AI Crop Edge Detection for Self-Propelled Harvester Steering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Manual steering of self-propelled harvesters can lead to incomplete crop processing due to human error, resulting in missed rows and inefficiency, and existing systems using laser sensors are costly.
Innovation Solution
A system employing a camera and computing unit with artificial intelligence, specifically a trained neural network, to detect and delineate the crop edge by segmenting images into 'plant population' and 'background' classes, enabling precise tracking and automatic steering of the harvester.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If manual steering is used to align the harvester with the crop, then the harvester can be operated with simple equipment, but the alignment precision deteriorates due to human error causing missed rows
Solution Approach 1:
The patent replaces manual mechanical steering with an automated optical detection and control system. A camera captures images of the crop edge, a computing unit processes these images to determine precise crop edge positions, and this information is used to automatically steer the harvester, eliminating human error in alignment while maintaining operational simplicity through automation.
2Measurement precision
If laser sensors are used to determine the crop edge, then the alignment precision improves, but the system cost deteriorates due to expensive laser sensor equipment
Solution Approach 1:
The patent replaces expensive laser sensors with a more cost-effective camera-based imaging system. The camera captures visual information of the crop edge, which is then processed by a computing unit to achieve precise alignment. This substitution significantly reduces system cost while maintaining the necessary measurement precision for accurate harvester alignment.
Solution Approach 2:
The patent substitutes optical laser-based detection with optical camera-based detection. Instead of using laser sensors to measure crop edge position, the system uses a camera to capture images that are subsequently processed by image analysis algorithms, achieving the same functional goal at lower cost.
3Device complexity
If manual steering with continuous driver input is used, then the system complexity remains low, but the reliability deteriorates due to high source of human error
Solution Approach 1:
The patent implements a self-steering system where the harvester automatically determines and maintains its position relative to the crop edge. The camera continuously monitors the crop edge, the computing unit processes the visual data to calculate position and orientation, and the steering system automatically adjusts the harvester's path without requiring continuous manual input, thereby eliminating human error while keeping the system relatively simple.
Solution Approach 2:
The patent establishes a closed-loop feedback system where the camera continuously captures crop edge position, the computing unit processes this information to determine deviation from the desired path, and the steering system adjusts accordingly. This real-time feedback mechanism ensures reliable harvesting by continuously correcting the harvester's position based on actual crop edge location.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The present application relates to a system (1) for determining a crop edge (2). In order to provide a system (1) for determining a crop edge (2) that overcomes the disadvantages of the prior art and simplifies the guidance of a self-propelled harvesting machine (1) into a crop stand (10), the invention provides that the system comprises a computing unit (3) and at least one camera (4), wherein the camera (4) is provided and configured to capture optical information in the form of discrete images (5) of a front environment (6) of an agricultural harvesting machine (7), wherein the computing unit (3) and the camera (4) are interconnected in a data-transmitting manner such that the images (5) can be transmitted to the computing unit (3), and wherein the computing unit (3) is provided and configured to process at least one of the images (5) using artificial intelligence.that a plant area (8) of a field (9) on which a stand of plants (10) is located is delimited from a remaining residual area (11) of the field (9) and the stand of plants (10) is determined in this way, wherein the computing unit (3) is provided and configured to determine the stand edge (2) of the stand of plants (10) based on the determination of the stand of plants (10). Furthermore, the present application relates to a self-propelled harvesting machine (1).