Laser Crop Row Identification for Stable Field Navigation
Find Innovative SolutionsGenerate 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
Engineering 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
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.
2Measurement precision
If multiple cameras are used to capture images from different positions, then measurement precision is improved, but device complexity increases
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.
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.
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
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.
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
Implementation Method 2
receive reflected light from the objects
Data Source
Figure 1
Figure 2
Figure 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.