Ground pickup path planning method and device based on unmanned aerial vehicle aerial photography system, computer equipment and storage medium

Abstract

The invention provides a ground pickup path planning method and device based on an unmanned aerial vehicle aerial photography system, computer equipment and a storage medium. The method comprises the following steps: acquiring a global boundary image of an aerial photography unmanned aerial vehicle; acquiring a global boundary classification parameter according to the global boundary image; performing boundary type likelihood loss processing on the global boundary classification parameter and a preset boundary classification parameter to obtain a boundary type loss amount; and sending a path updating loss signal to a ground pickup path planner according to the boundary class loss amount. After global boundary classification parameters are collected, the boundary distribution condition of a current to-be-planned area is determined, then the global boundary classification parameters are compared with standard boundary classification parameters, the boundary distribution difference degree of the to-be-planned area is determined conveniently, and finally, the to-be-planned area is planned according to the boundary distribution difference degree. And adjusting the moving path of the ground pickup host to enable the ground pickup path planned by the ground pickup host to fully cover the to-be-planned area.

Description

Technical Field

[0001] This disclosure relates to the field of image data technology, and in particular to a ground-based object pickup path planning method, apparatus, computer equipment, and storage medium based on an unmanned aerial vehicle (UAV) aerial photography system. Background Technology

[0002] With the rapid development of intelligent robots, lawn weeding and pond cleaning, as crucial aspects of debris removal, directly impact the utilization rate of lawns and ponds. Traditional lawn weeding and pond cleaning primarily rely on manual or semi-manual methods, which have several drawbacks. First, manual cleaning is labor-intensive and inefficient, making it difficult to meet the needs of large-scale lawn / pond maintenance. Second, while traditional robots improve efficiency, their cleaning paths are limited, often leaving some hard-to-reach areas in the lawn / pond unattended, requiring separate manual cleaning and resulting in low overall efficiency. Summary of the Invention

[0003] The purpose of this disclosure is to overcome the shortcomings of the prior art and provide a ground-based object pickup path planning method, apparatus, computer equipment, and storage medium based on a drone aerial photography system to effectively improve cleaning efficiency.

[0004] The purpose of this disclosure is achieved through the following technical solution: A ground-based object pickup path planning method based on a drone aerial photography system includes: performing a global boundary analysis of the ground using a drone aerial photography system, wherein the drone aerial photography system includes: an aerial photography drone and a ground-based object pickup host; the aerial photography drone is used to take global photos of the area to be planned to obtain a global boundary image of the area to be planned; the ground-based object pickup host is located within the area to be planned, and the ground-based object pickup host is used to plan a path within the area to be planned and to pick up objects along the planned path; The ground object pickup path planning method includes: Acquire the global boundary image of the aerial drone, which includes planar or three-dimensional images of the area to be planned; Obtain global boundary classification parameters based on the global boundary image; The global boundary classification parameters and the preset boundary classification parameters are subjected to boundary type likelihood loss processing to obtain the boundary class loss. Based on the boundary loss, a path update failure signal is sent to the ground pickup path planner to adjust the movement path of the ground pickup host.

[0005] In one embodiment, obtaining global boundary classification parameters based on the global boundary image includes: obtaining global boundary non-circularity based on the global boundary image.

[0006] In one embodiment, the global boundary classification parameters and preset boundary classification parameters are subjected to boundary type likelihood loss processing to obtain the boundary type loss amount, including: calculating the boundary circle likelihood loss of the global boundary non-circularity and the preset non-circularity to obtain the boundary non-circularity loss.

[0007] In one embodiment, sending a path update enable signal to the ground pickup path planner based on the boundary loss amount to adjust the movement path of the ground pickup host includes: detecting whether the boundary non-roundness loss is greater than or equal to a preset non-roundness loss; when the boundary non-roundness loss is greater than or equal to the preset non-roundness loss, sending a path update enable signal to the ground pickup path planner.

[0008] In one embodiment, obtaining global boundary classification parameters based on the global boundary image includes: obtaining global machine-edge distance based on the global boundary image.

[0009] In one embodiment, the global boundary classification parameters and preset boundary classification parameters are subjected to boundary type likelihood loss processing to obtain the boundary type loss amount, including: calculating the edge distance likelihood loss between the global machine-edge distance and the preset distance to obtain the edge distance loss.

[0010] In one embodiment, sending a path update failure signal to the ground pickup path planner based on the boundary loss amount to adjust the movement path of the ground pickup host includes: detecting whether the edge distance loss is greater than or equal to a preset distance loss; when the edge distance loss is less than the preset distance loss, sending a path update failure signal to the ground pickup path planner.

[0011] A ground-based object-picking path planning device is disclosed. The device employs the ground-based object-picking path planning method based on a UAV aerial photography system as described in any of the above embodiments. The device includes a global acquisition module, a boundary classification processing module, and a path adjustment module. The global acquisition module acquires global boundary images of the UAV. The boundary classification processing module obtains global boundary classification parameters based on the global boundary images. It then performs boundary type likelihood loss processing on the global boundary classification parameters and preset boundary classification parameters to obtain a boundary type loss. The path adjustment module sends a path update failure signal to the ground-based object-picking path planner based on the boundary type loss to adjust the movement path of the ground-based object-picking host.

[0012] A computer device includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program performing the following steps: Acquire the global boundary image of the aerial drone, which includes planar or three-dimensional images of the area to be planned; Obtain global boundary classification parameters based on the global boundary image; The global boundary classification parameters and the preset boundary classification parameters are subjected to boundary type likelihood loss processing to obtain the boundary class loss. Based on the boundary loss, a path update failure signal is sent to the ground pickup path planner to adjust the movement path of the ground pickup host.

[0013] A computer-readable storage medium having a computer program stored thereon, the computer program performing the following steps when executed by a processor: Acquire the global boundary image of the aerial drone, which includes planar or three-dimensional images of the area to be planned; Obtain global boundary classification parameters based on the global boundary image; The global boundary classification parameters and the preset boundary classification parameters are subjected to boundary type likelihood loss processing to obtain the boundary class loss. Based on the boundary loss, a path update failure signal is sent to the ground pickup path planner to adjust the movement path of the ground pickup host.

[0014] Compared with the prior art, this disclosure has at least the following advantages: After collecting the boundary classification parameters for the entire area, the current boundary distribution of the area to be planned by the aerial drone is determined. Then, the boundary classification parameters for the entire area are compared with the standard boundary classification parameters to determine the degree of difference in the boundary distribution of the area to be planned by the aerial drone. Finally, based on the degree of difference in the boundary distribution, the movement path of the ground pickup host is adjusted so that the ground pickup path planned by the ground pickup host fully covers the area to be planned, avoiding dead spots and effectively improving the cleaning efficiency. Attached Figure Description

[0015] To more clearly illustrate the technical solutions of the embodiments of this disclosure, the accompanying drawings used in the embodiments will be briefly described below. It should be understood that the following drawings only show some embodiments of this disclosure and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0016] Figure 1 This is a flowchart of a ground object pickup path planning method based on a UAV aerial photography system in one embodiment; Figure 2 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0017] To facilitate understanding of this disclosure, a more complete description will be given below with reference to the accompanying drawings, which illustrate preferred embodiments of the present disclosure. However, this disclosure can be implemented in many different forms and is not limited to the embodiments described herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of the disclosure.

[0018] It should be noted that when an element is referred to as being "fixed to" another element, it can be directly attached to the other element or there may be an intervening element. When an element is referred to as being "connected to" another element, it can be directly connected to the other element or there may be an intervening element. The terms "vertical," "horizontal," "left," "right," and similar expressions used herein are for illustrative purposes only and do not represent the only possible implementation.

[0019] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this disclosure belongs. The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of this disclosure. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0020] This disclosure relates to a ground-based object-collecting path planning method based on a drone aerial photography system. In one embodiment, the method includes acquiring a global boundary image of the drone, which includes a planar or three-dimensional image of the area to be planned; acquiring global boundary classification parameters based on the global boundary image; performing boundary type likelihood loss processing on the global boundary classification parameters and preset boundary classification parameters to obtain a boundary type loss; and sending a path update failure signal to the ground-based object-collecting path planner based on the boundary type loss to adjust the movement path of the ground-based object-collecting host. After acquiring the global boundary classification parameters, the current boundary distribution of the area to be planned by the drone is determined. Then, the global boundary classification parameters are compared with standard boundary classification parameters to determine the degree of difference in the boundary distribution of the area to be planned by the drone. Finally, based on the degree of difference in boundary distribution, the movement path of the ground-based object-collecting host is adjusted so that the ground-based object-collecting path planned by the ground-based object-collecting host fully covers the area to be planned, avoiding dead zones and effectively improving cleaning efficiency.

[0021] See also Figure 1This is a flowchart of a ground-based object-collecting path planning method based on a drone aerial photography system, according to an embodiment of this disclosure. The ground-based object-collecting path planning method includes some or all of the following steps: a drone aerial photography system is used to perform global boundary analysis on the ground. The drone aerial photography system includes: an aerial photography drone and a ground-collecting host; the aerial photography drone is used to take global photographs of the area to be planned to obtain a global boundary image of the area to be planned; the ground-collecting host is located within the area to be planned, and the ground-collecting host is used to plan a path within the area to be planned and to collect objects along the planned path.

[0022] The ground object pickup path planning method specifically includes the following steps: S100: Acquires the global boundary image of the aerial drone, which includes planar or three-dimensional images of the area to be planned.

[0023] In this embodiment, the global boundary image is an aerial image of the area to be planned taken by the drone. The drone acquires the global boundary image and transmits it back to the ground-based pickup host. For example, the drone takes photos and transmits them in real time, or it completes capturing all the aerial images of the area to be planned and returns to the ground-based pickup host. Image transmission occurs through the I / O port between the two devices. The drone can also be charged on the ground-based pickup host, which acts as a power source to charge the drone. Acquiring the global boundary image facilitates determining the boundary distribution of the area to be planned, thus facilitating the determination of the specific location of the ground-based pickup host within that area. The area corresponding to the global boundary image can be a planar area, such as a lawn; or it can be a three-dimensional area, such as a pool.

[0024] S200: Obtain global boundary classification parameters based on the global boundary image.

[0025] In this embodiment, the global boundary classification parameters are obtained based on the graphical analysis of the global boundary image, which is an aerial photograph of the area to be planned by the drone. The drone collects the global boundary image and transmits it back to the ground-based pickup host. For example, the drone takes pictures and transmits them in real time, or it completes the aerial photography of the area to be planned and returns to the ground-based pickup host, transmitting the image through the I / O port between the two. Collecting the global boundary image facilitates the determination of the boundary distribution of the area to be planned, thereby facilitating the determination of the specific location of the ground-based pickup host within the area. The global boundary classification parameters are the boundary classification data of the area to be planned; that is, the global boundary classification parameters represent the boundary type of the area to be planned, and thus correspond to the boundary distribution of the area to be planned. Collecting the global boundary classification parameters facilitates the determination of the regional shape distribution of the area to be planned, thereby facilitating the determination of the location of the ground-based pickup host within the area.

[0026] S300: Perform boundary type likelihood loss processing on the global boundary classification parameters and the preset boundary classification parameters to obtain the boundary class loss.

[0027] In this embodiment, the global boundary classification parameters are obtained based on the graphical analysis of the global boundary image, which is an aerial photograph of the area to be planned by the drone. The drone collects the global boundary image and transmits it back to the ground-based pickup host. For example, the drone takes pictures and transmits them in real time, or it completes the aerial photography of the area to be planned and returns to the ground-based pickup host, transmitting the image through the I / O port between the two. Collecting the global boundary image facilitates the determination of the boundary distribution of the area to be planned, thereby facilitating the determination of the specific location of the ground-based pickup host within the area. The global boundary classification parameters are the boundary classification data of the area to be planned; that is, the global boundary classification parameters represent the boundary type of the area to be planned, and thus correspond to the boundary distribution of the area to be planned. Collecting the global boundary classification parameters facilitates the determination of the regional shape distribution of the area to be planned, thereby facilitating the determination of the location of the ground-based pickup host within the area. The preset boundary classification parameters are the standard boundary classification data of the area to be planned. By processing the boundary type likelihood loss between the global boundary classification parameters and the preset boundary classification parameters, it is easy to determine the degree of difference in the boundary distribution of the area to be planned by the aerial drone.

[0028] S400: Send a path update failure signal to the ground pickup path planner according to the boundary loss amount to adjust the movement path of the ground pickup host.

[0029] In this embodiment, the boundary classification loss is obtained based on the global boundary classification parameters and the preset boundary classification parameters. The global boundary classification parameters are obtained through graphic analysis of the global boundary image, which is an aerial image of the area to be planned by the drone. The drone collects the global boundary image and transmits it back to the ground pickup host. For example, the drone takes pictures and transmits them in real time, or it completes the aerial photography of the area to be planned and returns to the ground pickup host, transmitting the image through the I / O port between the two. By collecting the global boundary image, it is easy to determine the boundary distribution of the area to be planned, thereby facilitating the determination of the specific location of the ground pickup host within the area to be planned. The global boundary classification parameters are the boundary classification data of the area to be planned, that is, the global boundary classification parameters are the boundary type of the area to be planned, and the global boundary classification parameters correspond to the boundary distribution of the area to be planned. By collecting the global boundary classification parameters, it is easy to determine the regional shape distribution of the area to be planned, thereby facilitating the determination of the location of the ground pickup host within the area to be planned. The preset boundary classification parameters are the standard boundary classification data of the area to be planned. By processing the boundary type likelihood loss between the global boundary classification parameters and the preset boundary classification parameters, the degree of difference in the boundary distribution of the area to be planned by the aerial drone can be easily determined. After obtaining the boundary type loss, the boundary distribution of the area to be planned where the ground pickup host is located is determined. At this time, according to the boundary distribution difference reflected by the boundary type loss, the movement path of the ground pickup host in the area to be planned is adjusted accordingly, so that the movement path of the ground pickup host fully covers the area to be planned, so as to avoid omissions or dead zones, thereby improving the pickup efficiency of the ground pickup host.

[0030] In the above embodiments, after collecting the global boundary classification parameters, the current boundary distribution of the planned area of ​​the aerial drone is determined. Then, the global boundary classification parameters are compared with the standard boundary classification parameters to determine the degree of difference in the boundary distribution of the planned area of ​​the aerial drone. Finally, based on the degree of difference in the boundary distribution, the movement path of the ground pickup host is adjusted so that the ground pickup path planned by the ground pickup host fully covers the planned area, avoiding dead ends and effectively improving the cleaning efficiency.

[0031] In one embodiment, obtaining global boundary classification parameters based on the global boundary image includes: obtaining the global boundary non-circularity based on the global boundary image. In this embodiment, the global boundary classification parameters are obtained based on graphic analysis of the global boundary image, which is an aerial image of the area to be planned taken by the aerial drone. The aerial drone collects the global boundary image and transmits it back to the ground pickup host. For example, the aerial drone takes pictures and transmits them in real time, or the aerial drone returns to the ground pickup host after completing the aerial image of the area to be planned, and the image is transmitted through the I / O port between the two. By collecting the global boundary image, it is easy to determine the boundary distribution of the area to be planned, thereby facilitating the determination of the specific location of the ground pickup host within the area to be planned. The global boundary classification parameters are the boundary classification data of the area to be planned, that is, the global boundary classification parameters are the boundary type of the area to be planned, and the global boundary classification parameters correspond to the boundary distribution of the area to be planned. By collecting the global boundary classification parameters, it is easy to determine the shape distribution of the area to be planned, thereby facilitating the determination of the location of the ground-based object pickup device within the planned area. The global boundary classification parameters include global boundary non-circularity, which is an index of the uniformity of the boundary distribution of the area to be planned; that is, the global boundary non-circularity represents the degree of continuous smoothness of the boundary of the area to be planned. By collecting the global boundary non-circularity, it is easy to determine the smoothness of the boundary of the area to be planned.

[0032] Further, the global boundary classification parameters and preset boundary classification parameters are subjected to boundary type likelihood loss processing to obtain the boundary type loss, including: calculating the boundary circle likelihood loss between the global boundary non-circularity and the preset non-circularity to obtain the boundary non-circularity loss. In this embodiment, the global boundary classification parameters are obtained based on the graphic analysis of the global boundary image, which is an aerial image of the area to be planned by the aerial photography drone. The aerial photography drone collects the global boundary image and transmits it back to the ground pickup host. For example, the aerial photography drone takes pictures and transmits them in real time, or the aerial photography drone returns to the ground pickup host after completing the aerial images of the area to be planned, and the image is transmitted through the I / O port between the two. By collecting the global boundary image, it is easy to determine the boundary distribution of the area to be planned, thereby facilitating the determination of the specific location of the ground pickup host within the area to be planned. The global boundary classification parameters are the boundary classification data of the area to be planned, that is, the global boundary classification parameters are the boundary type of the area to be planned, and the global boundary classification parameters correspond to the boundary distribution of the area to be planned. By collecting the global boundary classification parameters, it is easy to determine the regional shape distribution of the area to be planned, thereby facilitating the determination of the location of the ground-based pickup host within the area to be planned. The preset boundary classification parameters are the standard boundary classification data of the area to be planned. By performing boundary type likelihood loss processing on the global boundary classification parameters and the preset boundary classification parameters, it is easy to determine the degree of difference in the boundary distribution of the area to be planned by the aerial drone. The global boundary classification parameters include global boundary non-circularity, which is an index of the uniformity of the regional boundary distribution of the area to be planned, that is, the global boundary non-circularity is the degree of continuous smoothness of the regional boundary of the area to be planned. By collecting the global boundary non-circularity, it is easy to determine the smoothness of the regional boundary of the area to be planned. The preset non-circularity is a reference index for the uniformity of the regional boundary distribution of the area to be planned. By calculating the boundary circle likelihood loss between the global boundary non-circularity and the preset non-circularity, it is easy to determine the difference in the uniformity of the regional boundary distribution of the area to be planned.

[0033] Furthermore, a path update failure signal is sent to the ground pickup path planner based on the boundary classification loss to adjust the movement path of the ground pickup host. This includes: detecting whether the boundary non-circularity loss is greater than or equal to a preset non-circularity loss; and when the boundary non-circularity loss is greater than or equal to the preset non-circularity loss, sending a path update enable signal to the ground pickup path planner. In this embodiment, the boundary classification loss is obtained based on the global boundary classification parameters and the preset boundary classification parameters. The global boundary classification parameters are obtained based on the graphic analysis of the global boundary image, which is an aerial image of the area to be planned by the aerial photography drone. The aerial photography drone collects the global boundary image and transmits it back to the ground pickup host. For example, the aerial photography drone takes pictures and transmits them in real time, or the aerial photography drone returns to the ground pickup host after completing all the aerial images of the area to be planned, and the image is transmitted through the I / O port between the two. By collecting the global boundary image, it is easy to determine the boundary distribution of the area to be planned, thereby facilitating the determination of the specific location of the ground pickup host within the area to be planned. The global boundary classification parameters are the boundary classification data of the area to be planned, that is, the global boundary classification parameters are the boundary type of the area to be planned, and the global boundary classification parameters correspond to the boundary distribution of the area to be planned. By collecting the global boundary classification parameters, it is easy to determine the regional shape distribution of the area to be planned, thereby facilitating the determination of the position of the ground-based pickup host within the area to be planned. The preset boundary classification parameters are the standard boundary classification data of the area to be planned. By processing the boundary type likelihood loss between the global boundary classification parameters and the preset boundary classification parameters, it is easy to determine the degree of difference in the boundary distribution of the area to be planned by the aerial drone. After obtaining the boundary type loss, the boundary distribution of the area to be planned where the ground-based pickup host is located is determined. At this time, based on the boundary distribution difference reflected by the boundary type loss, the movement path of the ground-based pickup host within the area to be planned is adjusted accordingly, so that the movement path of the ground-based pickup host fully covers the area to be planned, avoiding omissions or blind spots, thereby improving the pickup efficiency of the ground-based pickup host. The global boundary classification parameters include global boundary non-circularity, which is an index of the uniformity of the regional boundary distribution of the area to be planned. Specifically, global boundary non-circularity represents the degree of continuous smoothness of the regional boundary of the area to be planned. Collecting global boundary non-circularity facilitates the determination of the smoothness of the regional boundary of the area to be planned. The preset non-circularity serves as a reference index for the uniformity of the regional boundary distribution of the area to be planned. Calculating the boundary likelihood loss between global boundary non-circularity and preset non-circularity facilitates the determination of the differences in the uniformity of the regional boundary distribution of the area to be planned.If the boundary non-circularity loss is greater than or equal to the preset non-circularity loss, it indicates that the uniformity of the boundary distribution of the area to be planned is poor, that is, the smoothness of the boundary of the area to be planned is low. At this time, a path update enable signal is sent to the ground pickup path planner to adjust and update the movement path of the ground pickup host, so that the movement path of the ground pickup host is more adapted to the current shape of the area to be planned. Specifically, the movement path of the ground pickup host is changed from the original loop path to a reciprocating path. The loop path consists of multiple paths parallel to the boundary of the area to be planned, and the reciprocating path consists of multiple Z-shaped paths.

[0034] In another embodiment, when the boundary non-circularity loss is less than a preset non-circularity loss, a path update disabling signal is sent to the ground pickup path planner. At this time, the regional boundary distribution uniformity of the area to be planned is relatively high. By sending a path update disabling signal to the ground pickup path planner, the movement path of the ground pickup host is maintained as a loop path.

[0035] In one embodiment, obtaining global boundary classification parameters based on the global boundary image includes: obtaining global edge distance based on the global boundary image. In this embodiment, the global boundary classification parameters are obtained based on graphic analysis of the global boundary image, which is an aerial image of the area to be planned by the aerial photography drone. The aerial photography drone collects the global boundary image and transmits it back to the ground pickup host. For example, the aerial photography drone takes pictures and transmits them in real time, or the aerial photography drone returns to the ground pickup host after completing the aerial images of the area to be planned, and the image is transmitted through the I / O port between the two. By collecting the global boundary image, it is easy to determine the boundary distribution of the area to be planned, thereby facilitating the determination of the specific location of the ground pickup host within the area to be planned. The global boundary classification parameters are the boundary classification data of the area to be planned, that is, the global boundary classification parameters are the boundary type of the area to be planned, and the global boundary classification parameters correspond to the boundary distribution of the area to be planned. By collecting the global boundary classification parameters, it is easy to determine the shape distribution of the area to be planned, thereby facilitating the determination of the location of the ground-based pickup unit within the area to be planned. The global boundary classification parameters include the global machine-edge distance, which is an index representing the distance between the boundary of the area to be planned and the ground-based pickup unit. In other words, the global machine-edge distance indicates the distance between the boundary of the area to be planned and the ground-based pickup unit. Collecting the global machine-edge distance facilitates the determination of the relative position between the boundary of the area to be planned and the ground-based pickup unit.

[0036] Further, the global boundary classification parameters and preset boundary classification parameters are subjected to boundary type likelihood loss processing to obtain the boundary loss amount, including: calculating the edge distance likelihood loss between the global edge distance and the preset distance to obtain the edge distance loss. In this embodiment, the global boundary classification parameters are obtained based on the graphic analysis of the global boundary image, which is the aerial image of the area to be planned by the aerial photography drone. The aerial photography drone collects the global boundary image and transmits it back to the ground pickup host. For example, the aerial photography drone takes pictures and transmits them in real time, or the aerial photography drone returns to the ground pickup host after completing the aerial image of the area to be planned, and the image is transmitted through the I / O port between the two. By collecting the global boundary image, it is easy to determine the boundary distribution of the area to be planned, thereby making it easier to determine the specific location of the ground pickup host in the area to be planned. The global boundary classification parameters are the boundary classification data of the area to be planned, that is, the global boundary classification parameters are the boundary type of the area to be planned, that is, the global boundary classification parameters correspond to the boundary distribution of the area to be planned. By collecting the global boundary classification parameters, the shape distribution of the area to be planned can be easily determined, thereby facilitating the determination of the location of the ground-based pickup host within the planned area. The preset boundary classification parameters are the standard boundary classification data for the area to be planned. By processing the boundary type likelihood loss between the global boundary classification parameters and the preset boundary classification parameters, the degree of difference in the boundary distribution of the area to be planned by the aerial drone can be easily determined. The global boundary classification parameters include the global drone-edge distance, which is an index representing the distance between the boundary of the area to be planned and the ground-based pickup host. In other words, the global drone-edge distance indicates the distance between the boundary of the area to be planned and the ground-based pickup host. By collecting the global drone-edge distance, the relative position of the boundary of the area to be planned and the ground-based pickup host can be easily determined. The preset distance is a reference index for the distance between the boundary of the area to be planned and the ground pickup host. By calculating the edge likelihood loss between the edge distance of the whole area and the preset distance, it is easy to determine the difference between the boundary of the area to be planned and the distance between the ground pickup host, thereby making it easier to determine the degree of overlap between the ground pickup host and the center of the area to be planned.

[0037] Furthermore, a path update failure signal is sent to the ground-based pickup path planner based on the boundary loss amount to adjust the movement path of the ground-based pickup host. This includes: detecting whether the boundary loss is greater than or equal to a preset boundary loss; and when the boundary loss is less than the preset boundary loss, sending a path update failure signal to the ground-based pickup path planner. In this embodiment, the boundary loss amount is obtained based on the global boundary classification parameters and the preset boundary classification parameters. The global boundary classification parameters are obtained based on the graphical analysis of the global boundary image, which is an aerial image of the area to be planned by the aerial photography drone. The aerial photography drone collects the global boundary image and transmits it back to the ground-based pickup host. For example, the aerial photography drone takes pictures and transmits them in real time, or the aerial photography drone returns to the ground-based pickup host after completing all the aerial images of the area to be planned, and the image is transmitted through the I / O port between the two. By collecting the global boundary image, it is easy to determine the boundary distribution of the area to be planned, thereby facilitating the determination of the specific location of the ground-based pickup host within the area to be planned. The global boundary classification parameters are the boundary classification data of the area to be planned, that is, the global boundary classification parameters are the boundary type of the area to be planned, and the global boundary classification parameters correspond to the boundary distribution of the area to be planned. By collecting the global boundary classification parameters, it is easy to determine the regional shape distribution of the area to be planned, thereby facilitating the determination of the position of the ground-based pickup host within the area to be planned. The preset boundary classification parameters are the standard boundary classification data of the area to be planned. By processing the boundary type likelihood loss between the global boundary classification parameters and the preset boundary classification parameters, it is easy to determine the degree of difference in the boundary distribution of the area to be planned by the aerial drone. After obtaining the boundary type loss, the boundary distribution of the area to be planned where the ground-based pickup host is located is determined. At this time, based on the boundary distribution difference reflected by the boundary type loss, the movement path of the ground-based pickup host within the area to be planned is adjusted accordingly, so that the movement path of the ground-based pickup host fully covers the area to be planned, avoiding omissions or blind spots, thereby improving the pickup efficiency of the ground-based pickup host. The global boundary classification parameters include the global machine-edge distance, which is the distance index between the boundary of the area to be planned and the ground pickup host. In other words, the global machine-edge distance is the distance between the boundary of the area to be planned and the ground pickup host. By collecting the global machine-edge distance, it is easy to determine the relative position between the boundary of the area to be planned and the ground pickup host.The preset distance serves as a reference index for the distance between the boundary of the area to be planned and the ground-based pickup host. By calculating the edge likelihood loss between the global edge distance and the preset distance, the difference in distance between the boundary of the area to be planned and the ground-based pickup host can be easily determined, thereby facilitating the determination of the degree of overlap between the ground-based pickup host and the regional center of the area to be planned. If the edge distance loss is less than the preset distance loss, it indicates that the distance between the boundary of the area to be planned and the ground-based pickup host is small, meaning that the ground-based pickup host is close to the boundary of the area to be planned, or that the ground-based pickup host is significantly deviated from the regional center of the area to be planned. In this case, a path update disabling signal is sent to the ground-based pickup path planner to maintain the ground-based pickup host's movement path as a loop path. This facilitates the initial loop path of the ground-based pickup host moving along the boundary of the area to be planned, forming a movement template for subsequent loop paths. Furthermore, the boundary of the area to be planned can be sampled a second time to improve the accuracy of the boundary positioning.

[0038] In another embodiment, if the machine edge distance loss is greater than or equal to a preset distance loss, it indicates that the distance between the boundary of the area to be planned and the ground pickup host is large, that is, it indicates that the ground pickup host is far away from the boundary of the area to be planned, and that the ground pickup host and the center of the area to be planned are basically coincident. At this time, a path update enable signal is sent to the ground pickup path planner to adjust and update the movement path of the ground pickup host, so that the movement path of the ground pickup host is more adapted to the current shape of the area to be planned. Specifically, the movement path of the ground pickup host is changed from the original loop path to a reciprocating path, wherein the loop path consists of multiple paths parallel to the boundary of the area to be planned, and the reciprocating path consists of multiple Z-shaped paths.

[0039] During the actual full-area photography of the planned area by the aerial drone, the planned area serves as the working area of ​​the ground pickup host. The full-area boundary image is the image of the ground pickup host within the planned area. In addition to the ground pickup host, there are often obstacles such as facilities or trees within the planned area. Moving directly according to the planned path can easily be blocked by these obstacles.

[0040] To improve obstacle avoidance performance, a path update failure signal is sent to the ground-based object-picking path planner based on the boundary loss amount to adjust the movement path of the ground-based object-picking host. This process also includes: Perform intra-domain boundary processing on the global boundary image to obtain the area of ​​the intra-domain obstacle unit; Detect whether the area of ​​the obstacle unit within the domain is greater than the preset unit area; When the area of ​​the obstacle unit within the domain is greater than the preset unit area, a boundary obstacle avoidance compensation path signal is sent to the ground object pickup path planner.

[0041] In this embodiment, the global boundary image undergoes intra-domain boundary processing. This involves identifying obstacles in the photographed area of ​​the global boundary image and calculating the size of the obstacle's location to obtain the location and area of ​​the obstacle within the planned area. The preset unit area is the obstacle identification area within the planned area; specifically, the preset unit area is the area of ​​the ground-based pickup host. If the intra-domain obstacle unit area is larger than the preset unit area, it indicates that the marked area within the planned area contains an obstacle, and this obstacle is much larger than the ground-based pickup host. In this case, the ground-based pickup host cannot overcome the obstacle. By sending an obstacle avoidance compensation path signal along the boundary to the ground-based pickup path planner, the host can avoid obstacles on subsequent planned movement paths. Specifically, when the ground-based pickup host moves to the location of the obstacle, it detours along the obstacle's boundary to re-move onto the subsequent path of the planned path.

[0042] Further, after detecting whether the area of ​​the obstacle unit within the domain is greater than the preset unit area, the process also includes: When the area of ​​the obstacle unit within the domain is less than or equal to the preset unit area, the object diameter cutting angle is obtained based on the global boundary image. Detect whether the object diameter cutting angle matches the preset cutting angle; When the object diameter cutting angle does not match the preset cutting angle, a path spacing adjustment signal is sent to the ground object pickup path planner.

[0043] In this embodiment, the area of ​​the obstacle unit within the domain is less than or equal to the area of ​​the preset unit, indicating that the object corresponding to the marked area in the area to be planned is slightly smaller than the ground-based pickup host. This indicates that the marked object in the area to be planned is an object that needs to be collected, such as garbage in the area to be planned. For the pickup of such objects, it is necessary to compare their location with the movement path. The object diameter tangent is the angle between the straight line connecting the object to be picked up in the area to be planned and the ground-based pickup host, and the tangent of the ground-based pickup host on its current movement path. The object diameter tangent is used to reflect the deviation of the object to be picked up from the current movement path of the ground-based pickup host. The preset chamfer is the angle corresponding to the object to be picked up on the moving path. If the object diameter chamfer does not match the preset chamfer, it indicates that the object to be picked up is seriously deviated from the current moving path of the ground picking host, that is, it indicates that the object to be picked up is not on the current moving path of the ground picking host. At this time, by sending a path spacing adjustment signal to the ground picking path planner, the distance between the current moving path of the ground picking host and the adjacent moving path is adjusted so that the moving path of the ground picking host passes through the object to be picked up, making it easier to clean the object to be picked up.

[0044] In another embodiment, the current movement path is maintained when the object diameter tangent matches a preset tangent.

[0045] All the aforementioned preset variables are set in the database for easy retrieval, and different preset variables are placed in different storage units, i.e., in different storage stacks. Moreover, the area of ​​obstacle units and the object diameter tangent within the domain can be acquired by the corresponding processor, for example, by the high-resolution camera of the global acquisition module.

[0046] In one embodiment, this disclosure also relates to a ground-based object-picking path planning device. The device employs the ground-based object-picking path planning method based on a UAV aerial photography system described in any of the above embodiments. The device includes: a global acquisition module, a boundary classification processing module, and a path adjustment module. The global acquisition module acquires a global boundary image of the UAV. The boundary classification processing module obtains global boundary classification parameters based on the global boundary image. It performs boundary type likelihood loss processing on the global boundary classification parameters and preset boundary classification parameters to obtain a boundary type loss. The path adjustment module sends a path update failure signal to the ground-based object-picking path planner based on the boundary type loss to adjust the movement path of the ground-based object-picking host.

[0047] In this embodiment, after the boundary classification processing module collects the global boundary classification parameters, it determines the current boundary distribution of the area to be planned by the aerial drone. Then, it compares the global boundary classification parameters with the standard boundary classification parameters to determine the degree of difference in the boundary distribution of the area to be planned by the aerial drone. Finally, the path adjustment module adjusts the movement path of the ground pickup host according to the above-mentioned degree of boundary distribution difference, so that the ground pickup path planned by the ground pickup host fully covers the area to be planned, avoiding dead ends and effectively improving the cleaning efficiency.

[0048] In one embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 2As shown, the computer device includes a processor, memory, and a network interface connected via a system bus. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The database stores data such as global boundary images, global boundary classification parameters, preset boundary classification parameters, and path update failure signals. The network interface is used for communication with external terminals via a network connection. When executed by the processor, the computer program implements a ground object pickup path planning method.

[0049] Those skilled in the art will understand that Figure 2 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0050] In one embodiment, this application also provides a computer device including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in the above-described method embodiments.

[0051] In one embodiment, this application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps in the above-described method embodiments.

[0052] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the methods described above. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, or optical storage, etc. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc.

[0053] The embodiments described above are merely illustrative of several implementations of this disclosure, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this disclosure, and these all fall within the protection scope of this disclosure. Therefore, the protection scope of this patent should be determined by the appended claims.

Claims

1. A ground-based object pickup path planning method based on an unmanned aerial vehicle (UAV) aerial photography system, characterized in that, include: A drone aerial photography system is used to perform global boundary analysis of the ground. The drone aerial photography system includes: Aerial photography drone, which is used to take pictures of the entire area to be planned in order to obtain the boundary image of the entire area to be planned. A ground-based object-collecting host is located within a planned area. The ground-based object-collecting host is used for path planning within the planned area and for collecting objects along the planned path. The ground object pickup path planning method includes: Acquire the global boundary image of the aerial drone, which includes planar or three-dimensional images of the area to be planned; Obtain global boundary classification parameters based on the global boundary image; The global boundary classification parameters and the preset boundary classification parameters are subjected to boundary type likelihood loss processing to obtain the boundary class loss. Based on the boundary loss, a path update failure signal is sent to the ground pickup path planner to adjust the movement path of the ground pickup host.

2. The ground object pickup path planning method based on an unmanned aerial vehicle (UAV) aerial photography system according to claim 1, characterized in that, Global boundary classification parameters are obtained from the global boundary image, including: The global boundary non-circularity is obtained based on the global boundary image.

3. The ground object pickup path planning method based on an unmanned aerial vehicle (UAV) aerial photography system according to claim 2, characterized in that, The global boundary classification parameters and preset boundary classification parameters are subjected to boundary type likelihood loss processing to obtain the boundary class loss, including: Calculate the boundary circle likelihood loss between the global boundary non-circularity and the preset non-circularity to obtain the boundary non-circularity loss.

4. The ground object pickup path planning method based on an unmanned aerial vehicle (UAV) aerial photography system according to claim 3, characterized in that, Based on the boundary loss amount, a path update failure signal is sent to the ground pickup path planner to adjust the movement path of the ground pickup host, including: Detect whether the boundary non-roundness loss is greater than or equal to the preset non-roundness loss; When the boundary non-circularity loss is greater than or equal to the preset non-circularity loss, a path update enable signal is sent to the ground pickup path planner.

5. The ground object pickup path planning method based on an unmanned aerial vehicle (UAV) aerial photography system according to claim 1, characterized in that, Global boundary classification parameters are obtained from the global boundary image, including: The global machine-edge distance is obtained based on the global boundary image.

6. The ground object pickup path planning method based on an unmanned aerial vehicle (UAV) aerial photography system according to claim 5, characterized in that, The global boundary classification parameters and preset boundary classification parameters are subjected to boundary type likelihood loss processing to obtain the boundary class loss, including: Calculate the margin likelihood loss between the global machine edge distance and the preset distance to obtain the margin loss.

7. The ground object pickup path planning method based on an unmanned aerial vehicle (UAV) aerial photography system according to claim 6, characterized in that, Based on the boundary loss amount, a path update failure signal is sent to the ground pickup path planner to adjust the movement path of the ground pickup host, including: Detect whether the edge distance loss is greater than or equal to the preset distance loss; When the edge distance loss is less than the preset distance loss, a path update failure signal is sent to the ground picking path planner.

8. A ground object pickup path planning device, wherein the ground object pickup path planning device adopts the ground object pickup path planning method based on an unmanned aerial vehicle (UAV) aerial photography system as described in any one of claims 1 to 7, characterized in that, include: A global acquisition module, which is used to acquire global boundary images of the aerial drone; A boundary classification processing module is used to obtain global boundary classification parameters based on the global boundary image; and to perform boundary type likelihood loss processing on the global boundary classification parameters and preset boundary classification parameters to obtain the boundary class loss. The path adjustment module is used to send a path update failure signal to the ground picking path planner according to the boundary loss amount, so as to adjust the movement path of the ground picking host.

9. A computer 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 steps of the method according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 7.

Images