A main route generation method based on spatial position relationship of seedling points under orthographic images

Abstract

This invention discloses a method for generating a main flight path based on the spatial relationship of seedling points under orthophoto photography, belonging to the field of agricultural engineering and intelligent navigation technology. This invention uses the set of seedling point coordinates extracted from orthophotos as input, and combines a dynamic search radius mechanism and a direction constraint strategy to gradually generate a continuous and reasonable main flight path. The system includes a seedling point identification module, a spatial filtering module, a direction constraint module, and a path generation module. The main flight path generated by this invention enables the main walking mechanism of the equipment to be precisely positioned between crop rows, while its arm-reaching working mechanism covers a preset number of crop rows. Furthermore, it directly uses the actual seedling point coordinates extracted from orthophotos as the basis for path generation, maximizing the fit between the actual growth positions of the crops. This allows the walking mechanism of the agricultural equipment to accurately traverse between seedling points, effectively avoiding the crop root area, significantly reducing the risk of damage to crops from the equipment wheels, and improving the safety of field operations.

Description

Technical Field

[0001] This invention relates to the fields of agricultural engineering and intelligent navigation technology, and in particular to a method for generating a main flight path based on the spatial positional relationship of seedling points under orthophoto photography. Background Technology

[0002] With the rapid development of intelligent agricultural equipment, autonomous navigation capability has become one of the key technologies for field operation robots to achieve efficient operation. Especially in the scenario of inter-row operation, the accuracy of navigation path directly affects the operation efficiency and crop protection effect.

[0003] Current agricultural navigation path generation methods are mostly based on image fitting or ensemble modeling strategies. While these methods have a certain degree of versatility, they often neglect the actual spatial relationships of crop seedlings, causing the generated paths to deviate from the actual seedling distribution and increasing the risk of damage to crop roots from equipment wheels. Furthermore, the complex field environment, uneven distribution of seedlings, undulating terrain, and image noise further exacerbate the difficulty of path planning.

[0004] To address the aforementioned problems, this invention proposes a main flight path generation method based on the positional relationships of seedling points under orthophoto photography. This method utilizes orthophoto images of the field acquired by UAVs or high-position cameras. Through image segmentation, boundary removal, seedling point identification, and result stitching, it constructs the spatial positional relationships of seedling points within the target plot and generates a continuous main flight path by combining a dynamic search radius and direction constraint strategy. This method can maximize the fit to the actual growth location of the plants, improve path accuracy and robustness, adapt to complex terrain and seedling point distribution environments, and provide safer and more reliable navigation paths for agricultural robots. Summary of the Invention

[0005] The purpose of this invention is to address the problem that existing agricultural navigation path generation methods generally employ set fitting strategies, which ignore or obscure the actual spatial relationships of crops during path construction, leading to generated paths that deviate from the actual distribution of seedlings and thus increasing the risk of damage to crop roots by equipment wheels. The invention proposes a main route generation method based on the spatial relationships of seedlings under orthophoto photography.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: A method for generating a main flight path based on the spatial relationship of seedling points under orthophoto photography includes the following steps: S1: Acquire crop orthophotos of the target plot, segment and detect boundaries in the orthophotos, remove areas without information, and obtain several effective small image blocks; S2: Identify seedling points for each effective small image block, extract seedling point coordinate information, and stitch the identification results together to construct a complete set of seedling points for the target plot and determine the spatial relationship of the seedling points. S3: Initialize path generation parameters, including: setting the current navigation point, initializing the search radius range, defining the maximum search radius and minimum step size, and setting the threshold for the number of seedling points to be explored; S4: In each iteration, calculate the spatial distance between each seedling point in the seedling point set and the current navigation point, filter out the candidate seedling point set that satisfies the current search radius constraint, and exclude seedling points that have been recorded in the visited set and seedling points whose direction deviates from the set from the candidate seedling point set. S5: Based on the preset reference direction, apply directional constraints to the filtered candidate seedling point set, select seedling points whose angle with the current navigation direction is within a preset range as target navigation points, add the target navigation points to the visited set, and update the current navigation point; S6: When no target navigation point is found within the preset range, enter turning mode, adjust the search area to a fan-shaped area perpendicular to the current heading direction, and re-select target navigation points within the fan-shaped area; S7: Repeat steps S4 to S6 until the path generation termination condition is met, and output the main route consisting of all target navigation points.

[0007] As a preferred embodiment, in step S4, if the number of candidate seedling points selected under the current search radius is less than the threshold for the number of pre-explored seedling points, the search radius is gradually expanded until the number of candidate seedling points reaches the threshold or the search radius reaches the maximum search radius.

[0008] As a preferred embodiment, the direction deviation set in step S4 is used to record points whose angle with the current main heading is within a preset range, but whose direction quadrant is inconsistent with the main route's forward direction, in order to avoid reverse jumps or reversals in the path.

[0009] As a preferred embodiment, in step S6, the search area in the turning mode is a fan-shaped area, and its angle range is preset to a threshold. The selection of the target navigation point prioritizes directional continuity and vertical distance from the current navigation point to ensure a smooth path transition.

[0010] As a preferred option, the path generation termination conditions in step S7 include: no valid candidate seedling points can be selected in both straight-going mode and turning mode, no valid candidate points are found in multiple consecutive rounds of searching, or the current search radius has reached the maximum search radius and the number of candidate sets is still less than the pre-explored seedling point number threshold.

[0011] As a preferred embodiment, the method is applicable to both offline centralized processing of orthophoto images taken by UAVs to generate the main flight path and sending it to the operating equipment, and to real-time calculation and path generation of equipment equipped with forward-looking cameras during field operations.

[0012] A main flight path generation system based on the spatial positional relationship of seedling points under orthophoto photography includes: Image acquisition and preprocessing module: used to acquire crop orthophotos of the target plot, and to segment and detect the boundaries of the orthophotos, remove informationless areas, and obtain several effective small image blocks; Seedling point identification and localization module: used to identify seedling points in each of the effective small image blocks, extract seedling point coordinate information, stitch the identification results together, construct a complete set of seedling points for the target plot, and determine the spatial relationship of seedling points; Parameter initialization module: used to initialize path generation parameters; Candidate point filtering module: Used in each iteration to filter the set of candidate seedling points based on the current navigation point and search radius, and exclude seedling points in the visited set and the direction deviation set; Target point determination module: used to select the target navigation point from the candidate seedling point set according to the direction constraints, update the current navigation point and record the visited set; Turning mode control module: used to switch to a fan-shaped area perpendicular to the current heading direction for searching when there are no candidate points in the straight direction; Path generation and termination module: Used to control the iteration process and output the main route consisting of all target navigation points when the termination condition is met.

[0013] As a preferred embodiment, the candidate point screening module is further configured to gradually expand the search radius until the threshold or the maximum search radius is reached when the number of candidate seedling points in the current search radius is less than the threshold for the number of pre-explored seedling points.

[0014] As a preferred embodiment, in the turning mode control module, the selection of the target navigation point prioritizes directional continuity and the perpendicular distance from the current navigation point.

[0015] Compared with the prior art, the beneficial effects of the present invention are as follows: 1. Precisely matches the actual location of crops, reducing the risk of crop damage; This invention abandons traditional path generation methods based on set fitting or fixed row spacing, and directly uses the actual seedling coordinates extracted from orthophoto images as the basis for path generation. The generated main path closely matches the actual growth position of the crop, enabling the walking mechanism of agricultural equipment to accurately traverse between seedlings, effectively avoiding the crop root area, significantly reducing the risk of damage to crops from equipment wheels, and improving the safety of field operations.

[0016] 2. Dynamically adapt to changes in seedling distribution to ensure path continuity; By introducing a dynamic search radius mechanism, this invention can adaptively adjust the search range according to local changes in seedling density. When seedlings are sparse, the algorithm automatically expands the search radius to find subsequent points and avoids path interruption. When seedlings are dense, it maintains an appropriate step size to ensure that the path is close to the seedling strip, effectively solving the path planning problem caused by uneven distribution of seedlings in the field and making the generated main route have good continuity.

[0017] 3. Intelligent directional constraints and corrections ensure consistent path direction; This invention sets up a direction deviation set to record abnormal seedling points that are inconsistent with the main heading direction quadrant and excludes them during candidate point screening, thus avoiding reverse jumps, reversals, or sawtooth deviations in the path. At the same time, a direction constraint screening is introduced in each iteration to ensure that the path extends stably along the crop row direction. Even when there is a slight deviation in crop planting, direction correction can be achieved, significantly improving the rationality and robustness of the path.

[0018] 4. Smooth turning transitions, adaptable to field boundary operations; When a sudden change in direction or approach to a plot boundary is detected, the invention automatically switches to turning mode, adjusting the search area to a fan-shaped region perpendicular to the heading, and prioritizing seedlings with strong directional continuity and small vertical distances as target points. This strategy ensures a smooth transition at the end of crop rows, avoiding abrupt sharp turns and providing reliable navigation guidance for agricultural equipment to turn around in the field.

[0019] 5. Dual application modes, balancing offline accuracy and online real-time performance; This invention supports both offline high-precision processing of orthophoto images captured by UAVs to generate a globally optimal master flight path before sending it to the operating equipment, and online real-time calculation and path generation in the field using equipment equipped with forward-looking cameras. This flexible application mode balances the high-precision processing required at the laboratory level with the real-time needs of field operations, demonstrating strong adaptability and broad application prospects.

[0020] 6. Compatible with boom-type agricultural machinery, further enhancing operational safety capabilities; For boom-type agricultural equipment, the main path generated by this invention enables the main walking mechanism of the equipment to be precisely positioned between crop rows, while its boom working mechanism covers a preset number of crop rows. This targeted path planning ensures the working width while minimizing the crushing of crops by the main body of the equipment, providing technical support for the precision operation of large agricultural machinery. Attached Figure Description

[0021] Figure 1 This is a flowchart of a main flight path generation method based on the spatial positional relationship of seedling points under orthophoto photography proposed in this invention; Figure 2This is a detailed technical flowchart of the "candidate seedling point screening" step in the main flight path generation method based on the spatial positional relationship of seedling points under orthophoto photography proposed in this invention. Figure 3 This is a schematic diagram of the "candidate seedling point screening" step in the main flight path generation method based on the spatial positional relationship of seedling points under orthophoto photography proposed in this invention. Figure 4 This is a schematic diagram of the main flight path passing through seedling points in a main flight path generation method based on the spatial positional relationship of seedling points under orthophoto photography proposed in this invention; Figure 5 This is a schematic diagram of the main flight path adapted to the boom-type agricultural machinery in the main flight path generation method based on the spatial positional relationship of seedling points under orthophoto photography proposed in this invention. Detailed Implementation

[0022] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0023] It should be understood that, when used in this specification and the appended claims, the terms "comprising" and "including" indicate the presence of the described features, integrals, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or collections thereof.

[0024] It should also be understood that the terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to limit the invention. As used in this specification and the appended claims, the singular forms “a,” “an,” and “the” are intended to include the plural forms unless the context clearly indicates otherwise.

[0025] It should also be further understood that the term "and / or" as used in this specification and the appended claims refers to any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0026] Example, refer to Figures 1 to 5 A method for generating a main flight path based on the spatial relationship of seedling points under orthophoto photography includes the following steps: Step 1: Use a drone or a device equipped with a high-position camera to acquire orthophoto images of crops in the field, and make a preliminary judgment based on the image size; Step 2: If the image size is large, the image is segmented into several small regions, and boundary detection is performed to remove regions with no information. If a high-position camera is used, real-time seedling identification and flight path construction are required. The main flight path generation logic remains unchanged, but parameters such as seedling points in the candidate area need to be fine-tuned according to the actual crop planting situation (agronomy). The current introduction mainly focuses on the processing flow after obtaining large-scale field orthographic images. Step 3: Perform seedling point identification within the segmented effective small block area, extract seedling point coordinate information, and construct a local seedling point set; Step 4: Combine the identification results of each small block, unify the coordinate system, and form a complete spatial distribution data of large seedling points; Step 5: Initialize the navigation point positions, set the search radius range, minimum step size, maximum search radius, and threshold for the number of seedling points to be explored; Step 6: In each iteration, calculate the spatial distance between the current navigation point and all seedling points, filter out the candidate seedling point set that meets the distance constraints, and exclude seedling points in the visited set and the direction deviation set; Step 7: Perform directional constraint filtering on the seedling points in the candidate set, limit the range of angles between the seedling points and the current navigation direction, select seedling points with strong directional continuity as target points, and update the navigation point positions; Step 8: Add the selected target point to the visited set and record the path extension process; Step 9: When there are no target points that meet the conditions in the navigation direction, it is determined that the boundary area may be approached, and the turning mode is entered. The search area is adjusted to a fan-shaped area perpendicular to the navigation direction (the angle is generally within ±30°). The target points are re-selected, and the shortest distance and reasonable direction are given priority to realize the path turning logic. Step 10: When no valid candidate seedling points can be selected in either the straight-ahead mode or the turning mode, the current path segment generation is considered complete; at this point, the recorded seedling point access order is the actual generated trajectory of the main route.

[0027] During path generation, this invention introduces a dynamic search radius mechanism to adapt to local changes in seedling density. When the number of candidate seedlings within the current search radius is less than a preset threshold for the number of pre-explored seedlings, the algorithm automatically and gradually expands the search radius until the number of candidate seedlings reaches the threshold or the maximum search radius. This mechanism ensures that the path continuity is maintained even in sparse seedling areas.

[0028] This invention establishes a direction deviation set to record potential points whose angle with the current main heading is within a preset range, but whose direction quadrant is inconsistent with the main route's forward direction. Selecting these potential points could lead to path reversals or detours. By including them in the direction deviation set and excluding them from the candidate points, path disorder is effectively avoided.

[0029] The main flight path generated by this invention is adaptable to reach-out agricultural machinery. In this application scenario, path planning ensures that the main walking mechanism of the equipment is positioned between crop rows, while the reach-out working mechanism can cover a preset number of crop rows. Figure 5 For example, the vehicle itself spans three rows of crops, and its arm spans one row of crops. By guiding the precise seedling position relationship, the damage to the crop roots caused by the wheels can be minimized.

[0030] The algorithm can be used for offline processing of images acquired by UAVs on a high-performance computer in the laboratory, or it can be deployed on devices equipped with forward-looking cameras to realize real-time path generation and navigation control in the field, thus taking into account the application requirements of offline high-precision processing and online real-time navigation.

[0031] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A main route generation method based on the spatial position relationship of seedling points under orthographic images, characterized in that, Includes the following steps: S1: Acquire crop orthophotos of the target plot, segment and detect boundaries in the orthophotos, remove areas without information, and obtain several effective small image blocks; S2: Identify seedling points for each effective small image block, extract seedling point coordinate information, and stitch the identification results together to construct a complete set of seedling points for the target plot and determine the spatial relationship of the seedling points. S3: Initialize path generation parameters, including: setting the current navigation point, initializing the search radius range, defining the maximum search radius and minimum step size, and setting the threshold for the number of seedling points to be explored; S4: In each iteration, calculate the spatial distance between each seedling point in the seedling point set and the current navigation point, filter out the candidate seedling point set that satisfies the current search radius constraint, and exclude seedling points that have been recorded in the visited set and seedling points whose direction deviates from the set from the candidate seedling point set. S5: Based on the preset reference direction, apply directional constraints to the filtered candidate seedling point set, select seedling points whose angle with the current navigation direction is within a preset range as target navigation points, add the target navigation points to the visited set, and update the current navigation point; S6: When no target navigation point is found within the preset range, enter turning mode, adjust the search area to a fan-shaped area perpendicular to the current heading direction, and re-select target navigation points within the fan-shaped area; S7: Repeat steps S4 to S6 until the path generation termination condition is met, and output the main route consisting of all target navigation points.

2. The main route generation method based on the spatial position relationship of seedling points in orthographic images according to claim 1, characterized in that, In step S4, if the number of candidate seedling points selected under the current search radius is less than the pre-exploration seedling point number threshold, the search radius is gradually expanded until the number of candidate seedling points reaches the threshold or the search radius reaches the maximum search radius.

3. The method for generating a main flight path based on the spatial positional relationship of seedling points under orthophoto photography according to claim 1, characterized in that, The direction deviation set in step S4 is used to record points whose angle with the current main heading is within a preset range, but whose direction quadrant is inconsistent with the main route's forward direction, in order to avoid reverse jumps or reversals in the path.

4. The method for generating a main flight path based on the spatial positional relationship of seedling points under orthophoto photography according to claim 1, characterized in that, In step S6, the search area in the turning mode is a fan-shaped area with its angle range preset to a threshold. The selection of the target navigation point prioritizes directional continuity and vertical distance from the current navigation point to ensure a smooth path transition.

5. The method for generating a main flight path based on the spatial positional relationship of seedling points under orthophoto photography according to claim 1, characterized in that, The path generation termination conditions in step S7 include: no valid candidate seedling points can be selected in both straight and turning modes, no valid candidate points are found in multiple consecutive rounds of searching, or the current search radius has reached the maximum search radius and the number of candidate points is still less than the pre-explored seedling point number threshold.

6. A method for generating a main flight path based on the spatial positional relationship of seedling points under orthophoto photography, as described in claims 1 to 5, characterized in that, The method is applicable to both offline centralized processing of orthophoto images taken by UAVs to generate the main flight path and send it to the operating equipment, and to real-time calculation and path generation of equipment equipped with forward-looking cameras during field operations.

7. The system proposed in any one of claims 1 to 6 for generating a main flight path based on the spatial positional relationship of seedling points under orthophoto photography, characterized in that, include: Image acquisition and preprocessing module: used to acquire crop orthophotos of the target plot, and to segment and detect the boundaries of the orthophotos, remove informationless areas, and obtain several effective small image blocks; Seedling point identification and localization module: used to identify seedling points in each of the effective small image blocks, extract seedling point coordinate information, stitch the identification results together, construct a complete set of seedling points for the target plot, and determine the spatial relationship of seedling points; Parameter initialization module: used to initialize path generation parameters; Candidate point filtering module: Used in each iteration to filter the set of candidate seedling points based on the current navigation point and search radius, and exclude seedling points in the visited set and the direction deviation set; Target point determination module: used to select the target navigation point from the candidate seedling point set according to the direction constraints, update the current navigation point and record the visited set; Turning mode control module: used to switch to a fan-shaped area perpendicular to the current heading direction for searching when there are no candidate points in the straight direction; Path generation and termination module: Used to control the iteration process and output the main route consisting of all target navigation points when the termination condition is met.

8. A main flight path generation system based on the spatial positional relationship of seedling points under orthophoto photography according to claim 7, characterized in that, The candidate point screening module is also used to gradually expand the search radius until the threshold is reached when the number of candidate seedling points in the current search radius is less than the threshold for the number of pre-explored seedling points.

9. A main flight path generation system based on the spatial positional relationship of seedling points under orthophoto photography according to claim 7, characterized in that, In the turning mode control module, the selection of the target navigation point prioritizes directional continuity and the perpendicular distance from the current navigation point.

Images