Digital Image Processing based proposed approach to Identify Different Bamboo Species

摘要

Assam is one of the biodiversity hotspot in India. Most of the village people of Assam somehow depended on bamboo trees found in this area. Identification of bamboo species is very crucial because every species has its unique use. Manual bamboo identification is time consuming and an erratic practice. In general, species are recognized from their parts like stem, internodes distance, leaves, culm-sheaths, flowers etc. This paper discusses a new approach for bamboo species recognition using digital image processing technique. The proposed work uses characteristics of stem and leaf for classification. Also an algorithm is developed to make our approach easier and eliminate the overlapping characteristics of tree parts like stem and leaf. A database is created by capturing the photographs of bamboo parts with a high resolution camera. The digital image processing techniques are used here for classification. Till today there is no specific software for bamboo species identification in this region of India.