Plant classification based on leaf Shape features using Neural Network

Plant classification becomes very much demanding research area for the use of taxonomist and botanist as well as in agricultural requirement. Using computerized techniques for performing these task is more effective. This research paper presents the classification of plants by using Feed forward Back Propagation Neural Network designed.This method performed in three steps first is leaf image processing to find the leaf shape patterns, leaf shape features extraction that follows classification. Leaf images of 220 different plant species are processed to find the leaf shape patterns. Leaf shape features are computed automatically from the leaf shape patterns of the leaf image samples. Neural network designed based on the leaf shape features for plant classification. This paper is organized in the sections, Introduction,Methods, Methodology and Conclusion Keywords—Leaf Image Processing,Leaf Shape,Plant Classification, Feed Forward Neural Network. I . INTRODUCTION Plants are very important part of our ecosystem. It maintains the balance of environment. The imbalance in environment is because of deforesting. Due to this imbalance we face the problem of natural disaster like draught. Plants are important source of development of human being and living of them. Plants are useful for various industries. They are useful in production of medicines they are also useful as foodstuff. Plant classification is process by which we can group together the plant species based on some physiological characteristics which are common by some criterion [4]. For plant classification trained taxonomist and botanist follows various tasks. They follow many methods such as morphological anatomy, cell biology and molecular biological approach [10]. These methods are required to follow many critical tasks. Performing these tasks is time consuming process and it requires more efforts. These processes are less efficient as it is difficult to remember one all the species in the world[3]. There are many features which classify the plant species with physiological characteristics leaf component of almost all species can be used for classifying the plant species as they are more readily available. This paper presents the performance of the neural network for classification of the plant species. Plant classification process is presented here based on the leaf shape features which are automatically extracted from the leaf image samples. This method is not required efforts to perform the critical task for classification.