Hyperspectral Crop Identification Using Multi-Band Spectral Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing precision agriculture methods using hyperspectral cameras to identify plant species exclude bands affected by chlorophyll content changes, limiting the ability to determine the growing state of trees and only determining tree species, which is not applicable to other agricultural crops like cereals and weeds.
Innovation Solution
A precision agriculture support system that uses a hyperspectral camera to measure spectral characteristics across visible and infrared bands, including mid-wavelength and long-wavelength infrared, to differentiate between desired crops and weeds by creating a database of spectral features for accurate identification.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If bands affected by chlorophyll content changes are excluded from measurement to determine tree species, then species identification accuracy is improved, but the ability to determine growing state is lost
Solution Approach 1:
The patent segments the spectral measurement into two distinct purposes: using certain bands for species identification and other bands for growing state determination. This allows the system to process spectral data for multiple objectives simultaneously without interference, resolving the contradiction between accurate species identification and growing state monitoring.
Solution Approach 2:
The patent makes the spectral measurement system universal by enabling it to perform multiple functions: species identification and growing state determination. By utilizing different spectral bands for different purposes, the system achieves multi-functionality, allowing both objectives to be accomplished with a single measurement system.
2Device complexity
If spectral measurement is limited to visible and near-infrared bands for tree species determination, then measurement system complexity is reduced, but applicability to other agricultural crops is limited
Solution Approach 1:
The patent extends the measurement system to include mid-infrared and long-infrared bands in addition to visible and near-infrared bands. This expansion makes the system universally applicable to various agricultural crops including cereals and weeds, not just trees, while maintaining manageable system complexity through integrated processing.
Solution Approach 2:
The patent adds new dimensions to the spectral measurement by incorporating mid-infrared and long-infrared bands beyond the traditional visible and near-infrared range. This dimensional expansion enables the system to detect crop-specific spectral features across different crop types, significantly enhancing versatility.
3Measurement precision
If full spectral range including mid-infrared and long-infrared bands is measured to identify crops and weeds, then identification accuracy is improved, but measurement precision for specific parameters may be reduced
Solution Approach 1:
The patent segments the full spectral range into distinct bands (visible, near-infrared, mid-infrared, long-infrared) and assigns specific bands to different identification purposes. This segmentation allows the system to process complex spectral data efficiently while maintaining high identification accuracy for both crops and weeds.
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 accurate identification of desired crops like rice plants and weeds, reducing herbicide costs and minimizing soil contamination by targeting herbicide application precisely, while also monitoring crop health to prevent yield reduction.
Implementation Method 1
spectrum in bands from visible rays to near-infrared rays of the reflected light from trees is measured by used of a hyperspectral camera
Data Source
AI summary
A precision agriculture support system is provided with a measuring device, a storage device and a plant species determining unit. The measuring device measures a first spectral characteristic of light derived from vegetation in a support target area. The storage device stores a database of spectrum according to species that shows a spectral characteristic of a desired crop. The plant species determining unit determines whether a plant included in the vegetation is the desired crop or not based on the database of spectrum according to species and a measurement result of the first spectral characteristic. The plant species determination unit further carries out distinction of agricultural crops, distinction of agricultural crops and weeds and the like. Furthermore, the precision agriculture support system identifies an area where abnormality is occurring, estimates a nature of the abnormality and carries out an early warning by providing a countermeasure against the abnormality.