Rotation-Invariant Markers for Power-Free Agricultural Machine Positioning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing agricultural working machines that use fiducial markers for navigation require a power source, limiting their functionality in tools that cannot be powered, and existing solutions for image-based navigation are cumbersome and require human verification.
Innovation Solution
The implementation of rotation-invariant recognition patterns that can be recognized by sensor devices without additional electrical energy, allowing for autonomous adjustment and control of agricultural machines using a sensor device and evaluation module for three-dimensional positioning and orientation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If LED fiducial markers are used to track the position of a towed implement, then the position tracking is improved, but a power source is required which limits functionality in non-powered implements
Solution Approach 1:
The patent replaces active electronic fiducial markers (LEDs requiring power) with passive optical recognition patterns that can be detected by sensor devices. This substitution eliminates the power source requirement while maintaining position tracking capability through optical detection of geometric patterns.
Solution Approach 2:
The patent uses simple, passive recognition patterns that can be printed or painted directly on implements rather than expensive electronic components. These patterns are inexpensive to produce and apply, enabling position tracking on any implement regardless of power availability.
2Measurement precision
If mobile devices are used to detect fiducial markers, then recognition capability is improved, but human verification is required which reduces automation
Solution Approach 1:
The patent implements autonomous recognition systems where the agricultural working machine itself performs fiducial marker detection and position determination without requiring human verification. The evaluation module automatically processes sensor data to determine machine position and orientation, enabling fully automated operation.
Solution Approach 2:
The patent establishes a closed-loop feedback system where sensor devices continuously detect recognition patterns, the evaluation module processes this data to determine position and orientation, and the control system automatically adjusts machine operations based on this information, eliminating the need for human intervention.
3Measurement precision
If rotation-invariant recognition patterns are used, then recognition accuracy under various orientations is improved, but pattern design complexity increases
Solution Approach 1:
The patent employs asymmetric geometric shapes (such as L-shapes, T-shapes, or arrows) within the recognition patterns that have distinct orientation characteristics. These asymmetric features allow the system to determine the precise rotation angle of the pattern, enabling accurate position and orientation calculation even when the implement is rotated to various angles.
Solution Approach 2:
The patent uses three-dimensional positioning of multiple recognition patterns on implements, adding spatial depth as another dimension for identification. By arranging patterns at different heights or positions in 3D space, the system can distinguish between different implements and determine their spatial orientation more accurately.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The present invention relates to an agricultural machine (1) for use in an agricultural field (2) with a sensor device (3) for generating an image (4) of an area (5) surrounding the agricultural machine (1). The present invention is based on the general idea that the agricultural machine (1) is adjusted depending on a rotationally invariant recognition pattern (7), which is also called a rotationally invariant fiducial marker.