Identification of plant species using non-imaging hyperspectral data

Hyperspectral non-imaging data provides the spectral range from 400-2500nm which has the ability to identify each and every unique materials on the surface. The plant species identification is critical task manually and computationally. In the present paper, we have proposed plant species identification system based on non-imaging hyperspectral data and designed our own database for experiment. Also we have identified various plant species and performed support vector machine (SVM) algorithm on it for recognition. The overall accuracy 91% was achieved through SVM.