Automated Plant Counting via Multispectral Imaging
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for counting plants in test plots are prone to errors and fatigue, leading to unreliable data due to manual counting by teams of people, which can significantly impact yield estimates and decision-making in plant testing and development programs.
Innovation Solution
An automated plant counting system that captures color and NIR image data, calculates pixel ratios, generates false color images, and identifies specific plant types based on distinguishing characteristics, using a self-propelled mobile platform equipped with multispectral cameras and a data processing system to accurately count plants in real-time.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If manual counting by teams of people is used, then the process is simple and requires basic equipment, but the data reliability deteriorates due to counting errors and fatigue
Solution Approach 1:
The patent replaces the manual mechanical counting process with an automated optical system using cameras and image processing algorithms. The system captures images of plants in the field and uses computer vision to automatically count and identify target plants, eliminating human fatigue and counting errors while maintaining operational simplicity through automated processing.
Solution Approach 2:
The system enables self-service by allowing the automated plant counting system to perform the counting task independently without requiring human operators to manually count each plant. The image processing algorithm automatically processes captured images, identifies target plants, and generates count data, making the system self-sufficient in performing the measurement function.
2Measurement precision
If automated plant counting system with multispectral cameras is used, then measurement precision improves through accurate plant identification, but device complexity increases due to multiple cameras and processing systems
Solution Approach 1:
The patent segments the plant counting task into distinct functional components: color camera for general plant detection, NIR camera for specific plant type identification, and processing algorithms for data fusion and analysis. This segmentation allows each component to specialize in a specific aspect of plant identification, improving overall measurement precision while organizing system complexity into manageable modules.
Solution Approach 2:
The system achieves multi-functionality by using a combined color and NIR imaging system that can perform both general plant detection and specific plant type identification simultaneously. The same hardware platform executes multiple functions (capturing visible light images, capturing NIR images, processing both datasets, and generating differentiated plant counts), reducing the need for separate specialized systems.
3Productivity
If hand counting methods are used, then equipment requirements are minimal, but productivity deteriorates due to time-consuming manual processes
Solution Approach 1:
The patent implements continuous useful action by capturing images continuously as the system moves through the field, rather than stopping to manually count plants. The automated system maintains continuous image capture and processing operations, allowing simultaneous data collection across multiple plots and significantly increasing counting speed while the modular architecture manages the associated system complexity.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
The system provides precise and reliable plant counts, reducing human error and fatigue, and enabling accurate yield estimates by automatically identifying and counting specific plant types within a field, thus enhancing the precision and reliability of plant testing and development programs.
Implementation Method 1
capturing color NIR image data of an entire field having plants of a selected type growing therein utilizing an automated plant counting system
Implementation Method 2
captures color and NIR image data, calculates pixel ratios, generates false color images
Data Source
AI summary
A method for recognizing individual plants of a selected type growing in a field, wherein the method comprises capturing color NIR image data of an entire field having plants of a selected type growing therein utilizing an automated plant counting system and calculating a ratio value between each pixel of the color image data and the corresponding pixel of the NIR image data utilizing a plant recognition algorithm executed via a data processing system of the plant counting system. The method additionally comprises generating, via execution of the plant recognition algorithm, a false color image of the field based on the calculated ratios for each pixel, and identifying, via execution of the plant recognition algorithm, all plants of the selected type in the false color image based on a plant distinguishing characteristic uniquely rendered for each individual plant of the selected type in the false color image.