Open AccessCrop Discrimination using Non-Imaging Hyperspectral Data
Author(s) -
Pooja Vinod Janse,
Ratnadeep R. Deshmukh
Publication year - 2021
Publication title -
international journal of engineering and advanced technology
Language(s) - English
Resource type - Journals
ISSN - 2249-8958
DOI - 10.35940/ijeat.e2802.0610521
Subject(s) - hyperspectral imaging , spectroradiometer , reflectivity , support vector machine , remote sensing , vegetation (pathology) , crop , similarity (geometry) , spectral bands , artificial intelligence , mathematics , pattern recognition (psychology) , environmental science , computer science , geography , agronomy , image (mathematics) , optics , biology , medicine , physics , pathology
Crop type discrimination is still very challengingtask for researchers using non-imaging hyperspectral data. It isbecause of spectral reflectance similarity between crops. In thisresearch work we have discriminated between four crops wheat,jowar, bajara and maize. We have tried to overcome the problemswhich have been faced my researchers. Initially by visualanalysis we have selected 22 reflectance band which shows theabsorption property of particular molecules and classificationtechnique is applied, but it has given us very poor result ofclassification. We observed only 24% classification accuracy. Sowe considered nine vegetation indices along with spectral bandsand achieved better classification accuracy. ASD FieldSpec 4Spectroradiometer device is used for capturing spectralreflectance data. We calculated nine different vegetation indicesand some selective reflectance bands are used for cropclassification. We have used Support Vector Machine (SVM) forclassification.