Exploring the Potential of EnMAP Hyperspectral Data for Crop Classification: Technique and Performance Evaluation
Hyperspectral remote sensing is one powerful component that contributes to precision agriculture. The study explores the potential of EnMAP hyperspectral data for the classification of crop types in Goroimari, Kamrup, Assam. The targeted crops namely Sali rice, Rabi maize, mustard and potato are the...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Copernicus Publications
2025-07-01
|
| Series: | ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences |
| Online Access: | https://isprs-annals.copernicus.org/articles/X-G-2025/291/2025/isprs-annals-X-G-2025-291-2025.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Hyperspectral remote sensing is one powerful component that contributes to precision agriculture. The study explores the potential of EnMAP hyperspectral data for the classification of crop types in Goroimari, Kamrup, Assam. The targeted crops namely Sali rice, Rabi maize, mustard and potato are the major Kharif crops grown in the area. A standard protocol for spectra and metadata collection was prepared to be followed while collecting the field spectra. A spectral library of the crops is prepared by collecting spectra in the field using SVC HR-1024 spectroradiometer. After preparing the spectral library as a reference for identification, EnMAP hyperspectral data of the study area, acquired at a similar time with the field data collection is obtained. Utilizing the high spectral resolution of EnMAP with ground field spectra of targeted crops, a combination of end member extraction, endmember spectral matching and Spectral Angle Mapper (SAM) algorithm was implemented to effectively differentiate between various crop types. Three additional classes were identified i.e., crop residue, fallow and sandbar class. The identified endmember classes were taken as the reference spectra to classify the area using SAM. An overall classification accuracy of 88.43 % was achieved. The study demonstrates the potential of EnMAP hyperspectral data in discriminating crop type. |
|---|---|
| ISSN: | 2194-9042 2194-9050 |