Orchard Mobile Robot Navigation Using Tree-Line Grid Mapping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
There is a need for a method and apparatus that enables mobile robots to autonomously, accurately, and stably navigate within orchard environments, as traditional methods face challenges in efficiently performing agricultural tasks like weeding, pest control, and harvesting due to the complexity of orchard layouts and the need for precise navigation.
Innovation Solution
The method involves converting 3D point cloud data into a 2D grid map, using local and global scan matching to determine the robot's position, extracting tree centers, generating effective lines based on tree arrangements, and creating driving paths that include straight and rotation paths, while considering obstacles and tree spacing, utilizing a processor and interface device to implement these steps.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If mobile robots are deployed in orchard environments, then labor consumption is reduced and productivity is improved, but navigation accuracy and stability deteriorate due to complex orchard layouts and tree arrangements
Solution Approach 1:
The patent divides the orchard environment into discrete grid cells, with each cell representing a specific spatial unit. This segmentation allows the robot to process complex orchard layouts systematically by evaluating occupancy probabilities and tree positions in each grid cell, thereby maintaining navigation stability while improving productivity.
Solution Approach 2:
The patent converts three-dimensional point cloud data into two-dimensional grid maps, projecting 3D spatial information onto a 2D plane. This dimensionality reduction simplifies the complex 3D orchard environment into a manageable 2D representation, enabling reliable path planning and navigation while maintaining accuracy in tree position detection.
2Device complexity
If traditional navigation methods are used in orchards, then device complexity is kept simple, but measurement precision and path accuracy deteriorate due to inability to accurately detect tree positions and navigate through rows
Solution Approach 1:
The patent replaces traditional mechanical navigation systems with laser-based point cloud scanning and computational processing. By using laser scanners to capture 3D environmental data and processing it through algorithms to generate 2D grid maps and detect tree positions, the system achieves high measurement precision without significantly increasing mechanical complexity.
Solution Approach 2:
The patent transforms raw point cloud data into structured grid map representations by changing the data format and dimensional parameters. This parameter transformation converts unstructured 3D point clouds into organized 2D grid cells with occupancy probabilities, enabling accurate tree position detection while maintaining computational efficiency.
3Measurement precision
If 3D point cloud data is processed directly, then measurement precision is high, but computational complexity and processing time increase, affecting real-time navigation performance
Solution Approach 1:
The patent reduces computational complexity by projecting three-dimensional point cloud data onto a two-dimensional grid plane. This dimensionality reduction transforms the complex 3D processing task into a simpler 2D grid evaluation task, maintaining essential spatial information while significantly reducing processing time for real-time navigation.
Solution Approach 2:
The patent extracts only the essential information from the complete point cloud data by generating a simplified grid map that captures occupancy probabilities and tree positions. This extraction process removes redundant data while retaining critical navigation information, achieving real-time processing performance without sacrificing measurement precision.
Data Source
AI summary
A method and an apparatus for recognition a position and generating a path for autonomous driving of a mobile robot in an orchard environment are provided. The method converts three-dimensional (3D) point cloud data for the orchard environment into a 2D grid map, obtains a position of the mobile robot by performing local scan matching and global scan matching on the 2D grid map, and extracts a tree from the 2D grid map based on occupancy of each grid of the 2D grid map. Then, an effective line based on center points of extracted trees is obtained, a destination based on the obtained effective line is generated, and a driving path according to the generated destination and the position of the mobile robot is generated.