Designing a clustered database for identification of leaves
Komal Asrani, R. N. Jain · 2013
Plants have an important role in our life, but due to environmental changes, many species of plants are facing extinction. It is very important to treasure this great wealth by maintaining complete details of all types of plant which will help in understanding the aspect of their survival. A leaf plays an important role in the identification of plant. Edge details of a leaf are detected using edge detection algorithms and can be stored in the form of feature vectors. Hence to improve the effectiveness of identification of leaf, we are hereby proposing to generate Clustered Database. This paper introduces an approach of Clustering based on Eccentricity. For clustering database, K-means algorithm is used. Experimental results show the effectiveness of forming clusters by calculating entropy and purity. From experimental results, it was found that the proposed approach of clustering is quite effective and would enhance the retrieval efficiency.