Classification of Plant Leaves by Near-infrared Spectroscopy Using ANN and Wavelet

Near-infrared (NIR) spectroscopy is a kind of non-destructive technology, which has many advantages. Infrared spectrum is highly characteristic and can be used to analyze and identify. However, researches and applications of NIR spectroscopy in plant classification and growth monitoring are still rare. This paper conducted NIR spectroscopy experiments on Cinnamomum camphora and Aceraceaedie leaves to obtain their spectrum curves, and wavelet is used to perform pre-processing of obtained data, with efficient compression of spectrum data, then the classification model of the leaves is established using artificial neural network (ANN), with good classification results, which showed that NIR spectroscopy can be used in classification of different plants.