Low-altitude flight collision detection method based on conical ray detection

Abstract

The invention discloses a low-altitude flight collision detection method based on conical ray detection. Acquiring environmental space data and position and attitude information of the unmanned aerial vehicle in a three-dimensional space; the flight direction of the unmanned aerial vehicle is obtained according to attitude information processing of the unmanned aerial vehicle in the three-dimensional space, the position of the unmanned aerial vehicle serves as a conical vertex, the flight direction serves as a central axis, and a conical detection area is constructed in combination with the field angle of the unmanned aerial vehicle and a preset detection distance; discretizing the conical detection area to generate a plurality of sub-rays covering the conical detection area; detecting a spatial relationship between each sub-ray and a spatial obstacle according to the environmental spatial data to obtain spatial relationship information of each sub-ray; and adjusting the attitude information of the unmanned aerial vehicle according to the spatial relationship information of each sub-ray. A conical ray detection model is introduced, and compared with traditional single-straight-line ray detection, missing detection and false detection are reduced.

Description

Technical Field

[0001] This invention relates to the field of low-altitude unmanned aerial vehicle (UAV) flight safety technology and intelligent obstacle avoidance, and in particular to a low-altitude flight collision detection method based on cone-shaped ray detection. Background Technology

[0002] Currently, low-altitude unmanned aerial vehicles (UAVs) are widely used in urban environments, complex terrains, and specific mission scenarios (such as inspection, surveying, and security monitoring). To ensure the safety of UAVs during flight and avoid collisions with obstacles such as buildings, vegetation, and power lines, collision detection methods based on sensors (such as lidar, ultrasonic sensors, and infrared ranging) or 3D modeling are commonly used both domestically and internationally.

[0003] In existing technologies, typical methods include: Direct sensor detection: The drone is equipped with lidar or Doppler radar to scan for obstacles ahead of its flight path, enabling obstacle avoidance. This method offers high real-time performance, but it suffers from the following problems: Limited detection range: The sensor's scanning angle is limited, unable to cover all directions, especially the lateral and overhead areas. Resolution is hardware-dependent and limited by sensor accuracy and sampling frequency, posing a risk of missed detections in long-range or small obstacle detection. High cost: High-precision radar equipment is expensive, increasing the overall cost of the drone.

[0004] Detection based on 3D geographic models: Using 3D terrain data or urban building models, the relative positions of the drone and obstacles are calculated to determine whether there is a collision risk. This type of method typically uses 3D visualization platforms (such as Cesium) for path simulation and combines GIS data for spatial analysis.

[0005] Its main shortcomings include: a disconnect between data and detection algorithms: some methods rely solely on model rendering to determine visibility, lacking high-precision integration with spatial databases, resulting in insufficient accuracy. The detection models are simplistic: most employ straight-line ray detection, failing to consider the coverage characteristics of conical rays, making it difficult to simulate the actual detection range of aircraft sensors. Data real-time performance is poor: when the amount of spatial data is large (such as high-density building vector data), detection calculations are time-consuming, making it difficult to achieve high real-time judgment during flight.

[0006] PostGIS Spatial Analysis Methods: As a commonly used spatial database extension, PostGIS possesses rich spatial query and analysis capabilities, which can be used to calculate buffers, distances, intersection relationships, etc. However, in existing applications, PostGIS often lacks real-time interaction with 3D visualization platforms, and collision detection mostly remains in the offline analysis stage, making it difficult to cope with the dynamic environmental changes of low-altitude flight.

[0007] In summary, existing technologies for low-altitude flight collision detection generally suffer from the following prominent problems: The detection accuracy relies on a single technology (hardware or software), and the lack of multi-source fusion leads to high false positive and false negative rates. The detection range and morphological modeling are not realistic enough to fully simulate the aircraft's detection characteristics. The low coupling between spatial data and the 3D rendering platform makes it difficult to balance real-time performance and accuracy. The failure to effectively distinguish between "visible obstacles" and "nearby obstacles" results in obstacle avoidance decisions including numerous irrelevant targets or overlooking critical risks.

[0008] Therefore, there is an urgent need for a method that can deeply integrate the dynamic rendering capabilities of a 3D visualization platform with the high-precision computing capabilities of a spatial database, and achieve high precision and real-time performance in low-altitude flight collision detection through a detection model that is more in line with physical characteristics (such as a cone ray). Summary of the Invention

[0009] To address the problems in the background art, this invention proposes a low-altitude flight collision detection method based on cone-shaped ray detection.

[0010] The technical solution adopted in this invention is: The method of the present invention includes the following steps: S1. Acquire environmental spatial data and the position and attitude information of the UAV in three-dimensional space; S2. Based on the attitude information of the UAV in three-dimensional space, the flight direction of the UAV is obtained. The position of the UAV is used as the vertex of the cone, the flight direction is used as the central axis, and the cone detection area is constructed by combining the UAV's field of view and the preset detection distance. S3. Discretize the cone-shaped detection area to generate multiple sub-rays covering the cone-shaped detection area; S4. Based on the environmental space data, detect the spatial relationship between each sub-ray and spatial obstacles to obtain the spatial relationship information of each sub-ray. S5. Adjust the attitude information of the UAV based on the spatial relationship information of each sub-ray.

[0011] The environmental spatial data is obtained in any one or more of the following ways: oblique photography, lidar scanning, BIM building model, and topographic mapping data.

[0012] The location of the UAV in three-dimensional space includes the UAV's latitude, longitude, and altitude information; The attitude information of a UAV in three-dimensional space includes heading angle, pitch angle, and roll angle.

[0013] Specifically, S3 involves dividing the cone-shaped detection area based on the drone's position and a preset angle, and generating sub-rays along each angle.

[0014] Specifically, S4 involves sampling the space within the cone-shaped detection area along the extension direction of each sub-ray based on environmental spatial data, and then determining the spatial relationships. If, within the preset detection distance, there is obstacle data in the environmental space data at the sampling position corresponding to the sub-ray, it is determined that there is a collision risk, and the first sampling point that meets this condition is recorded as the collision point. If, within the preset detection distance range, there is no obstacle data in the environmental space data at the sampling position corresponding to the sub-ray, and the spatial distance between the sampling point and the obstacle is less than or equal to the preset safe distance, it is determined that there is a close-range risk, and the corresponding sampling point, obstacle, and distance information between the corresponding sampling point and the obstacle are recorded. If, within the preset detection distance range, there are no obstacle data in the environmental space data at the sampling position corresponding to the sub-ray, and the spatial distance between any sampling point and any obstacle is greater than the preset safety distance, then it is determined to be safe. The obstacle data includes building models, terrain models, and tree models.

[0015] Specifically, S5 is: If a collision risk or close-range risk is determined, the drone's flight path is adjusted, and steps S1 to S4 are repeated until it is determined to be safe. If it is determined to be safe, the drone's attitude information will not be changed.

[0016] A computer device includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement the steps of the low-altitude flight collision detection method based on cone ray detection.

[0017] A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the low-altitude flight collision detection method based on cone ray detection.

[0018] The beneficial effects of this invention are: Significantly improved detection accuracy: This invention introduces a conical ray detection model, which can comprehensively cover the detection range in front of and to the sides of the aircraft, realistically simulating the physical detection characteristics of the aircraft's sensors. Compared with traditional single-line ray detection, it reduces missed detections and false detections. Utilizing PostGIS's three-dimensional spatial analysis functions (such as ST_3DIntersects and ST_3DDWithin), it achieves sub-meter accuracy obstacle collision determination, improving the reliability of the detection results.

[0019] High degree of spatial data integration: It deeply integrates Cesium's 3D visualization and rendering capabilities with PostGIS's spatial data management capabilities, achieving a seamless connection between visualization and precise calculation. Through a real-time spatial data interaction mechanism, the detection model can instantly acquire the latest obstacle information and adapt to dynamic environmental changes.

[0020] Comprehensive detection coverage: The conical ray detection range allows for parameterized definition of detection distance and field of view, flexibly adapting to different equipment models and mission requirements. The detection grid can adaptively adjust according to resolution requirements, ensuring sufficient spatial coverage density at both long and short distances.

[0021] Superior real-time performance and computational efficiency: Combining PostGIS spatial indexing (GiST / BRIN) technology and a cone-ray segmentation clipping algorithm, millisecond-level collision detection response is achieved in large-scale 3D data environments. Parallel computing and regional data clipping avoid global data scanning, improving detection efficiency and reducing server load.

[0022] Highly scalable and adaptable: Suitable for various low-altitude flight scenarios, including densely built-up urban areas, power line inspection corridors, and mountainous terrain. Conical ray parameters, detection thresholds, and early warning strategies can all be flexibly configured, facilitating integration with different flight control systems. Attached Figure Description

[0023] Figure 1 This is a schematic diagram of the collision detection process in this embodiment; Figure 2 This is a diagram showing the collision detection effect in this embodiment. Detailed Implementation

[0024] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0025] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and do not limit the scope of protection of this invention.

[0026] like Figure 1 As shown, the method of the present invention includes the following steps: S1. Acquire environmental spatial data of the low-altitude flight area and the current position and attitude information of the UAV in three-dimensional space; Specifically, the acquisition and modeling of spatial data for low-altitude flight environments involves obtaining spatial information about low-altitude flight areas from multi-source data sources (such as oblique photography, lidar scanning, BIM building models, and topographic mapping data). Data types include 3D building models, polygonal obstacles, point cloud data, and terrain data. The data undergoes coordinate unification and spatial accuracy correction, using a unified projected coordinate system (e.g., EPSG:4326 or EPSG:3857). The processed spatial data is then imported into a PostGIS database, and a corresponding spatial index (GiST or BRIN) is created to accelerate spatial query speed.

[0027] First, the flight status parameters of the UAV, such as its current position, flight altitude, heading angle, pitch angle and speed, are acquired in real time through the flight control system and then located in the Cesium 3D scene.

[0028] like Figure 2 As shown, the location and flight direction of the drone are represented by blue triangles in the scene. The scene data comes from terrain DEM, 3D building model (CityGML / 3DTiles), and obstacle vector data stored in PostGIS.

[0029] In this embodiment, data preparation involves storing 3,000 3D building models (including height information) in PostGIS and establishing a GiST index. The drone flies at an altitude of 80 meters, with a heading angle of 45°, a horizontal field of view of 60°, a detection range of 200 meters, and a safety radius of R=250 meters.

[0030] S2. Based on the attitude information of the UAV in three-dimensional space, the flight direction of the UAV is obtained. The current position of the UAV is used as the vertex of the cone, the flight direction is used as the central axis, and the cone detection area is constructed by combining the UAV's field of view and the preset detection distance. S3. Discretize the cone-shaped detection area to generate multiple sub-rays covering the cone-shaped detection area; Specifically, the cone-shaped ray detection model is constructed as follows: In Cesium, the cone-shaped detection area is constructed with the current position of the drone as the apex of the cone and the flight direction as the central axis.

[0031] Cone parameter definition: Vertex position: Real-time 3D coordinates of the UAV (latitude, longitude + altitude).

[0032] Central axis direction: calculated from the UAV attitude (heading angle, pitch angle, roll angle).

[0033] Cone angle: Set according to the drone's sensor field of view (FOV), such as 60° for the horizontal field of view and 40° for the vertical field of view.

[0034] Maximum detection range: Set according to mission requirements, for example, 200 meters.

[0035] The conical ray is discretized into several sub-rays to form a dense detection grid, so that every spatial region within the detection range is covered.

[0036] Based on the UAV's flight direction and preset detection parameters (detection radius R, horizontal field of view θ, horizontal resolution Δα, vertical resolution Δβ), a cone-shaped ray detection area is generated in Cesium. For example... Figure 2 As shown, the cone-shaped rays radiate forward from the position of the aircraft. The direction of each ray is calculated from the heading angle, pitch angle, and resolution parameters, forming a three-dimensional cone composed of multiple rays.

[0037] S4. Based on the environmental space data, detect the spatial relationship between each sub-ray and spatial obstacles to obtain the spatial relationship information of each sub-ray. Specifically, the fusion detection of conical rays and spatial data involves calculating the 3D spatial path of each sub-ray and transmitting it to the backend in real time. In PostGIS, each ray is converted into a 3D line segment object generated by ST_MakeLine, and spatial functions such as ST_3DIntersects and ST_3DDWithin are used to determine its 3D intersection relationship with obstacles.

[0038] Testing process: Input: Sub-ray geometry objects, obstacle geometry objects (buildings, terrain, etc.).

[0039] Determine if there is an intersection: If ST_3DIntersects(ray, obstacle) returns true, then record the coordinates and distance of the collision point.

[0040] Determine proximity: If ST_3DDWithin(ray, obstacle, safe distance) returns true, then trigger an early warning.

[0041] All detection results are returned to the Cesium front end for visualization, with collision points highlighted and evasion commands generated in the drone control system.

[0042] During the testing process, the optimization and real-time performance of test results are ensured. Spatial data cropping: Based on the current location of the UAV and the range of the cone-shaped envelope, ST_Intersects is used in the PostGIS to crop the data to the region, retaining only the obstacle data within the detection range, thus reducing the amount of computation.

[0043] Parallel computing: For collision detection of multiple sub-rays, multi-threaded or asynchronous tasks are used for parallel processing to improve real-time performance.

[0044] Dynamic updates: When the position or attitude of the drone changes, the cone ray is reconstructed in real time and collision detection is performed again.

[0045] In the PostGIS database, based on the aircraft's location and the outer cube extent of the conical ray, spatial queries such as ST_Intersects or ST_3DDWithin are used to quickly filter out obstacle data that may intersect the ray, thereby reducing subsequent computational load. For example... Figure 2 As shown, the gray area represents the outer perimeter of the cone-shaped ray, and the red area represents the set of candidate obstacles retrieved from the database.

[0046] For each cone-shaped ray, a 3D intersection test is performed with the candidate obstacle. The intersection point position is calculated using PostGIS's ST_3DIntersects or a custom bounding box segmentation and clipping algorithm, and the distance to the nearest intersection point is recorded. For example... Figure 2 As shown, green rays represent collision-free paths, while red rays represent collisions and mark the intersections.

[0047] First Channel: Obtaining the set of visible obstacles A based on cone-shaped rays. In the Cesium 3D scene, multiple discrete sub-rays (forming a cone-shaped probe) are emitted with the drone's current position as the vertex and its flight direction as the central axis; Using Cesium's raycasting; For each detected collision point (i.e. the intersection of a ray and the surface of a 3D model), the obstacle entity (such as a building, power line tower, etc.) to which the collision point belongs is determined through a spatial reverse lookup mechanism (such as model ID mapping, spatial nearest neighbor query, or attribute binding). If they intersect, record the corresponding obstacle IDs and summarize them to obtain the set of visible obstacles A.

[0048] Second channel: Obtaining the set of nearby obstacles B based on location neighborhood. Using the real-time three-dimensional position of the UAV as the center of a sphere (or ellipsoid), set a safety radius R (e.g., 250 meters). In PostGIS, execute ST_3DDWithin(drone location, obstacle, R) to filter out all nearby obstacles; The set of nearby obstacles, B, is obtained by summing them up.

[0049] Phase 3: Integration of Sets and Intersections Calculate the final obstacle list C = A ∩ B; Collision warnings, distance calculations, and visual markers are only performed for obstacles in set C. If C is empty, the current path is considered safe.

[0050] Optimization and Real-Time Assurance of Detection Results Before performing step 3, you can first clip the full PostGIS data by cone-shaped bounding box or neighborhood range (ST_Intersects) to further narrow the query range; The construction of sets A and B can be performed in parallel; Dynamic update mechanism: When the UAV pose changes, A, B and their intersection C are updated synchronously.

[0051] Generate set A Cesium emits 1800 sub-rays (1° resolution); The backend detected that 12 buildings were hit by rays, forming a set A={B3, B7, B15,...,B210}.

[0052] Generate set B Use the drone's location as the center to query all buildings within 250 meters; A total of 48 buildings were returned, forming a set B = {B1, B3, B5, ..., B301}. Three-dimensional collision detection of rays and obstacles. Intersection and Fusion Calculate C = A ∩ B = {B3, B15, B210} (3 buildings in total); Collision distance calculations and warnings are only performed for these three buildings.

[0053] Results verification: The total detection time was 38ms (of which A construction took 22ms, B construction took 10ms, intersection took 2ms, and the rest was I / O). Compared with the traditional "ray detection only" method (which missed detecting 2 side buildings) and the "neighborhood query only" method (which falsely reported 15 buildings outside the field of view), this solution has a 0% missed detection rate and a 85% reduction in false alarm rate.

[0054] S5. Adjust the UAV's attitude information and update the UAV's flight path based on the spatial relationship information of each sub-ray.

[0055] In this embodiment, a 3D model of the cone-shaped ray and the detection results are drawn on the Cesium front end for dynamic display. When a collision risk is detected, the system automatically calculates an alternative flight path and displays it as a dashed line in the 3D scene. Combined with the flight control system interface, avoidance commands are sent to the UAV to achieve automatic obstacle avoidance or assisted piloting.

[0056] This invention utilizes a conical ray to simulate the detection field of view of a UAV and fuses this 3D detection model with high-precision PostGIS spatial data. Through algorithms such as spatial intersection and distance judgment, it achieves high-precision identification and early warning of obstacles in low-altitude flight paths. Compared with traditional straight-line rays, the conical detection range is closer to the physical characteristics of aircraft sensors. Combined with PostGIS spatial indexing, it enables real-time collision detection and visualization in a large-scale 3D data environment.

[0057] To avoid misjudgments caused by insufficient accuracy of obstacle models, this invention introduces a buffer zone strategy, which involves extending a certain distance (e.g., 0.5 meters) outward from the obstacle surface before performing collision detection, thereby enhancing the safety margin.

[0058] Meanwhile, the ray resolution can be dynamically adjusted according to flight speed and mission accuracy to ensure sufficient collision reaction time even at high speeds.

[0059] After the detection is complete, the collision information is returned to the flight control system, including: collision distance, azimuth angle, obstacle type, etc., and the danger zone is highlighted in real time in the Cesium scene. Figure 2 As shown, the red area is the high-risk collision zone, the yellow area is the medium-risk buffer zone, and the green area is the safe flight zone.

[0060] The method of this invention constructs a cone-shaped ray detection model through the three-dimensional visualization platform Cesium and integrates it with the three-dimensional spatial data stored in the PostGIS spatial database for analysis, thereby achieving high-precision, real-time judgment of potential collision risks of low-altitude aircraft in complex environments.

[0061] Environmental spatial data for low-altitude flight areas are obtained in any one or more of the following ways: oblique photography, lidar scanning, BIM building model, and topographic mapping data.

[0062] The location of a drone in three-dimensional space includes its latitude, longitude, and altitude information; The attitude information of a UAV in three-dimensional space includes heading angle, pitch angle, and roll angle.

[0063] S3 specifically involves dividing the cone-shaped detection area based on the drone's position, according to preset angles and distances, and generating sub-rays along each division angle.

[0064] S4 specifically involves sampling the space within the cone-shaped detection area along the extension direction of each sub-ray based on environmental spatial data, and then determining spatial relationships: If, within the preset detection distance, there is obstacle data in the environmental space data at the sampling position corresponding to the sub-ray, it is determined that there is a collision risk, and the first sampling point that meets this condition is recorded as the collision point. If, within the preset detection distance range, there is no obstacle data in the environmental space data at the sampling position corresponding to the sub-ray, and the spatial distance between any sampling point and any obstacle is less than or equal to the preset safe distance, then it is determined that there is a close-range risk, and the corresponding sampling point, obstacle, and distance information between the corresponding sampling point and obstacle are recorded when the spatial distance is less than or equal to the preset safe distance. If, within the preset detection distance range, there are no obstacle data in the environmental space data at the sampling position corresponding to the sub-ray, and the spatial distance between any sampling point and any obstacle is greater than the preset safety distance, then it is determined to be safe. Obstacle data includes building models, terrain models, trees, or other pre-defined solid object models.

[0065] S5 specifically refers to: If a collision risk or close-range risk is determined, the drone's flight path is adjusted to move away from the collision risk area, and steps S1 to S4 are repeated until it is determined to be safe. If it is determined to be safe, the drone's attitude information will not be changed.

[0066] This method constructs a conical ray model and integrates it with high-precision spatial data to achieve high-precision, real-time detection and judgment of potential obstacles in the flight path, significantly improving the accuracy and reliability of collision detection in low-altitude flight environments. It is a cross-application technology of UAV three-dimensional spatial perception and geographic information system (GIS) data processing.

[0067] This invention combines three-dimensional spatial obstacle data from a Geographic Information System (GIS) and employs an intersection fusion strategy of "ray collision obstacle set" and "location neighborhood obstacle set" to achieve high-precision, low-redundancy, and real-time identification of potential collision targets.

[0068] Using the method of this invention, in simulated urban low-altitude flight tests (60° field of view, 200-meter detection radius, 1° ray resolution), the collision detection time is controlled within 40ms among 3000 three-dimensional obstacle models stored in PostGIS, and the missed detection rate is less than 1%, meeting the engineering requirements for real-time flight obstacle avoidance.

[0069] The above detailed embodiments illustrate the technical solution and beneficial effects of the present invention. It should be understood that the above description is only the most preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, additions, and equivalent substitutions made within the scope of the principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A low-altitude flight collision detection method based on cone-shaped ray detection, characterized in that, The method includes the following steps: S1. Acquire environmental spatial data and the position and attitude information of the UAV in three-dimensional space; S2. Based on the attitude information of the UAV in three-dimensional space, the flight direction of the UAV is obtained. The position of the UAV is used as the vertex of the cone, the flight direction is used as the central axis, and the cone detection area is constructed by combining the UAV's field of view and the preset detection distance. S3. Discretize the cone-shaped detection area to generate multiple sub-rays covering the cone-shaped detection area; S4. Based on the environmental space data, detect the spatial relationship between each sub-ray and spatial obstacles to obtain the spatial relationship information of each sub-ray. S5. Adjust the attitude information of the UAV based on the spatial relationship information of each sub-ray.

2. The low-altitude flight collision detection method based on cone-shaped ray detection according to claim 1, characterized in that: The environmental spatial data is obtained in any one or more of the following ways: oblique photography, lidar scanning, BIM building model, and topographic mapping data.

3. The low-altitude flight collision detection method based on cone-shaped ray detection according to claim 1, characterized in that: The location of the UAV in three-dimensional space includes the UAV's latitude, longitude, and altitude information; The attitude information of a UAV in three-dimensional space includes heading angle, pitch angle, and roll angle.

4. The low-altitude flight collision detection method based on cone-shaped ray detection according to claim 1, characterized in that: Specifically, S3 involves dividing the cone-shaped detection area based on the drone's position and a preset angle, and generating sub-rays along each angle.

5. The low-altitude flight collision detection method based on cone-shaped ray detection according to claim 1, characterized in that: Specifically, S4 involves sampling the space within the cone-shaped detection area along the extension direction of each sub-ray based on environmental spatial data, and then determining the spatial relationships. If, within the preset detection distance, there is obstacle data in the environmental space data at the sampling position corresponding to the sub-ray, it is determined that there is a collision risk, and the first sampling point that meets this condition is recorded as the collision point. If, within the preset detection distance range, there is no obstacle data in the environmental space data at the sampling position corresponding to the sub-ray, and the spatial distance between the sampling point and the obstacle is less than or equal to the preset safe distance, it is determined that there is a close-range risk, and the corresponding sampling point, obstacle, and distance information between the corresponding sampling point and the obstacle are recorded. If, within the preset detection distance range, there are no obstacle data in the environmental space data at the sampling position corresponding to the sub-ray, and the spatial distance between any sampling point and any obstacle is greater than the preset safety distance, then it is determined to be safe. The obstacle data includes building models, terrain models, and tree models.

6. The low-altitude flight collision detection method based on cone-shaped ray detection according to claim 1, characterized in that: Specifically, S5 is: If a collision risk or close-range risk is determined, the drone's flight path is adjusted, and steps S1 to S4 are repeated until it is determined to be safe. If it is determined to be safe, the drone's attitude information will not be changed.

7. 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 low-altitude flight collision detection method based on cone ray detection as described in any one of claims 1 to 6.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the low-altitude flight collision detection method based on cone ray detection as described in any one of claims 1 to 6.

Images