Automated Phytotoxicity Assessment Using Pixelwise Image Classification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for assessing phytotoxicity in agricultural crops are time-consuming and produce inconsistent results due to manual visual assessments by experts, lacking efficiency and accuracy.
Innovation Solution
An assessment system utilizing machine learning models to analyze color images of plants, classify pixels into healthy, chlorotic, necrotic, and bleached classes, and compute ratings for chlorosis, necrosis, stunting, and deformations, with photogrammetric and lidar data for stunting assessments, and image regression models for general phytotoxicity ratings.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If manual visual assessment by expert scientists is used, then phytotoxicity ratings can be provided, but the process is time-consuming and produces inconsistent results
Solution Approach 1:
The patent replaces manual visual assessment by expert scientists with an automated image analysis system. The system uses machine learning models trained on image data to automatically detect and rate phytotoxicity symptoms, eliminating the need for human experts to visually inspect plants in the field. This substitution of mechanical/manual system with automated system directly addresses both the time consumption and inconsistency issues of manual assessment.
Solution Approach 2:
The patent creates a digital copy of the plant assessment process through machine learning models that are trained on labeled image data. The models learn to replicate the assessment decisions that would be made by expert scientists, but do so automatically and consistently. This copying of the expert assessment logic into an automated system allows for rapid, consistent phytotoxicity rating without requiring repeated manual field visits.
2Measurement precision
If manual visual assessment by multiple scientists is used, then phytotoxicity can be evaluated, but results are inconsistent across different scientists
Solution Approach 1:
The patent ensures homogeneity in assessment results by using a single automated machine learning model that applies consistent criteria to all plant images. Unlike multiple human scientists who may have different interpretations and rating standards, the AI model provides uniform, objective assessment across all samples. The model's decision boundaries and classification criteria remain constant, ensuring that phytotoxicity measurements are precise and consistent regardless of which 'assessor' processes the data.
3Productivity
If automated image analysis is used, then assessment speed increases, but the system complexity increases
Solution Approach 1:
The patent segments the complex phytotoxicity assessment task into distinct, manageable components handled by specialized machine learning models. Instead of using a single monolithic complex system, the assessment is divided into separate detection tasks (e.g., chlorosis detection, necrosis detection, stunting detection) each handled by dedicated models. This segmentation reduces the complexity of individual components while maintaining high productivity through automated processing of multiple symptoms simultaneously.
Data Source
AI summary
An assessment system receives color images of plants and produces rating values of chlorosis, necrosis, bleaching, stunting, deformations, and/or an overall general assessment of the plants. Phytotoxicity damage can be identified via the color images using pixelwise classifications from imagery of crop canopy into healthy, chlorotic, necrotic, and bleached classes.