Dynamic Plant Imaging Lighting for Shadowed Crop Canopies
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing machine learning models for agricultural tasks are hindered by uneven lighting in captured image data, which affects accuracy, particularly as plants grow and block overhead light, leading to insufficient illumination of lower portions.
Innovation Solution
A reinforcement learning model is used to dynamically adjust lighting on agricultural robots or modular sensor packages, adjusting lights based on image data analysis to improve illumination evenness, using a policy network to identify areas of uneven illumination and generate lighting adjustments.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Illumination intensity
If overhead light is used to illuminate crops, then the lighting setup is simple and energy-efficient, but the lower portions of growing plants become insufficiently illuminated as plants grow taller
Solution Approach 1:
The lighting system is segmented into multiple independent light sources positioned at different locations (overhead, side, and bottom lights) that can be individually controlled. This allows targeted illumination of specific plant zones that would otherwise be in shadow, resolving the contradiction by dividing the lighting function into multiple segments rather than using a single overhead source.
Solution Approach 2:
The patent transitions from traditional single-dimension overhead lighting to multi-dimensional lighting by adding light sources from multiple spatial dimensions (top, sides, and bottom). This dimensional expansion ensures all plant surfaces receive adequate illumination regardless of growth stage, solving the limitation of overhead-only lighting.
2Measurement precision
If static lighting configuration is used, then the system is simple to operate, but image quality deteriorates as plants grow and block light
Solution Approach 1:
The lighting system transitions from static to dynamic operation, with light sources that automatically adjust their intensity and activation based on real-time image analysis. The system continuously adapts to changing plant growth and light blocking conditions, maintaining optimal image quality without manual intervention.
Solution Approach 2:
The system implements closed-loop feedback by analyzing captured images to identify shadowed or unevenly illuminated regions, then using this information to automatically adjust the lighting configuration. This feedback mechanism ensures continuous optimization of image quality while eliminating manual lighting adjustments.
3Illumination intensity
If multiple lights are added to illuminate all plant portions, then illumination evenness improves, but energy consumption increases
Solution Approach 1:
Instead of uniformly illuminating the entire scene, the system applies lighting locally to specific regions that require it. The multi-light configuration targets shadowed areas and unevenly illuminated portions selectively, providing optimal illumination where needed while minimizing energy waste in already-well-lit areas.
Solution Approach 2:
The system dynamically changes lighting parameters (intensity, activation state) based on real-time conditions. By adjusting light output to match actual illumination needs detected through image analysis, the system maintains high illumination evenness while optimizing energy consumption according to environmental and plant growth conditions.
Data Source
AI summary
Various implementations include processing an instance of image data using a reinforcement learning policy model to generate illumination output, where the illumination output indicates one or more lights of an agricultural robot or modular sensor package to adjust based on uneven illumination in the instance of image data. In many implementations, the initial instance of image data is captured using one or more sensors of an agricultural robot or modular sensor package, where the initial instance of image data captures one or more crops in a portion of a plot of land. In various implementations, the agricultural robot or modular sensor package can adjust one or more lights based on the illumination output, and can capture an updated instance of image data of the given one or more crops with the updated illumination.