A method, device and medium for monitoring farmland status for farmland protection

Abstract

The present application discloses a method, device and medium for monitoring the state of cultivated land for cultivated land protection, the method comprising: a first edge computing device acquires remote sensing images and non-remote sensing images of a monitoring area; through a preset image recognition algorithm, the cultivated land area and non-cultivated land area of ​​the remote sensing image are determined, and a three-dimensional regional model corresponding to the monitoring area is constructed according to the cultivated land area and non-cultivated land area of ​​the remote sensing image; and according to a preset rule, the non-remote sensing image is sent to a second edge computing device, and an image label corresponding to the non-remote sensing image returned by the second edge computing device is received; according to the non-remote sensing image and the image label, the three-dimensional regional model is updated, so as to monitor the cultivated land state of the monitoring area through the updated three-dimensional regional model. The first edge computing device can update the three-dimensional regional model constructed by the remote sensing image according to the analysis result of the non-remote sensing image, thereby improving the accuracy of cultivated land state monitoring.

Description

Technical Field

[0001] The present application relates to the field of farmland technology, and in particular to a farmland status monitoring method, equipment and medium for farmland protection. Background Art

[0002] Cultivated land is the main place for human beings to engage in agricultural production activities and provides the main food source for human survival. Therefore, timely and effective arable land extraction plays an important role in formulating scientific and reasonable agricultural planning.

[0003] At present, farmland construction and production conditions are usually monitored by using high-altitude remote sensing. However, no matter what the status is, the satellite orbit and attitude cannot be adjusted quickly, and the cloud thickness and light factors often interfere, which often makes the remote sensing images collected in a single time unusable, or the terrain of the monitoring area may be relatively complex. Monitoring the cultivated land status in the monitoring area only through remote sensing images will have errors, resulting in low accuracy in monitoring the cultivated land status in the monitoring area. Summary of the invention

[0004] The embodiments of the present application provide a method, device and medium for monitoring the state of cultivated land for farmland protection, which are used to solve the problem of low accuracy in monitoring the state of cultivated land in a monitoring area through remote sensing images.

[0005] The present application embodiment adopts the following technical solutions:

[0006] On the one hand, an embodiment of the present application provides a method for monitoring the state of cultivated land for cultivated land protection, the method comprising: a first edge computing device acquires a remote sensing image and a non-remote sensing image of a monitoring area; the non-remote sensing image is a two-dimensional image; through a preset image recognition algorithm, the cultivated land area and the non-cultivated land area of ​​the remote sensing image are determined, and according to the cultivated land area and the non-cultivated land area of ​​the remote sensing image, a three-dimensional regional model corresponding to the monitoring area is constructed; and according to preset rules, the non-remote sensing image is sent to a second edge computing device, and an image label corresponding to the non-remote sensing image returned by the second edge computing device is received; according to the non-remote sensing image and the image label, the three-dimensional regional model is updated to monitor the cultivated land state of the monitoring area through the updated three-dimensional regional model.

[0007] In one example, the updating of the three-dimensional regional model according to the non-remote sensing image and the image label specifically includes: extracting the cultivated land area of ​​the non-remote sensing image according to the image label in the non-remote sensing image; performing an intersection operation on the cultivated land area of ​​the remote sensing image and the cultivated land area of ​​the non-remote sensing image to determine a suspected cultivated land area; the suspected cultivated land area refers to the non-overlapping area between the cultivated land area of ​​the remote sensing image and the cultivated land area of ​​the non-remote sensing image; obtaining the latest historical remote sensing image from a dynamic periodic monitoring library, matching the suspected cultivated land area in the latest historical remote sensing image, and determining the matching area in the latest historical remote sensing image; and updating the three-dimensional regional model according to the matching area.

[0008] In one example, matching the suspected cultivated land area in the latest historical remote sensing image to determine the matching area in the latest historical remote sensing image specifically includes: generating a first minimum rectangular area in the remote sensing image that includes the suspected cultivated land area; calculating the overlapping area between the remote sensing image and the latest historical remote sensing image based on the shooting angle and shooting position of the remote sensing image and the latest historical remote sensing image; and matching the first minimum rectangular area in the overlapping area of ​​the latest historical remote sensing image to determine the matching area in the latest historical remote sensing image.

[0009] In one example, it is determined that the suspected cultivated land area is in the cultivated land area of ​​the non-remote sensing image, and the three-dimensional regional model is updated according to the matching area, specifically including: if the matching area is the non-cultivated land area in the historical remote sensing image, a second minimum rectangular area is generated in the non-remote sensing image that contains the suspected cultivated land area; the position parameters of the second minimum rectangular area in the non-remote sensing image are sent to an unmanned aerial vehicle device, and the unmanned aerial vehicle device is linked to capture a video image of the second minimum rectangular area; the video image is acquired, and the video image is analyzed by a preset image analysis algorithm to determine the image label of the matching area; and the three-dimensional regional model is updated according to the image label of the matching area.

[0010] In one example, after updating the three-dimensional regional model according to the non-remote sensing image and the image label, the method further includes: high-pass filtering the historical remote sensing image to detect edge information in the historical remote sensing image; extracting buildings and roads in the high-pass filtered remote sensing image by inputting the high-pass filtered remote sensing image into a pre-built image classification neural network model to obtain a classified remote sensing image; superimposing the classified remote sensing image with the non-remote sensing image to determine the unsuperimposed buildings and roads in the non-remote sensing image; and determining the illegal buildings and roads in the cultivated land area based on the unsuperimposed buildings and roads.

[0011] In one example, determining the illegal buildings and illegal roads in the cultivated land area based on the non-overlapping buildings and roads specifically includes: determining the boundary longitude and latitude information corresponding to the non-overlapping buildings and roads in the non-remote sensing image; searching the boundary longitude and latitude information in the basic land database to determine the basic land types corresponding to the non-overlapping buildings and roads in the basic land database; if the non-overlapping buildings and roads are inconsistent with the basic land types, determining the non-overlapping buildings and roads as illegal buildings and illegal roads; and issuing warnings for the illegal buildings and illegal roads in the three-dimensional area model according to the first warning rule.

[0012] In one example, after updating the three-dimensional regional model according to the non-remote sensing image and the image label, the method also includes: if the monitored area is a mountainous area, determining whether there are crops in all pixels at any pixel position in the cultivated land area of ​​the remote sensing image and the cultivated land area of ​​the non-remote sensing image; representing the cultivated land area of ​​the remote sensing image and the cultivated land area of ​​the non-remote sensing image as binary images respectively; a pixel value of a first preset value indicates the presence of crops, and a pixel value of a second preset value indicates the absence of crops; adding the binary images of the cultivated land area of ​​the remote sensing image and the cultivated land area of ​​the non-remote sensing image respectively to obtain a sum image; determining the pixel position in the sum image whose pixel value is greater than the set value as a abandoned farmland area, and in the three-dimensional regional model, issuing an early warning for the abandoned farmland area according to the second early warning rule.

[0013] In one example, a time period corresponding to when a remote sensing device inspects the monitoring area is determined; environmental information of the monitoring area within the time period is obtained from a meteorological station; the environmental information is analyzed to determine whether the remote sensing device is blocked by obstacles when inspecting the monitoring area; if so, the non-remote sensing image is sent to a second edge computing device.

[0014] On the other hand, an embodiment of the present application provides a farmland status monitoring device for farmland protection, which is applied to a first edge computing device and includes: at least one processor; and a memory connected to the at least one processor in communication; wherein the memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can: obtain remote sensing images and non-remote sensing images of a monitoring area; the non-remote sensing images are two-dimensional images; determine the farmland area and non-cultivated land area of ​​the remote sensing image through a preset image recognition algorithm, and construct a three-dimensional regional model corresponding to the monitoring area based on the farmland area and non-cultivated land area of ​​the remote sensing image; and according to preset rules, send the non-remote sensing image to a second edge computing device, and receive the image label corresponding to the non-remote sensing image returned by the second edge computing device; update the three-dimensional regional model based on the non-remote sensing image and the image label, so as to perform farmland status monitoring on the monitoring area through the updated three-dimensional regional model.

[0015] On the other hand, an embodiment of the present application provides a non-volatile computer storage medium for monitoring the state of cultivated land for cultivated land protection, storing computer executable instructions, and applied to a first edge computing device, wherein the computer executable instructions are configured as follows: the first edge computing device obtains remote sensing images and non-remote sensing images of a monitoring area; the non-remote sensing images are two-dimensional images; the cultivated land area and non-cultivated land area of ​​the remote sensing image are determined through a preset image recognition algorithm, and a three-dimensional regional model corresponding to the monitoring area is constructed based on the cultivated land area and non-cultivated land area of ​​the remote sensing image; and according to preset rules, the non-remote sensing image is sent to a second edge computing device, and an image label corresponding to the non-remote sensing image returned by the second edge computing device is received; the three-dimensional regional model is updated based on the non-remote sensing image and the image label, so as to perform cultivated land state monitoring on the monitoring area through the updated three-dimensional regional model.

[0016] At least one of the above technical solutions adopted in the embodiments of the present application can achieve the following beneficial effects:

[0017] The first edge computing device in the embodiment of the present application not only analyzes the remote sensing image of the monitoring area and constructs a three-dimensional regional model of the monitoring area, but also obtains non-remote sensing images of the monitoring area and sends the non-remote sensing images to the second edge computing device. The second edge computing device analyzes the non-remote sensing images and feeds back the analysis results to the first edge computing device.

[0018] Therefore, the first edge computing device can update the three-dimensional regional model constructed by the remote sensing image according to the analysis results of the non-remote sensing image, thereby adjusting the cultivated land status in the monitored area and improving the accuracy of cultivated land status monitoring. At the same time, the second edge computing device bears the computational pressure of non-remote sensing image analysis, reducing the computational pressure of the first edge computing device and improving the efficiency of cultivated land monitoring through the three-dimensional regional model. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the technical solution of the present application, some embodiments of the present application will be described in detail below in conjunction with the accompanying drawings, in which:

[0020] Figure 1 A flow chart of a method for monitoring the state of cultivated land for cultivated land protection provided in an embodiment of the present application;

[0021] Figure 2 A schematic diagram of the structure of a farmland status monitoring device for farmland protection provided in an embodiment of the present application. DETAILED DESCRIPTION

[0022] In order to make the purpose, technical solution and advantages of the present application clearer, the technical solution of the present application will be clearly and completely described below in conjunction with specific embodiments and corresponding drawings. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in the field without creative work are within the scope of protection of the present application.

[0023] Some embodiments of the present application are described in detail below with reference to the accompanying drawings.

[0024] Figure 1 A schematic diagram of a process flow of a method for monitoring farmland status for farmland protection provided in an embodiment of the present application. Certain input parameters or intermediate results in the process flow allow for manual intervention and adjustment to help improve accuracy.

[0025] Figure 1 The process in may include the following steps:

[0026] S101: The first edge computing device obtains remote sensing images and non-remote sensing images of a monitoring area.

[0027] Among them, the remote sensing image can be a satellite remote sensing image or an aerial remote sensing image. The non-remote sensing image is obtained by taking a video image of the monitoring area by a drone. In addition, the non-remote sensing image is a two-dimensional image, and the remote sensing image is a three-dimensional image.

[0028] It should be noted that the first edge computing device can be configured according to actual needs and is not limited here.

[0029] S102: Determine the cultivated land area and the non-cultivated land area of ​​the remote sensing image through a preset image recognition algorithm, and construct a three-dimensional regional model corresponding to the monitoring area according to the cultivated land area and the non-cultivated land area of ​​the remote sensing image.

[0030] Before the first edge computing device performs image recognition on the remote sensing image, it first preprocesses the remote sensing image to obtain a preprocessed remote sensing image. The preprocessing includes radiation correction, geometric correction, and cloud removal.

[0031] Then, the preprocessed remote sensing image is input into a preset image recognition algorithm, so as to obtain the cultivated land area and the non-cultivated land area of ​​the remote sensing image in the remote sensing image.

[0032] S103: According to a preset rule, the non-remote sensing image is sent to the second edge computing device, and an image tag corresponding to the non-remote sensing image returned by the second edge computing device is received.

[0033] Among them, after receiving the non-remote sensing image, the second edge computing device inputs the non-remote sensing image into a preset image extraction algorithm to obtain the category of land objects in the non-remote sensing image, and assigns corresponding image labels to the land object categories, thereby feeding back the image labels to the first edge computing device.

[0034] S104: updating the three-dimensional regional model according to the non-remote sensing image and the image label, so as to monitor the cultivated land status in the monitoring area through the updated three-dimensional regional model.

[0035] Among them, the first edge computing device can classify the image labels in the non-remote sensing image, thereby extracting the cultivated land area and the non-cultivated land area of ​​the non-remote sensing image, and then update the three-dimensional regional model through the cultivated land area and the non-cultivated land area of ​​the non-remote sensing image.

[0036] It should be noted that although the embodiments of the present application are based on Figure 1 Steps S101 to S104 are described in sequence, but this does not mean that steps S101 to S104 must be performed in a strict order. Figure 1 The order shown in the figure is to introduce and explain step S101 to step S104 in sequence, in order to facilitate those skilled in the art to understand the technical solution of the embodiment of the present application. In other words, in the embodiment of the present application, the order between step S101 to step S104 can be appropriately adjusted according to actual needs.

[0037] pass Figure 1According to the method, the first edge computing device not only analyzes the remote sensing image of the monitoring area and constructs a three-dimensional regional model of the monitoring area, but also obtains the non-remote sensing image of the monitoring area and sends the non-remote sensing image to the second edge computing device. The second edge computing device analyzes the non-remote sensing image and feeds the analysis result back to the first edge computing device.

[0038] Therefore, the first edge computing device can update the three-dimensional regional model constructed by the remote sensing image according to the analysis results of the non-remote sensing image, thereby adjusting the cultivated land status in the monitored area and improving the accuracy of cultivated land status monitoring. At the same time, the second edge computing device bears the computational pressure of non-remote sensing image analysis, reducing the computational pressure of the first edge computing device and improving the efficiency of cultivated land monitoring through the three-dimensional regional model.

[0039] based on Figure 1 The present application also provides some specific implementation schemes and extension schemes of the method, which will be described below.

[0040] In some embodiments of the present application, in order to reduce unnecessary computing pressure while effectively ensuring the accuracy of remote sensing images, the first edge computing device sends non-remote sensing images to the second edge computing device according to preset rules.

[0041] Specifically, the first edge computing device determines the time period corresponding to the inspection of the monitoring area by the remote sensing device, obtains the environmental information of the monitoring area in the time period from the meteorological station, and analyzes the environmental information to determine whether the remote sensing device is blocked by obstacles when inspecting the monitoring area. If so, the non-remote sensing image is sent to the second edge computing device. If not, the computing performance of the first edge computing device is evaluated, and when the performance exceeds a preset threshold, the non-remote sensing image is sent to the second edge computing device.

[0042] In some embodiments of the present application, when updating the three-dimensional regional model according to the non-remote sensing image and the image label, in order to reduce the computing pressure of the first edge computing device and ensure the accuracy of the update, the first edge computing device extracts the cultivated land area in the non-remote sensing image according to the image label in the non-remote sensing image, and then performs an intersection operation on the cultivated land area in the remote sensing image and the cultivated land area in the non-remote sensing image, thereby obtaining the non-overlapping area between the cultivated land area in the remote sensing image and the cultivated land area in the non-remote sensing image, and the non-overlapping area is used as the suspected cultivated land area. For example, in the remote sensing image, there may be an area with missing information in the remote sensing image due to the influence of meteorological conditions. The missing area will be used as a non-cultivated land area in the remote sensing image. At this time, the non-overlapping area is in the cultivated land area of ​​the non-remote sensing image.

[0043] Then, the first edge computing device obtains the latest historical remote sensing image from the dynamic cycle monitoring library, matches the suspected cultivated land area in the latest historical remote sensing image, and determines the matching area in the latest historical remote sensing image. Finally, the three-dimensional regional model is updated according to the matching area.

[0044] It should be noted that the first edge computing device stores the analyzed remote sensing image in the dynamic period monitoring library, so that it can be used as a reference remote sensing image to monitor the status of cultivated land in the monitoring area.

[0045] Furthermore, since the boundaries of the suspected cultivated land area may vary, in order to improve the accuracy of matching the suspected cultivated land area, first, the first edge computing device will generate a first minimum rectangular area containing the suspected cultivated land area in the remote sensing image.

[0046] Secondly, since the shooting angle and shooting position of the remote sensing image and the latest historical remote sensing image are not the same, in order to reduce the difference between the two remote sensing images and reduce the complexity of retrieving the first minimum rectangular area in the historical remote sensing image, the first edge computing device will calculate the overlapping area of ​​the remote sensing image and the latest historical remote sensing image based on the shooting angle and shooting position of the remote sensing image and the latest historical remote sensing image.

[0047] Then, in the overlapping area of ​​the latest historical remote sensing image, the first minimum rectangular area is matched to determine the matching area in the latest historical remote sensing image.

[0048] In some embodiments of the present invention, it is considered that no matter what the weather conditions are, the satellite orbit and attitude cannot be adjusted quickly, there are no opportunities for multiple shooting, and interference from cloud thickness and light factors often lead to missing remote sensing images. In this case, the suspected cultivated land area is very likely to be the missing cultivated land area.

[0049] At this time, the suspected cultivated land area will be located in the cultivated land area of ​​the non-remote sensing image, that is, the non-remote sensing image can show the missing cultivated land area. In addition, since the meteorological conditions of the latest historical remote sensing image and the meteorological conditions of the currently acquired remote sensing image are likely to be different, the suspected cultivated land area should be matched in the cultivated land area of ​​the latest historical remote sensing image, that is, the matching area is the cultivated land area in the historical remote sensing image.

[0050] Therefore, if it is determined that the suspected cultivated land area is located in the cultivated land area of ​​the non-remote sensing image, and the matching area is the cultivated land area in the historical remote sensing image, the suspected cultivated land area is changed to the cultivated land area in the remote sensing image, and the three-dimensional regional model is updated.

[0051] However, if the first minimum rectangular area is matched in the non-cultivated land area of ​​the latest historical remote sensing image, that is, the matching area is the non-cultivated land area in the historical remote sensing image, at this time, in order to predict the accuracy of the suspected cultivated land area, the first edge computing device will generate a second minimum rectangular area containing the suspected cultivated land area in the non-remote sensing image, and then send the position parameters of the second minimum rectangular area in the non-remote sensing image to the drone device, and link the drone device to shoot a video image of the second minimum rectangular area.

[0052] Then, after acquiring the video image, the video image is analyzed by a preset image analysis algorithm to determine the image label of the matching area, and finally the three-dimensional area model is updated according to the image label of the matching area.

[0053] In some embodiments of the present application, when the suspected cultivated land area is located in the cultivated land area of ​​the remote sensing image, it is possible that the first edge computing device has misidentified the remote sensing image, that is, the non-cultivated land area is identified as the cultivated land area. At this time, the non-remote sensing image is used as a reference image, so it is assumed that the accuracy of the non-remote sensing image is higher than that of the non-remote sensing image.

[0054] Therefore, when the land use type of the matching area of ​​the first minimum rectangular area in the historical remote sensing image is a non-cultivated land area, the suspected cultivated land area is changed to a non-cultivated land area in the remote sensing image, and the three-dimensional regional model is updated.

[0055] However, when the land use type of the matching area of ​​the first minimum rectangular area in the historical remote sensing image is a cultivated land area, the first edge computing device will generate a second minimum rectangular area in the non-remote sensing image that contains the suspected cultivated land area, and then send the position parameters of the second minimum rectangular area in the non-remote sensing image to the drone device, and link the drone device to shoot a video image of the second minimum rectangular area. Then, after acquiring the video image, the video image is analyzed by a preset image analysis algorithm to determine the image label of the matching area, and finally the three-dimensional area model is updated according to the image label of the matching area.

[0056] In some embodiments of the present application, after obtaining an accurate three-dimensional regional model, in order to protect the cultivated land area, illegal buildings and illegal roads in the cultivated land area are monitored, especially in urban cultivated land, where illegal buildings and illegal roads are increasing. In addition, since the non-remote sensing image is a two-dimensional image, it can clearly show the roads and buildings in the current monitoring area, so the non-remote sensing image is used as a reference image.

[0057] Specifically, the first edge computing device will perform high-pass filtering on the historical remote sensing image to detect edge information in the historical remote sensing image. That is, the image edge information can be detected by high-pass filtering, which is conducive to extracting building and road information.

[0058] Then, by inputting the high-pass filtered remote sensing image into the pre-built image classification neural network model, the buildings and roads in the high-pass filtered remote sensing image are extracted to obtain the classified remote sensing image. Then, the classified remote sensing image is superimposed with the non-remote sensing image to determine the unsuperimposed buildings and roads in the non-remote sensing image. Finally, the illegal buildings and illegal roads in the cultivated land area are determined based on the unsuperimposed buildings and roads.

[0059] Furthermore, in order to determine whether the unoverlapping buildings and roads were independently built by illegal persons.

[0060] The first edge computing device determines the boundary longitude and latitude information corresponding to the non-overlapping buildings and roads in the non-remote sensing image, searches the boundary longitude and latitude information in the basic land database, and determines the basic land types corresponding to the non-overlapping buildings and roads in the basic land database.

[0061] Then, if the unoverlapped buildings and roads are inconsistent with the basic land use type, the unoverlapped buildings and roads are determined to be illegal buildings and illegal roads, that is, after the latest historical remote sensing image is obtained, illegal persons have independently built illegal buildings or illegal roads in the cultivated land area.

[0062] Finally, in the three-dimensional area model, warnings are issued for illegal buildings and illegal roads according to the first warning rule, for example, by highlighting illegal buildings and illegal roads by color.

[0063] In some embodiments of the present application, mountainous areas are prone to problems such as drought. Considering the maximization of land use and the promotion of ecological environmental development, the abandoned farmland area is predicted, and the abandoned farmland area is predicted by binary images. The logic is simple and the calculation pressure is reduced.

[0064] Specifically, the first edge computing device determines whether crops exist in all pixels at any pixel position in the cultivated land area of ​​the remote sensing image and the cultivated land area of ​​the non-remote sensing image, and then represents the cultivated land area of ​​the remote sensing image and the cultivated land area of ​​the non-remote sensing image as binary images, respectively. A pixel value of the first preset value indicates the presence of crops, and a pixel value of the second preset value indicates the absence of crops. For example, a pixel value of 1 indicates the presence of crops, and a pixel value of 0 indicates the absence of crops.

[0065] Then, the binary images of the cultivated land area in the remote sensing image and the cultivated land area in the non-remote sensing image are added respectively to obtain a sum image, and the pixel positions in the sum image whose pixel values ​​are greater than the set value are determined as abandoned farmland areas, and in the three-dimensional regional model, an early warning is issued for the abandoned farmland areas according to the second early warning rule.

[0066] Based on the same idea, some embodiments of the present application also provide devices and non-volatile computer storage media corresponding to the above methods.

[0067] Figure 2 A structural diagram of a farmland status monitoring device for farmland protection provided in an embodiment of the present application, applied to a first edge computing device, includes:

[0068] at least one processor; and,

[0069] a memory communicatively connected to the at least one processor; wherein,

[0070] The memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to:

[0071] Obtain remote sensing images and non-remote sensing images of the monitoring area; non-remote sensing images are two-dimensional images;

[0072] Determine the cultivated land area and the non-cultivated land area of ​​the remote sensing image by using a preset image recognition algorithm, and construct a three-dimensional regional model corresponding to the monitoring area according to the cultivated land area and the non-cultivated land area of ​​the remote sensing image; and

[0073] According to a preset rule, the non-remote sensing image is sent to the second edge computing device, and an image label corresponding to the non-remote sensing image returned by the second edge computing device is received;

[0074] The three-dimensional regional model is updated according to the non-remote sensing image and the image label, so as to monitor the cultivated land status in the monitoring area through the updated three-dimensional regional model.

[0075] Some embodiments of the present application provide a non-volatile computer storage medium for monitoring farmland status for farmland protection, storing computer executable instructions, and applied to a first edge computing device, wherein the computer executable instructions are configured as follows:

[0076] Obtain remote sensing images and non-remote sensing images of the monitoring area; non-remote sensing images are two-dimensional images;

[0077] Determine the cultivated land area and the non-cultivated land area of ​​the remote sensing image by using a preset image recognition algorithm, and construct a three-dimensional regional model corresponding to the monitoring area according to the cultivated land area and the non-cultivated land area of ​​the remote sensing image; and

[0078] According to a preset rule, the non-remote sensing image is sent to the second edge computing device, and an image label corresponding to the non-remote sensing image returned by the second edge computing device is received;

[0079] The three-dimensional regional model is updated according to the non-remote sensing image and the image label, so as to monitor the cultivated land status in the monitoring area through the updated three-dimensional regional model.

[0080] Each embodiment in this application is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the device and medium embodiments, since they are basically similar to the method embodiments, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiments.

[0081] The devices and media provided in the embodiments of the present application correspond one-to-one to the methods. Therefore, the devices and media also have similar beneficial technical effects as the corresponding methods. Since the beneficial technical effects of the methods have been described in detail above, the beneficial technical effects of the devices and media will not be repeated here.

[0082] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0083] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0084] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0085] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process in the computer or other programmable device. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0086] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0087] The memory may include non-permanent storage in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.

[0088] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0089] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0090] The above is only an embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the technical principle of the present application should all fall within the protection scope of the present application.

Claims

1. A method for monitoring the state of cultivated land for cultivated land protection, characterized in that: The method comprises: The first edge computing device obtains a remote sensing image and a non-remote sensing image of the monitoring area; the non-remote sensing image is a two-dimensional image; Determine the cultivated land area and the non-cultivated land area of ​​the remote sensing image by using a preset image recognition algorithm, and construct a three-dimensional regional model corresponding to the monitoring area according to the cultivated land area and the non-cultivated land area of ​​the remote sensing image; and According to a preset rule, the non-remote sensing image is sent to a second edge computing device, and an image label corresponding to the non-remote sensing image returned by the second edge computing device is received; The three-dimensional regional model is updated according to the non-remote sensing image and the image label, so as to monitor the cultivated land status in the monitoring area through the updated three-dimensional regional model.

2. The method according to claim 1, characterized in that The updating of the three-dimensional region model according to the non-remote sensing image and the image label specifically includes: In the non-remote sensing image, extracting the cultivated land area of ​​the non-remote sensing image according to the image label; Performing an intersection operation on the cultivated land area of ​​the remote sensing image and the cultivated land area of ​​the non-remote sensing image to determine a suspected cultivated land area; the suspected cultivated land area refers to a non-overlapping area between the cultivated land area of ​​the remote sensing image and the cultivated land area of ​​the non-remote sensing image; Acquire the latest historical remote sensing image from the dynamic cycle monitoring library, match the suspected cultivated land area in the latest historical remote sensing image, and determine the matching area in the latest historical remote sensing image; The three-dimensional region model is updated according to the matching region.

3. The method according to claim 2, characterized in that The matching of the suspected cultivated land area in the latest historical remote sensing image to determine the matching area in the latest historical remote sensing image specifically includes: Generate a first minimum rectangular area in the remote sensing image that includes the suspected cultivated land area; Calculating an overlapping area between the remote sensing image and the latest historical remote sensing image according to the shooting angles and shooting positions of the remote sensing image and the latest historical remote sensing image; In the overlapping area of ​​the latest historical remote sensing image, the first minimum rectangular area is matched to determine the matching area in the latest historical remote sensing image.

4. The method according to claim 2, characterized in that: Determining that the suspected cultivated land area is in the cultivated land area of ​​the non-remote sensing image, and updating the three-dimensional regional model according to the matching area specifically includes: If the matching area is a non-cultivated land area in the historical remote sensing image, generating a second minimum rectangular area in the non-remote sensing image that includes the suspected cultivated land area; Sending the position parameters of the second minimum rectangular area in the non-remote sensing image to the drone device, and linking the drone device to shoot a video image of the second minimum rectangular area; Acquire the video image, analyze the video image using a preset image analysis algorithm, and determine the image label of the matching area; The three-dimensional area model is updated according to the image label of the matching area.

5. The method according to claim 2, characterized in that: After updating the three-dimensional region model according to the non-remote sensing image and the image label, the method further includes: Performing high-pass filtering on the historical remote sensing image to detect edge information in the historical remote sensing image; By inputting the high-pass filtered remote sensing image into a pre-built image classification neural network model, buildings and roads in the high-pass filtered remote sensing image are extracted to obtain a classified remote sensing image; Overlaying the classified remote sensing image with the non-remote sensing image to determine unoverlapped buildings and roads in the non-remote sensing image; According to the non-overlapping buildings and roads, illegal buildings and illegal roads in the cultivated land area are determined.

6. The method according to claim 5, characterized in that The determining of illegal buildings and illegal roads in the cultivated land area according to the non-overlapping buildings and roads specifically includes: In the non-remote sensing image, determining the boundary longitude and latitude information corresponding to the non-overlapping buildings and roads respectively; In the basic land database, the boundary longitude and latitude information is searched to determine the basic land types corresponding to the non-overlapping buildings and roads in the basic land database; If the non-overlapping buildings and roads are inconsistent with the basic land type, the non-overlapping buildings and roads are determined to be illegal buildings and illegal roads; In the three-dimensional area model, an early warning is issued for the illegal buildings and the illegal roads according to a first early warning rule.

7. The method according to claim 1, characterized in that After updating the three-dimensional region model according to the non-remote sensing image and the image label, the method further includes: If the monitoring area is a mountainous area, determining whether crops exist in all pixels at any pixel position in the cultivated land area of ​​the remote sensing image and the cultivated land area of ​​the non-remote sensing image; The cultivated land area of ​​the remote sensing image and the cultivated land area of ​​the non-remote sensing image are represented as binary images respectively; a pixel value of a first preset value indicates the presence of crops, and a pixel value of a second preset value indicates the absence of crops; Adding the binary images of the cultivated land area of ​​the remote sensing image and the cultivated land area of ​​the non-remote sensing image respectively to obtain a sum image; The pixel position in the sum image whose pixel value is greater than the set value is determined as the abandoned farmland area, and in the three-dimensional regional model, an early warning is issued for the abandoned farmland area according to the second early warning rule.

8. The method according to claim 1, characterized in that The sending the non-remote sensing image to the second edge computing device according to a preset rule specifically includes: Determine a time period corresponding to when the remote sensing equipment inspects the monitoring area; Obtaining environmental information of the monitoring area during the time period from the meteorological station; Analyzing the environmental information to determine whether the remote sensing device is blocked by obstacles when inspecting the monitoring area; If so, the non-remote sensing image is sent to the second edge computing device.

9. A farmland status monitoring device for farmland protection, characterized in that: Applied to the first edge computing device, including: at least one processor; and, a memory communicatively connected to the at least one processor; wherein, The memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to: Acquire remote sensing images and non-remote sensing images of the monitoring area; the non-remote sensing images are two-dimensional images; Determine the cultivated land area and the non-cultivated land area of ​​the remote sensing image by using a preset image recognition algorithm, and construct a three-dimensional regional model corresponding to the monitoring area according to the cultivated land area and the non-cultivated land area of ​​the remote sensing image; and According to a preset rule, the non-remote sensing image is sent to a second edge computing device, and an image label corresponding to the non-remote sensing image returned by the second edge computing device is received; The three-dimensional regional model is updated according to the non-remote sensing image and the image label, so as to monitor the cultivated land status in the monitoring area through the updated three-dimensional regional model.

10. A non-volatile computer storage medium for monitoring farmland status for farmland protection, storing computer executable instructions, characterized in that: Applied to a first edge computing device, the computer executable instructions are configured to: Acquire remote sensing images and non-remote sensing images of the monitoring area; the non-remote sensing images are two-dimensional images; Determine the cultivated land area and the non-cultivated land area of ​​the remote sensing image by using a preset image recognition algorithm, and construct a three-dimensional regional model corresponding to the monitoring area according to the cultivated land area and the non-cultivated land area of ​​the remote sensing image; and According to a preset rule, the non-remote sensing image is sent to a second edge computing device, and an image label corresponding to the non-remote sensing image returned by the second edge computing device is received; The three-dimensional regional model is updated according to the non-remote sensing image and the image label, so as to monitor the cultivated land status in the monitoring area through the updated three-dimensional regional model.

Images