Laser Crop Row Identification for Stable Field Navigation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing crop row identification methods using camera images struggle with lighting conditions and seasonal changes, making it difficult to stably and accurately identify crop rows.

Innovation Solution

A work vehicle equipped with a distance sensor using laser light to acquire distance information, generating an overhead image, and a crop row identification program that creates virtual points and straight lines to identify crop rows, reducing the influence of non-crop objects and environmental factors.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Extent of automation

If image recognition technology is used to identify crop rows, then automation is improved, but measurement precision deteriorates under variable lighting conditions

Engineering Contradiction:
ImproveautomationVSAvoidmeasurement precision
Core Design Contradiction:
Extent of automationVSMeasurement precision

Solution Approach 1:

The patent applies dynamics by making the image processing adaptive to varying lighting conditions. The system dynamically adjusts its row identification algorithm based on real-time analysis of lighting patterns, shadow directions, and intensity gradients in the captured images. This allows the automated system to maintain high measurement precision across different times of day and weather conditions, resolving the contradiction between automation and precision.

Inventive Principle:
Principle #15Dynamics

2Measurement precision

If multiple cameras are used to capture images from different positions, then measurement precision is improved, but device complexity increases

Engineering Contradiction:
Improvemeasurement precisionVSAvoiddevice complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent merges multiple image capture functions into a single camera system by utilizing different focal lengths and capture modes. Instead of requiring multiple physical cameras, the system uses one camera that can capture both wide-angle overhead views and telephoto close-up views, thereby achieving multi-perspective measurement precision while avoiding the complexity of coordinating multiple camera systems.

Inventive Principle:
Principle #5Merging (Combining)

Solution Approach 2:

The camera system is designed with multi-functionality, serving multiple purposes: capturing overhead row orientation, detecting individual plant positions, identifying row boundaries, and analyzing lighting conditions. This universal camera replaces what would traditionally require multiple specialized sensors, reducing device complexity while maintaining comprehensive measurement precision.

Inventive Principle:
Principle #6Universality (Multi-functionality)

3Loss of substance

If fertilization is performed based on visual identification, then loss of substance is reduced, but reliability deteriorates due to inaccurate row identification

Engineering Contradiction:
Improveloss of substanceVSAvoidreliability
Core Design Contradiction:
Loss of substanceVSReliability

Solution Approach 1:

The patent implements feedback by continuously monitoring identified row positions and comparing them against expected patterns. The system uses the captured images to verify row identification accuracy before triggering fertilization commands, and adjusts its identification algorithm based on feedback from subsequent image analyses. This feedback loop ensures high reliability in row identification, preventing misplaced fertilization that would cause substance loss.

Inventive Principle:
Principle #23Feedback

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

Enables stable and accurate identification of crop rows regardless of lighting conditions and seasonal changes, improving the reliability and precision of crop row detection.

Implementation Method 1

a distance sensor configured to irradiate a plurality of objects present in the field with laser light and receive reflected light from the objects

Methodology Applied
Scientific EffectLaser: Laser

Implementation Method 2

receive reflected light from the objects

Methodology Applied
Scientific EffectLight reflection: Reflection

Data Source

PatentEP3766322B1Work vehicle and crop row identification program
Publication Date: 2023.05.03 YANMAR POWER TECH CO LTD
  • EP3766322B1 patent drawingFigure 1
  • EP3766322B1 patent drawingFigure 2
  • EP3766322B1 patent drawingFigure 3

AI summary

A work vehicle includes a distance sensor configured to irradiate a plurality of objects present in a field with laser light and receive reflected light from the objects to thereby acquire distance information of each of the objects, an overhead image generating unit 51 configured to detect a position of each of the objects on the basis of the distance information and generate an overhead image including the detected position as a detection point, a row candidate region generating unit 52 configured to combine two or more detection points from among the detection points to thereby generate one or more row candidate regions, and a crop row identifying unit 53 configured to create a plurality of virtual points in each of the row candidate regions, estimate a straight line corresponding to the row region candidate from the plurality of virtual points, and identify a crop row from the straight line.