Field Anomaly Mapping Using ML Image Classification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for detecting lodging, bare soil, and weeds in agricultural fields are inefficient and costly, making them difficult to scale to commercial operations, and lack a systematic approach for automated detection and classification.
Innovation Solution
A machine learning approach using digital images from UAVs and ground vehicles to detect and classify lodging, bare soil, and weeds, employing image calibration, stitching, grid generation, and classification, integrated with a computer system for anomaly mapping.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If expensive sensors such as LiDAR and hyperspectral sensors are used for lodging and weed detection, then measurement precision is improved, but device complexity and cost increase
Solution Approach 1:
The patent replaces expensive, complex sensors (LiDAR, hyperspectral sensors) with inexpensive, readily available consumer cameras. This substitution maintains adequate detection capability for agricultural monitoring while dramatically reducing system complexity and cost, making the solution scalable to commercial operations.
Solution Approach 2:
The patent substitutes sophisticated post-processing methods (Surface from Motion Digital Surface Model) with machine learning-based image classification. This replacement simplifies the processing pipeline while maintaining or improving detection accuracy for lodging, bare soil, and weeds.
2Measurement precision
If sophisticated post-processing such as Surface from Motion is applied, then measurement precision is improved, but productivity decreases
Solution Approach 1:
The patent replaces time-consuming sophisticated post-processing (Surface from Motion) with efficient machine learning image classification algorithms. This substitution maintains detection precision for field anomalies while dramatically increasing processing throughput, enabling scalable deployment across multiple fields.
Solution Approach 2:
The patent changes the processing approach from complex geometric modeling (Surface from Motion) to statistical pattern recognition (machine learning classification). This parameter change in the processing methodology reduces computational complexity and processing time while maintaining accuracy.
3Measurement precision
If manual detection methods are used for lodging and weeds, then measurement precision is maintained, but productivity decreases
Solution Approach 1:
The patent replaces manual detection methods with automated machine learning-based image classification. This substitution maintains detection accuracy for lodging, bare soil, and weeds while dramatically increasing processing speed and throughput, enabling commercial-scale operations.
Solution Approach 2:
The patent implements an automated system that performs detection and classification without human intervention. The machine learning model automatically processes images, identifies anomalies, and generates results, eliminating the need for manual field assessment while maintaining accuracy.
4Ease of operation
If existing imaging approaches are applied, then ease of operation is improved, but measurement precision worsens due to lack of systematic approach
Solution Approach 1:
The patent implements a systematic multi-step processing pipeline: image calibration, stitching, grid generation, and classification. This segmentation of the detection process into distinct, automated stages improves both operational simplicity and measurement precision by ensuring consistent application of each processing step.
Solution Approach 2:
The patent creates a universal machine learning model that simultaneously detects and classifies multiple types of field anomalies (lodging, bare soil, weeds) in a single processing run. This multi-functional approach maintains ease of operation while improving measurement precision through systematic comprehensive analysis.
Data Source
AI summary
A computer-implemented method for generating an improved map of field anomalies using digital images and machine learning models is disclosed. In an embodiment, a method comprises: obtaining a shapefile that defines boundaries of an agricultural plot and boundaries of the field containing the plot; obtaining a plurality of plot images within the field from one or more image capturing devices that are located within the boundaries of the field; calibrating and pre-processing the plurality of plot images to create a plot map of the agricultural plot at a plot level; based on the plot map of the agricultural plot, generating a plot grid; based on the plot grid and the plot map, generating a plurality of plot tiles; based on the plurality of plot tiles, generating, using a first machine learning model and a plurality of first image classifiers corresponding to one or more first anomalies, a set of classified plot images that depicts at least one anomaly; based on the set of classified plot images, generating a plot anomaly map for the agricultural plot; transmitting the plot anomaly map to one or more controllers that control one or more agricultural machines or database systems to perform agricultural functions on the agricultural plot.