A method for UAV flight path planning based on 3D reality models

Abstract

This invention relates to a method for UAV flight path planning based on a 3D reality model, comprising the following specific steps: 3D reality modeling based on oblique photogrammetry, specifically including data acquisition planning, image preprocessing, aerial triangulation, dense matching, and texture mapping; and flight path planning based on the 3D reality model, specifically including importing the 3D model, setting the task area, setting flight parameters, and obstacle avoidance and safety settings. Through the deep integration of 3D reality modeling and intelligent flight path planning, safety and efficiency are improved. The high-precision 3D model generated using oblique photogrammetry technology can intuitively display three-dimensional information such as terrain undulations and obstacle distribution, enabling the UAV to autonomously avoid obstacles in complex environments and reducing collision risks. Dynamic optimization of the flight path through quantitative parameters improves operational efficiency and data acquisition completeness. Flight parameters can be quickly adjusted to adapt to different task requirements, ensuring operational safety while improving data quality and operational efficiency.

Description

Technical Field

[0001] This invention relates to the field of unmanned aerial vehicle (UAV) technology, specifically to a UAV flight path planning method based on a three-dimensional real-scene model. Background Technology

[0002] With the in-depth application of technologies such as cloud computing, the Internet of Things, big data, and artificial intelligence in the water conservancy industry, dam safety monitoring is also developing towards multi-source information fusion, intelligent model analysis, real-time online evaluation, three-dimensional visualization, and intelligent auxiliary decision-making. Among them, drone inspection based on image recognition analysis can identify and warn of abnormal situations such as water accumulation and drainage in the corridor, cracks and seepage in the dam body, deformation and damage on the dam surface, deformation and stability of upstream and downstream bank slopes, and floating objects on the reservoir surface. Compared with manual inspection, drone inspection greatly improves the efficiency and accuracy of inspection.

[0003] According to Chinese invention application number CN202210991219.7, a method, device, equipment and storage medium for planning unmanned aerial vehicle (UAV) patrol routes were proposed. The invention describes that "by acquiring a three-dimensional real-scene model of the target area and route planning parameters, and planning the route in the three-dimensional real-scene model according to the route planning parameters, the UAV can not only capture the desired image of the target object when patrolling along the planned route, but also effectively avoid surrounding obstacles, thus improving the safety and effectiveness of UAV patrols."

[0004] The current UAV patrol route planning method, device, equipment, and storage medium not only lack a clear and specific route planning scheme, parameter optimization process, and obstacle avoidance method, but also lack quantitative indicators for safety mechanisms, resulting in high risks in the planned UAV routes and affecting the flight safety of UAVs. Therefore, a UAV route planning method based on a three-dimensional real-scene model is proposed to solve the above problems. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a method for UAV flight path planning based on a 3D real-scene model. This method has advantages such as good safety performance and solves the problem that not only are there no clear and specific flight path planning schemes, parameter optimization processes, and obstacle avoidance methods, but the safety mechanisms also lack quantitative indicators, resulting in high risks in the planned UAV flight paths and affecting UAV flight safety.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a method for UAV route planning based on a 3D real-scene model, comprising the following specific steps: S1. Three-dimensional reality modeling based on oblique photogrammetry; S1.1. Data Acquisition Planning; S1.2. Image preprocessing; S1.3. Aerial triangulation of images; S1.4. Dense matching; S1.5. Texture mapping; S2. Route planning based on 3D reality models; S2.1. Import the 3D model; S2.2. Set the task area; S2.3. Set flight parameters; S2.4 Obstacle Avoidance and Safety Settings.

[0007] Furthermore, the data acquisition planning method in step S1.1 is as follows: Based on the modeling accuracy requirements and the specific modeling object, select a suitable aircraft and camera, plan a reasonable flight path, and calculate a suitable aerial photography interval based on the camera focal length, sensor pixel size, flight altitude, and required ground resolution. The calculation formula is: ; In the formula, For camera focal length, For sensor pixel size, For flight altitude, Image width, For the required ground resolution. For aerial photography intervals, This is an empirical coefficient.

[0008] Further, the image preprocessing method in step S1.2 is as follows: Image preprocessing includes distortion correction and uniform illumination and color processing. Distortion correction uses in-camera orientation elements and distortion parameters for correction. The correction formula involves radial distortion and tangential distortion models. For a point on the image ( , ), corrected coordinates ( , The formula for calculating ) is: ; ; ; In the formula, the coordinates of the principal point are ( , ), focal length is The radial distortion parameters are respectively and The tangential distortion parameters are respectively and .

[0009] Furthermore, the aerial triangulation method described in step S1.3 is as follows: The relative relationship between adjacent images is restored through feature extraction and feature matching. Feature extraction employs the SIFT algorithm, including scale-space extremum detection, precise keypoint localization, and determination of the principal direction of the keypoints. The calculation formula is: In the formula, and The images are respectively in and Gradient of direction.

[0010] Furthermore, the dense matching method described in step S1.4 is as follows: a semi-global matching algorithm is used to optimize the matching result by aggregating the matching costs from multiple directions. The formula for calculating the total matching cost is: In the formula, For matching cost, For the total matching cost, For pixels, For parallax, For pixels In the field, This is the penalty function.

[0011] Furthermore, the texture mapping method described in step S1.5 is as follows: A texture is applied to the constructed 3D model mesh, and the mapping of each vertex on the texture image is defined. Coordinates are used to map a 2D image to a 3D model, with values ​​ranging from [value range missing]. .

[0012] Furthermore, the method for importing the 3D model in step S2.1 is as follows: the generated 3D real-world model is imported into the flight route planning software, and the calculation formula for transforming the model coordinate system to the world coordinate system using the coordinate transformation matrix is: In the formula, This is the coordinate transformation matrix. For the model coordinate system, Use the world coordinate system.

[0013] Furthermore, the method for setting the task area in step S2.2 is as follows: The task area is defined on the 3D model, with three selection options: polygon, circle, and rectangle. For polygonal areas, the vertex coordinates are used to define the area. definition, For a circular region, the coordinates of the center are used. and radius Definition: For a rectangular region, the coordinates of its diagonal vertices are used to define the region. and definition.

[0014] Furthermore, the method for setting flight parameters in step S2.3 is as follows: flight parameters include flight altitude. Heading overlap rate Lateral overlap rate and flight speed Heading overlap and lateral overlap rate The calculation method is as follows: ; In the formula, The length of the overlapping area in the heading. The length of the flight path covered by a single image. The width of the lateral overlapping area. The lateral coverage width of a single image.

[0015] Furthermore, the obstacle avoidance and safety setting method described in step S2.4 is as follows: marking obstacles on the 3D model and setting a safety height. and emergency return point safe distance According to the size of the drone ,speed and reaction time The calculation method is as follows: In the formula, For deceleration.

[0016] Compared with the prior art, the technical solution of this application has the following beneficial effects: This UAV flight path planning method based on 3D reality models improves safety and efficiency through the deep integration of 3D reality modeling and intelligent flight path planning. It uses oblique photogrammetry technology to generate high-precision 3D models, which can intuitively display three-dimensional information such as terrain undulations and obstacle distribution, enabling UAVs to autonomously avoid obstacles in complex environments and reducing collision risks. By dynamically optimizing flight paths through quantitative parameters, it improves operational efficiency and data acquisition completeness. Flight parameters can be quickly adjusted to adapt to different mission requirements, ensuring operational safety while improving data quality and operational efficiency. Attached Figure Description

[0017] Figure 1 This is a flowchart of the present invention; Figure 2 This is a flowchart of the three-dimensional reality modeling process of the present invention; Figure 3 This is a flowchart of the route planning process of the present invention; Figure 4 Example diagram of planning a mission route on a 3D reality model. Detailed Implementation

[0018] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0019] Please see Figure 1-4 The method for UAV route planning based on a 3D reality model in this embodiment includes the following specific steps: S1. 3D reality modeling based on oblique photogrammetry: S1.1. Data Acquisition Planning: Based on the modeling accuracy requirements and the specific modeling object, select appropriate aircraft and cameras, plan reasonable flight routes, and calculate suitable aerial photography intervals based on camera focal length, sensor pixel size, flight altitude, and required ground resolution. The calculation formula is as follows: ; In the formula, For camera focal length, For sensor pixel size, For flight altitude, Image width, For the required ground resolution. For aerial photography intervals, This is an empirical coefficient.

[0020] S1.2. Image Preprocessing: Image preprocessing includes distortion correction and homogenization. Distortion correction uses in-camera orientation elements and distortion parameters. The correction formula involves radial and tangential distortion models. For a point on the image ( , ), corrected coordinates ( , The formula for calculating ) is: ; ; In the formula, the coordinates of the principal point are ( , ), focal length is The radial distortion parameters are respectively and The tangential distortion parameters are respectively and .

[0021] S1.3. Aerial Triangulation of Images: This involves recovering the relative relationships between adjacent images through feature extraction and feature matching. Feature extraction employs the SIFT algorithm, including scale-space extremum detection, precise keypoint localization, and determination of the principal orientation of keypoints. The calculation formula is: In the formula, and The images are respectively in and Gradient of direction.

[0022] S1.4. Dense Matching: A semi-global matching algorithm is used to optimize the matching result by aggregating the matching costs from multiple directions. The formula for calculating the total matching cost is as follows: In the formula, For matching cost, For the total matching cost, For pixels, For parallax, For pixels In the field, This is the penalty function.

[0023] S1.5. Texture Mapping: Apply textures to the constructed 3D model mesh by defining the mapping of each vertex onto the texture image. Coordinates are used to map a 2D image to a 3D model, with values ​​ranging from [value range missing]. .

[0024] S2. Route planning based on 3D reality models: S2.1. Importing the 3D Model: Import the generated 3D reality model into the flight path planning software. The formula for transforming the model coordinate system to the world coordinate system using the coordinate transformation matrix is ​​as follows: In the formula, This is the coordinate transformation matrix. For the model coordinate system, Use the world coordinate system.

[0025] S2.2. Setting the Task Area: Define the task area on the 3D model, choosing from three options: polygon, circle, or rectangle. For polygonal areas, the vertex coordinates are used to define the area. definition, For a circular region, the coordinates of the center are used. and radius Definition: For a rectangular region, the coordinates of its diagonal vertices are used to define the region. and definition.

[0026] S2.3. Set flight parameters: Flight parameters include flight altitude Heading overlap rate Lateral overlap rate and flight speed Heading overlap and lateral overlap rate The calculation method is as follows: ; In the formula, The length of the overlapping area in the heading. The length of the flight path covered by a single image. The width of the lateral overlapping area. The lateral coverage width of a single image.

[0027] S2.4. Obstacle Avoidance and Safety Settings: Mark obstacles on the 3D model and set the safety height. and emergency return point safe distance According to the size of the drone ,speed and reaction time The calculation method is as follows: In the formula, For deceleration.

[0028] The working principle of the above embodiments is as follows: This UAV flight path planning method based on 3D reality models improves safety and efficiency through the deep integration of 3D reality modeling and intelligent flight path planning. It uses oblique photogrammetry technology to generate high-precision 3D models, which can intuitively display three-dimensional information such as terrain undulations and obstacle distribution, enabling UAVs to autonomously avoid obstacles in complex environments and reducing collision risks. By dynamically optimizing flight paths through quantitative parameters, it improves operational efficiency and data acquisition completeness. Flight parameters can be quickly adjusted to adapt to different mission requirements, ensuring operational safety while improving data quality and operational efficiency.

[0029] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0030] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for UAV flight path planning based on a 3D reality model, characterized in that, The specific steps include the following: S1. Three-dimensional reality modeling based on oblique photogrammetry; S1.

1. Data Acquisition Planning; S1.

2. Image preprocessing; S1.

3. Aerial triangulation of images; S1.

4. Dense matching; S1.

5. Texture mapping; S2. Route planning based on 3D reality models; S2.

1. Import the 3D model; S2.

2. Set the task area; S2.

3. Set flight parameters; S2.4 Obstacle Avoidance and Safety Settings.

2. The UAV flight path planning method based on a three-dimensional real-scene model according to claim 1, characterized in that, The data acquisition planning method in step S1.1 is as follows: Based on the modeling accuracy requirements and the specific modeling object, select a suitable aircraft and camera, plan a reasonable flight path, and calculate a suitable aerial photography interval based on the camera focal length, sensor pixel size, flight altitude, and required ground resolution. The calculation formula is: ; In the formula, For camera focal length, For sensor pixel size, For flight altitude, Image width, For the required ground resolution For aerial photography intervals, This is an empirical coefficient.

3. The UAV flight path planning method based on a three-dimensional real-scene model according to claim 1, characterized in that, The image preprocessing method described in step S1.2 is as follows: Image preprocessing includes distortion correction and uniform illumination and color processing. Distortion correction uses in-camera orientation elements and distortion parameters for correction. The correction formula involves radial distortion and tangential distortion models. For a point on the image ( , ), corrected coordinates ( , The formula for calculating ) is: ; ; In the formula, the coordinates of the principal point are ( , ), focal length is The radial distortion parameters are respectively and The tangential distortion parameters are respectively and .

4. The UAV flight path planning method based on a three-dimensional real-scene model according to claim 1, characterized in that, The aerial triangulation method described in step S1.3 is as follows: The relative relationships between adjacent images are restored through feature extraction and feature matching. Feature extraction employs the SIFT algorithm, including scale-space extremum detection, precise keypoint localization, and determination of the principal orientation of the keypoints. The calculation formula is: In the formula, and The images are respectively in and Gradient of direction.

5. The UAV flight path planning method based on a three-dimensional real-scene model according to claim 1, characterized in that, The dense matching method described in step S1.4 is as follows: a semi-global matching algorithm is used to optimize the matching result by aggregating the matching costs from multiple directions. The formula for calculating the total matching cost is: In the formula, For matching cost, For the total matching cost, For pixels, For parallax, For pixels In the field, This is the penalty function.

6. The UAV flight path planning method based on a three-dimensional real-scene model according to claim 1, characterized in that, The texture mapping method described in step S1.5 is as follows: A texture is applied to the constructed 3D model mesh, and the mapping of each vertex on the texture image is defined. Coordinates are used to map a 2D image to a 3D model, with values ​​ranging from [value range missing]. .

7. The UAV flight path planning method based on a three-dimensional real-scene model according to claim 1, characterized in that, The method for importing the 3D model in step S2.1 is as follows: The generated 3D reality model is imported into the flight route planning software. The formula for transforming the model coordinate system to the world coordinate system using the coordinate transformation matrix is: In the formula, This is the coordinate transformation matrix. For the model coordinate system, Use the world coordinate system.

8. The UAV flight path planning method based on a three-dimensional real-scene model according to claim 1, characterized in that, The method for setting the task area in step S2.2 is as follows: The task area is defined on the 3D model, with three selection options: polygon, circle, and rectangle. For polygonal areas, the vertex coordinates are used to define the area. definition, For a circular region, the coordinates of the center are used. and radius Definition: For a rectangular region, the coordinates of its diagonal vertices are used to define the region. and definition.

9. The UAV flight path planning method based on a three-dimensional real-scene model according to claim 1, characterized in that, The method for setting flight parameters in step S2.3 is as follows: Flight parameters include flight altitude. Heading overlap rate Lateral overlap rate and flight speed Heading overlap and lateral overlap rate The calculation method is as follows: ; In the formula, The length of the overlapping area in the heading. This refers to the length of the flight path covered by a single image. The width of the lateral overlapping area. The lateral coverage width of a single image.

10. The UAV flight path planning method based on a three-dimensional real-scene model according to claim 1, characterized in that, The obstacle avoidance and safety setting method described in step S2.4 is as follows: mark obstacles on the 3D model and set a safety height. and emergency return point safe distance According to the size of the drone ,speed and reaction time The calculation method is as follows: In the formula, For deceleration.

Images