Autonomous Crop Weeding with Image-Based Blade Offset Correction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current agricultural technologies lack an efficient and autonomous method for weeding crops in fields, often requiring manual labor and leading to inefficiencies and damage to desired crops.
Innovation Solution
An autonomous machine equipped with a light module, cameras, and a weeding module that navigates along crop rows, detects target plants, calculates precise opening and closing locations for the weeding module, and actuates the blades to selectively remove weeds while avoiding desired crops.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If manual weeding is used, then weeds can be removed, but labor efficiency is low and crops may be damaged
Solution Approach 1:
The autonomous vehicle performs weeding operations independently without human intervention. The system detects crops and weeds using imaging sensors, navigates autonomously along crop rows, and activates weeding tools automatically based on real-time plant identification, enabling the system to service itself through automated decision-making and execution
Solution Approach 2:
Manual mechanical weeding operations are replaced with an automated system combining imaging sensors for detection, computer vision algorithms for plant identification, autonomous navigation for movement control, and automated weeding tools for weed removal, substituting human labor with an integrated electromechanical system
2Productivity
If conventional weeding methods are used, then weeds are removed, but desired crops may be damaged
Solution Approach 1:
The system applies different treatments to different locations by identifying individual plants as either crops or weeds. The imaging sensors capture detailed images of each plant, the processing system determines its type locally, and the weeding tools are activated only for weed locations while avoiding crop locations, enabling selective weeding at specific local positions
Solution Approach 2:
An imaging sensor system acts as an intermediary between the autonomous vehicle and the plants. The sensors capture images of crops and weeds, the processing system analyzes these images to distinguish plant types, and this intermediate detection layer enables precise differentiation before weeding action is taken, preventing direct contact with crops
3Measurement precision
If autonomous detection is implemented, then precise weed identification is achieved, but system complexity increases
Solution Approach 1:
The autonomous vehicle integrates multiple functions into a single platform: imaging sensors serve both navigation and weed detection purposes, the processing system handles both crop identification and weeding decisions, and the same vehicle platform performs both transportation and weeding operations, reducing overall system complexity through functional consolidation
Solution Approach 2:
The system uses imaging sensors to create optical copies (images) of the plants, which are then processed digitally to identify crops and weeds. This copying approach allows non-contact detection and analysis, achieving high measurement precision without physical interaction that would complicate the mechanical system
4Reliability
If real-time weeding is performed, then weed control is improved, but processing time increases
Solution Approach 1:
The system continuously captures images of crops and weeds as the autonomous vehicle moves through the field, performing detection and identification in advance before reaching the weeding position. This preliminary detection allows the control system to prepare weeding commands ahead of time, reducing actual processing delay when weeds need to be removed
Solution Approach 2:
The imaging sensors continuously capture plant images throughout the vehicle's movement, and the processing system continuously analyzes these images in real-time. This continuous detection and processing ensures no weeds are missed while maintaining efficient operation, as the system never stops its useful action of detecting and preparing to weed
Data Source
AI summary
A method for weeding crops includes, at an autonomous machine: recording an image at the front of the autonomous machine; detecting a target plant and calculating an opening location for the target plant longitudinally offset and laterally aligned with the location of the first target plant; driving a weeding module to laterally align with the opening location; tracking the opening location relative to a longitudinal reference position of the weeding module; when the weeding module longitudinally aligns with the first opening location, actuating the blades of the first weeding module to an open position; recording an image proximal to the weeding module; and in response to detecting the blades of the weeding module in the open position: calculating an offset between the opening location and a reference position of the weeding module, based on the image; and updating successive opening locations calculated by the autonomous machine based on the offset.