Ground feature change detection technology based on unmanned aerial vehicle remote sensing image

Abstract

The invention discloses a ground object change detection technology based on an unmanned aerial vehicle remote sensing image. The method comprises the following steps: shooting a group of remote sensing images of a target area through an unmanned aerial vehicle according to acquired aerial photography parameters to obtain a current remote sensing image group of the target area; according to the GPS information of each remote sensing image in the current remote sensing image group and the last remote sensing image group, matching to obtain a remote sensing image pair set of two remote sensing images with the same shooting points in the two remote sensing image groups; performing registration and geometric correction on each remote sensing image pair in the remote sensing image pair set to obtain a to-be-used remote sensing image pair set; performing change detection on two remote sensing images in the remote sensing image pair set to be used by using a change recognition algorithm, and recording and storing a ground object change detection result of each shooting point; and combining the vector files of all the shooting points of which the change detection results are changed, and generating a vector file of the ground feature change of the whole target area. The calculation amount is greatly reduced, and the application scene is wide.

Description

Technical Field

[0001] The present invention relates to the technical field of ground object change detection, and in particular to a ground object change detection method, device and computer-readable storage medium based on unmanned aerial vehicle remote sensing images. Background Art

[0002] Remote sensing image-based ground object change detection is an image analysis technology that compares remote sensing images taken at different times of the target area. By comparing the changes in two remote sensing images, it is possible to identify whether the ground objects in the target area have changed. It plays an important role in remote sensing image interpretation.

[0003] like Figure 1 As shown, the existing ground object change detection scheme based on remote sensing images is to use the original remote sensing image group of the target area obtained last time and the panoramic image of the target area obtained by the last stitching as the reference image, and then preprocess the data of the original remote sensing image group taken this time. The key steps of data preprocessing include but are not limited to registration, adjustment elimination, and geometric correction of each remote sensing image, and then stitch the preprocessed remote sensing image group to obtain a panoramic image of the target area taken this time, and then use the change detection and recognition algorithm to compare and analyze the changed area through the panoramic image of the target area taken this time and the panoramic image of the target area taken last time, and output the vector results of the ground object changes in the entire target area.

[0004] In the process of stitching the original remote sensing image group into a panoramic large picture of the target area, the existing technical solution not only has many key steps in the preprocessing process, but also, in order to be able to use the remote sensing images in the original remote sensing image group to stitch a panoramic large picture of the target area, a high degree of overlap is required for each remote sensing image in the remote sensing image group, which in turn leads to a sharp increase in the amount of data, further increasing the amount of data calculation, and the larger the area of ​​the target area, the more obvious the increase in the amount of calculation, which has obvious limitations for application scenarios with high timeliness requirements or where computer hardware resources are difficult to meet. Summary of the invention

[0005] In order to solve the technical problems that the existing technical solutions have large data calculation volume and are limited to application scenarios with high timeliness or low computer hardware resources, the embodiments of the present invention provide a method, device and computer-readable storage medium for detecting ground object changes based on drone remote sensing images.

[0006] The specific technical solutions are as follows:

[0007] In a first aspect, an embodiment of the present invention provides a method for detecting ground object changes based on UAV remote sensing images, comprising the following steps:

[0008] Obtain fixed aerial photography parameters of the target area, and take a group of remote sensing images of the target area through a drone according to the corresponding aerial photography parameters to obtain a current remote sensing image group of the target area; wherein the aerial photography parameters include fixed aerial photography route information and information of fixed shooting points on the aerial photography route;

[0009] According to the GPS information of each remote sensing image in the current remote sensing image group and the remote sensing image group obtained last time, two remote sensing images with the same shooting point in the two remote sensing image groups are matched, and the two remote sensing images with the same shooting point are regarded as a remote sensing image pair, thereby obtaining a remote sensing image pair set of the current remote sensing image group and the remote sensing image group obtained last time;

[0010] Performing registration and geometric correction on each remote sensing image pair in the remote sensing image pair set so that two remote sensing images in each remote sensing image pair are aligned at the pixel level, thereby obtaining a remote sensing image pair set to be used;

[0011] Using a change recognition algorithm, change detection is performed on two remote sensing images in each remote sensing image pair in the remote sensing image pair set to be used, and the ground object change detection result of each shooting point is recorded and saved; wherein, if the two remote sensing images in the remote sensing image pair are different, the change detection result is that a change has occurred, and a vector file with a marked change range is generated; if the two remote sensing images in the remote sensing image pair are the same, the change detection result is that there is no change;

[0012] All the vector files of the shooting points whose change detection results are changes are merged to generate a vector file of the changes of the objects in the entire target area.

[0013] Optionally, the aerial photography parameters also include information on the attitude and orientation of the shooting device;

[0014] When registering each remote sensing image pair in the remote sensing image pair set, the adopted matching algorithm does not consider the influence of rotation and scaling of the remote sensing image on the registration.

[0015] Optionally, when registering each remote sensing image pair in the remote sensing image pair set, a matching algorithm adopted is a Moravec matching algorithm.

[0016] Optionally, before obtaining fixed aerial photography parameters of the target area and photographing a group of remote sensing images of the target area by a drone according to the corresponding aerial photography parameters to obtain a current remote sensing image group of the target area, it also includes setting fixed aerial photography parameters for each target area; wherein the setting of the aerial photography parameters includes planning and setting a fixed aerial photography route for each target area, planning and setting fixed shooting points on the aerial photography route, and setting the attitude and orientation of the shooting equipment.

[0017] Optionally, when planning and setting a fixed aerial photography route for each target area and planning and setting fixed shooting points on the aerial photography route, the following overlap requirements need to be met:

[0018] The degree of overlap between each remote sensing image taken at each shooting point and the remote sensing images taken at the front, rear, left and right shooting points adjacent to the shooting point is greater than zero and less than or equal to five percent; wherein the front shooting point adjacent to the shooting point is the shooting point corresponding to the flight direction of the UAV.

[0019] In a second aspect, an embodiment of the present invention further provides an electronic device, including a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory communicate with each other via the communication bus;

[0020] Memory, used to store computer programs;

[0021] The processor is used to implement the steps of a method for detecting ground object changes based on unmanned aerial vehicle remote sensing images as described in the first aspect when executing the program stored in the memory.

[0022] In a third aspect, an embodiment of the present invention further provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of a method for detecting changes in objects based on drone remote sensing images as described in the first aspect above are implemented.

[0023] The method for detecting changes in land objects provided by the embodiment of the present invention adopts the same aerial photography parameters each time the target area is photographed, and is combined with the change comparison of a single remote sensing image of the same shooting point in the remote sensing image group taken twice before and after the target area, so as to realize the detection of land object changes in the entire target area. In the whole process of each land object change detection, there is no step of splicing the local remote sensing images of the target area to obtain a large panoramic picture of the target area. In this way, not only only registration and geometric correction need to be performed when preprocessing each remote sensing image in the remote sensing image group obtained by the two previous shots, but more importantly, the overlap requirements between each remote sensing image in the remote sensing image group are also greatly reduced, and the number of remote sensing images in the remote sensing image group is also greatly reduced, thereby effectively improving the data processing speed of each land object change detection in the remote sensing image preprocessing and change recognition detection links, with low requirements on computer hardware resources and a wide range of application scenarios.

[0024] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the written description, claims, and drawings.

[0025] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0027] Figure 1 A flowchart of an existing ground object change detection solution based on remote sensing images is provided;

[0028] Figure 2 A schematic diagram of a flow chart of a method for detecting ground object changes based on UAV remote sensing images provided by an embodiment of the present invention;

[0029] Figure 3 A schematic diagram of the structure of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0030] The preferred embodiments of the present invention are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0031] like Figure 2 As shown, the embodiment of the present application provides a method for detecting ground object changes based on drone remote sensing images, the method comprising the following steps:

[0032] Step S100, obtain fixed aerial photography parameters of the target area, and use a drone to shoot a group of remote sensing images of the target area according to the aerial photography parameters to obtain the current remote sensing image group of the target area; wherein, the aerial photography parameters of each aerial photography of the same target area are fixed, and the aerial photography parameters include aerial photography route information and shooting point information on the aerial photography route. In other words, the aerial photography route of each drone aerial photography of the target area a is the same, the shooting points on the aerial photography route are the same (i.e., the exposure points on the route are the same), and the shooting equipment posture is the same (such as the north direction is directly above the remote sensing image sensor), ensuring that the number of remote sensing images obtained for each shooting of the same target area is the same, the geographical location of each remote sensing image corresponds one-to-one, and the shooting angle is the same, so as to facilitate the comparison of changes in a single remote sensing image of the current remote sensing image group of the subsequent target area and the remote sensing image group obtained last time.

[0033] Step S200: According to the GPS information of each remote sensing image in the current remote sensing image group and the remote sensing image group obtained in the last shooting, two remote sensing images with the same shooting points in the two remote sensing image groups are matched, and the two remote sensing images with the same shooting points are regarded as a remote sensing image pair to obtain a remote sensing image pair set of the current remote sensing image group and the remote sensing image group obtained in the last shooting.

[0034] Step S300: performing registration and geometric correction on each remote sensing image pair in the remote sensing image pair set so that the two remote sensing images in each remote sensing image pair are aligned at the pixel level, thereby obtaining a remote sensing image pair set to be used.

[0035] In this step, since the remote sensing image group obtained for the same target area in step S100 is obtained according to a fixed aerial photography route, shooting point and shooting angle, and there is no need to subsequently stitch the remote sensing images in each remote sensing image group into a large panoramic image of the target area, this embodiment does not need to pre-adjust the current remote sensing image group with the last remote sensing image group when pre-processing the remote sensing images.

[0036] In some embodiments, the aerial photography parameters also include the posture and orientation information of the shooting device. Since the shooting device always maintains the same posture during each shooting, there is no need to consider the matching difficulties caused by the rotation and scaling of the remote sensing images when aligning each pair of remote sensing images. Therefore, the embodiment of the present application adopts a simpler and more efficient Moravec matching algorithm when performing alignment. However, the prior art cannot guarantee that the posture and orientation of the shooting device are the same when shooting twice. It is necessary to consider the impact of rotation and scaling on the alignment, so a more complex method has to be used. The more commonly used method is a matching algorithm such as SITF.

[0037] Step S400, using the change recognition algorithm to perform change detection on the two remote sensing images in each remote sensing image pair in the standby remote sensing image pair set, record and save the change detection result of each shooting point (GPS information), if the two remote sensing images in the remote sensing image pair are different, that is, the ground object corresponding to the remote sensing image shot at the shooting point has changed, the change detection result is a change, and a vector file marking the change range is generated; if the two remote sensing images in the remote sensing image pair are the same, the change detection result is no change, and the next remote sensing image pair in the standby remote sensing image pair set is processed until all remote sensing image pairs in the standby remote sensing image pair set have completed change recognition.

[0038] Step S500: merge all vector files of the changed shooting points with the change detection results to generate a vector file of the change of the entire target area, and output the vector file of the change of the entire target area as the result of the current ground object change detection of the target area. Since the merging of vector files reduces the amount of data compared to the splicing of remote sensing images, the speed of obtaining the result of ground object change detection is very fast.

[0039] The present application provides a method for detecting changes in objects based on UAV remote sensing images. The same aerial photography parameters are used each time the target area is photographed. Under this premise, combined with the change comparison of a single remote sensing image of the same shooting point in the remote sensing image group taken twice before and after the target area, the detection of changes in objects in the entire target area can be achieved. In the entire process of each detection of object changes, there is no step of splicing the local remote sensing images of the target area to obtain a panoramic view of the target area. In this way, not only does it only require registration and geometric correction when preprocessing each remote sensing image in the remote sensing image group obtained by the two previous shots, but more importantly, the overlap requirements between each remote sensing image in the remote sensing image group are also greatly reduced, and the number of remote sensing images in the remote sensing image group is also greatly reduced, thereby effectively improving the data processing efficiency of each object change detection in the remote sensing image preprocessing and change recognition detection links, thereby achieving rapid acquisition of the results of each object change detection in the target area.

[0040] Since the embodiment of the present application no longer stitches a large panoramic image of the target area, compared with the high requirements of the prior art on the overlap between each remote sensing image in the remote sensing image group (generally not less than 30%), the embodiment of the present application does not have a high overlap requirement for the overlap of each remote sensing image in the remote sensing image group taken each time. In order to ensure that there are no blind spots in the monitoring of the target area and to maximize the data processing efficiency, the embodiment of the present application, before executing step S100, also includes setting fixed aerial photography parameters for each target area; wherein the setting of aerial photography parameters includes planning and setting a fixed aerial photography route for each target area, planning and setting fixed shooting points on the aerial photography route, and setting the attitude and orientation of the shooting equipment.

[0041] When planning and setting a fixed aerial photography route for each target area and a fixed shooting point on the aerial photography route, the following overlap requirements need to be met: the overlap of each remote sensing image taken at each shooting point with the remote sensing images taken at the front, back, left and right shooting points adjacent to the shooting point (the orientation is based on the flight direction of the UAV, the front is the flight direction of the UAV, the opposite is the back, and the left and right are the sides of the UAV) is greater than zero and less than or equal to 5%, which is much lower than the overlap requirements for aerial photography of other remote sensing images. In this way, the number of remote sensing images in each remote sensing image group will be greatly reduced, and the data processing volume of steps S200 to S400 will also be greatly reduced, especially step S400, thereby effectively improving the data processing efficiency of the entire ground object change detection.

[0042] and Figure 2 Corresponding to the embodiment shown, the embodiment of the present invention further provides an electronic device, see Figure 3, including a processor 301, a communication interface 302, a memory 303 and a communication bus 304, wherein the processor 301, the communication interface 302, and the memory 303 communicate with each other through the communication bus 304;

[0043] Memory 303, used for storing computer programs;

[0044] The processor 301 is used to implement any one of the ground feature change detection method steps in the above-mentioned embodiments when executing the program stored in the memory.

[0045] It can be seen from this embodiment that, by applying the electronic device provided by the embodiment of the present invention, the same aerial photography parameters are used each time the target area is photographed, and combined with the change comparison of a single remote sensing image of the same shooting point in the remote sensing image group taken twice before and after the target area, the whole-area ground object change detection of the target area can be realized. In the whole process of each ground object change detection, there is no link of splicing the local remote sensing images of the target area to obtain a large panoramic picture of the target area. In this way, not only only registration and geometric correction need to be performed when preprocessing each remote sensing image group obtained by the two previous and subsequent shots, but more importantly, the overlap requirement between each remote sensing image in the remote sensing image group is also greatly reduced, and the number of remote sensing images in the remote sensing image group is also greatly reduced, thereby effectively improving the data processing speed of each ground object change detection in the remote sensing image preprocessing and change recognition detection links, with low requirements on computer hardware resources and a wide range of usage scenarios.

[0046] and Figure 1 Corresponding to the embodiment shown, an embodiment of the present invention further provides a computer-readable storage medium, in which a computer program is stored. When the computer program is executed by a processor, any step of the method for detecting changes in land objects in the above-mentioned embodiment is implemented.

[0047] It can be seen from this embodiment that by applying the storage medium provided by the embodiment of the present invention, the same aerial photography parameters are used each time the target area is photographed, and combined with the change comparison of a single remote sensing image of the same shooting point in the remote sensing image group taken twice before and after the target area, the whole-area ground object change detection of the target area can be realized. In the whole process of each ground object change detection, there is no link of splicing the local remote sensing images of the target area to obtain a large panoramic picture of the target area. In this way, not only only registration and geometric correction need to be performed when preprocessing each remote sensing image in the remote sensing image group obtained by the two previous shots, but more importantly, the overlap requirement between each remote sensing image in the remote sensing image group is also greatly reduced, and the number of remote sensing images in the remote sensing image group is also greatly reduced, thereby effectively improving the data processing speed of each ground object change detection in the remote sensing image preprocessing and change recognition detection links, with low requirements on computer hardware resources and a wide range of usage scenarios.

[0048] 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.

[0049] 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. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0050] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A method for detecting ground object changes based on UAV remote sensing images, characterized in that: include: Obtain fixed aerial photography parameters of the target area, and take a group of remote sensing images of the target area through a drone according to the corresponding aerial photography parameters to obtain a current remote sensing image group of the target area; wherein the aerial photography parameters include fixed aerial photography route information and information of fixed shooting points on the aerial photography route; According to the GPS information of each remote sensing image in the current remote sensing image group and the remote sensing image group obtained last time, two remote sensing images with the same shooting point in the two remote sensing image groups are matched, and the two remote sensing images with the same shooting point are regarded as a remote sensing image pair, thereby obtaining a remote sensing image pair set of the current remote sensing image group and the remote sensing image group obtained last time; Performing registration and geometric correction on each remote sensing image pair in the remote sensing image pair set so that two remote sensing images in each remote sensing image pair are aligned at the pixel level, thereby obtaining a remote sensing image pair set to be used; Using a change recognition algorithm, change detection is performed on two remote sensing images in each remote sensing image pair in the remote sensing image pair set to be used, and the ground object change detection result of each shooting point is recorded and saved; wherein, if the two remote sensing images in the remote sensing image pair are different, the change detection result is that a change has occurred, and a vector file with a marked change range is generated; if the two remote sensing images in the remote sensing image pair are the same, the change detection result is that there is no change; All the vector files of the shooting points whose change detection results are changes are merged to generate a vector file of the changes of the objects in the entire target area.

2. The method according to claim 1, characterized in that: The aerial photography parameters also include information on the attitude and orientation of the shooting device; When registering each remote sensing image pair in the remote sensing image pair set, the adopted matching algorithm does not consider the influence of rotation and scaling of the remote sensing image on the registration.

3. The method according to claim 2, characterized in that When registering each remote sensing image pair in the remote sensing image pair set, the matching algorithm adopted is the Moravec matching algorithm.

4. The method according to claim 2, characterized in that: Before obtaining fixed aerial photography parameters of the target area and photographing a group of remote sensing images of the target area by a drone according to the corresponding aerial photography parameters to obtain a current remote sensing image group of the target area, it also includes setting fixed aerial photography parameters for each target area; wherein the setting of the aerial photography parameters includes planning and setting a fixed aerial photography route for each target area, planning and setting fixed shooting points on the aerial photography route, and setting the attitude and orientation of the shooting equipment.

5. The method according to claim 4, characterized in that When planning and setting a fixed aerial photography route for each target area and planning and setting fixed shooting points on the aerial photography route, the following overlap requirements need to be met: The degree of overlap between each remote sensing image taken at each shooting point and the remote sensing images taken at the front, rear, left and right shooting points adjacent to the shooting point is greater than zero and less than or equal to five percent; wherein the front shooting point adjacent to the shooting point is the shooting point corresponding to the flight direction of the UAV.

6. An electronic device, characterized in that: It includes a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory communicate with each other through the communication bus; The memory is used to store computer programs; The processor is used to implement the method steps described in any one of claims 1-5 when executing the program stored in the memory.

7. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method steps described in any one of claims 1 to 5 are implemented.

Images