Cereal Grain Yield Prediction Using Spikelet Detection Models

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing trait extraction techniques for cereal grains are arduous, time-consuming, prone to human error, and subjective, making accurate crop yield prediction challenging.

Innovation Solution

Automated extraction of cereal grain traits using a spikelet detection model, which includes a neural network trained on images of cereal grain plots to identify spikelets, determine heading values, and generate predicted yields based on normalized spikelet counts and growth curves.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If automated spikelet detection model is used, then measurement precision and reliability are improved, but device complexity increases

Engineering Contradiction:
Improveheading value determination accuracyVSAvoidneural network model complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent replaces manual visual inspection and human judgment with an automated computer vision system using a neural network model. The spikelet detection model automatically identifies and counts spikelets in images, eliminating the need for human observers to manually assess heading values, thereby improving measurement precision and consistency.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The system creates a digital representation (copy) of the physical spikelets through image capture and processing. The neural network model analyzes these image copies to determine spikelet presence and count, allowing accurate heading value determination without direct human intervention in the physical assessment process.

Inventive Principle:
Principle #26Copying

2Productivity

If manual trait extraction is used, then ease of operation is maintained, but productivity and measurement precision deteriorate

Engineering Contradiction:
Improvetrait extraction speedVSAvoidautomated system operation complexity
Core Design Contradiction:
ProductivityVSEase of operation

Solution Approach 1:

The system performs self-service by automatically capturing images, detecting spikelets, counting them, and determining heading values without requiring manual operation for each measurement step. The automated pipeline processes multiple images sequentially, enabling high-throughput trait extraction while maintaining simplicity in overall system operation.

Inventive Principle:
Principle #25Self-service

3Measurement precision

If multiple images are processed sequentially, then measurement precision is improved, but loss of time increases

Engineering Contradiction:
Improvespikelet count accuracyVSAvoidprocessing time for multiple images
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The system maintains continuous useful action by processing images in an automated sequential pipeline without interruption. Each image is processed immediately after capture, with the neural network model continuously detecting and counting spikelets, thereby minimizing idle time while ensuring accurate measurements through multiple observations.

Inventive Principle:
Principle #20Continuity of useful action

Data Source

PatentEP4211665B1Determining cereal grain crop yield based on cereal grain trait value(s)
Publication Date: 2026.02.18 DEERE & CO
  • EP4211665B1 patent drawingFigure 1
  • EP4211665B1 patent drawingFigure 2
  • EP4211665B1 patent drawingFigure 3

AI summary

Techniques are disclosed that enable generating a predicted yield for a cereal grain crop based on one or more traits extracted from image(s) of the cereal grain crop. Various implementations include determining a heading trait value based on the number of identified spikelets, where the spikelets are identified by processing the image(s) of the cereal grain crop using a spikelet detection model. Additional or alternative implementations include generating a predicted cereal grain crop yield based on one or more additional or alternative trait values such as one or more heading values, one or more projected leaf area values, one or more stand spacing values, one or more wheat rust values, one or more maturity detection values, one or more intercropping phenotyping values extracted cereal grains intercropped with other crops, one or more additional or alternative trait output values, and/or combinations thereof.