the optical flow method. Threshold segmentation and morphological operations are performed on the optical flow vector field of the first infrared image to obtain the detection result of the first infrared image; The step of obtaining the optical flow vector field of the first infrared image based on the spatiotemporal correlation between the first image pyramid and the second image pyramid using the optical flow method includes: An optical flow field energy function is constructed for the first image pyramid. The energy function includes a data term and a smoothing term. The smoothing term is constructed based on the spatiotemporal correlation between the first image pyramid and the second image pyramid. Solve for the optical flow field energy function, and then solve for the optical flow vector field of each layer of the first image pyramid in turn; The optical flow vector field of the lowest layer image obtained by solving is used as the optical flow vector field of the first infrared image; The data item is constructed based on the assumptions of constant brightness and spatial consistency, and its expression is as follows: In the formula, Let β = 0.001 be the penalty function, and V = [u(x,y,t),v(x,y,t),1] T Let ρ represent the optical flow vector of the pixel at coordinates (x, y) in the image at time t, and let J be the size of the local neighborhood centered on the current pixel. ρ It is a local spatiotemporal derivative smoothing term. J is the image brightness gradient vector. ρ and The expressions are as follows: In the formula, f = f(x,y,t) is the gray value of the pixel at coordinates (x,y) in the image at time t, and G ρ The Gaussian kernel is ρ; The expression for the smoothing term is as follows: In the formula, and These represent the rate of change of velocity of a pixel at coordinates (x, y) in the horizontal and vertical directions between two adjacent frames, respectively, and the regularization coefficient. The smoothing factor represents the optical flow field, where C and ε are fixed constants; The optical flow energy function of the first image pyramid is: In the formula, 'a' represents the size of the square region in the image.

2. The method according to claim 1, characterized in that, The first image pyramid of the first infrared image and the second image pyramid of the second infrared image are constructed in the following manner: The first infrared image and the second infrared image are downsampled according to a preset rule until a preset number of layers are sampled, thus obtaining the first image pyramid and the second image pyramid.

3. The method according to claim 1, characterized in that, The process of solving the optical flow field energy function, and sequentially solving the optical flow vector field of each layer of the first image pyramid, includes: Traverse each layer of the first image pyramid sequentially in ascending order of image feature scale, and perform the following for each traversed layer: Solve for the optical flow field energy function to obtain the optical flow vector increment of the image layer; The sum of the optical flow vector increment of the image layer and the initial value of the optical flow vector of the image layer is used as the optical flow vector field of the image layer. The optical flow vector field of the image layer is used as the initial value of the optical flow vector of the next image layer, and so on, until the optical flow vector field of the bottom layer of the first image pyramid is solved.

4. The method according to claim 1, characterized in that, The optical flow field energy function is obtained by solving the Euler-Lagrange equations.

5. The method according to claim 1, characterized in that, The step of performing threshold segmentation and morphological operations on the optical flow vector field of the first infrared image to obtain the detection result of the first infrared image includes: The grayscale value of each pixel in the first infrared image is determined based on the optical flow vector field of the first infrared image. Based on the relationship between the gray value of each pixel and the segmentation threshold, the first infrared image is thresholded to obtain a binary image of the first infrared image. Morphological operations are performed on the binary image to obtain the detection result of the first infrared image.

6. A device for detecting high-altitude cirrus clouds, characterized in that, The apparatus for implementing the method as described in any one of claims 1-5 comprises: The acquisition module is used to acquire a first infrared image and a second infrared image, wherein the first infrared image and the second infrared image are two adjacent frames of infrared images; A construction module is used to construct a first image pyramid of the first infrared image and a second image pyramid of the second infrared image, wherein the number of layers and the size of each layer of the first image pyramid and the second image pyramid are the same; The solution module is used to obtain the optical flow vector field of the first infrared image based on the spatiotemporal correlation between the first image pyramid and the second image pyramid using the optical flow method; The segmentation module is used to perform threshold segmentation and morphological operations on the optical flow vector field of the first infrared image to obtain the detection result of the first infrared image.

7. An electronic device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the method as described in any one of claims 1-5.

8. A storage medium having a computer program stored thereon, characterized in that, When the computer program is executed in the computer, it causes the computer to perform the method of any one of claims 1-5.

Images