FEATURE EXTRACTION AND DENSE LAYER CLASSIFICATION USING CONVOLUTIONAL NEURAL NETWORK

Abstract

Butterflies are the important creature of the ecosystem.The relation involving plants and butterflies are playing a vital role in maintaining many natural processes of the ecosystem.They are sometimes called flying flowers of flowers with variant colors.Butterflies strongly help in the process of pollination.Due to many human activities such as the use of pesticides in the plants, destruction of habitats, unawareness of the importance of butterflies to the ecosystem decline the population of this beautiful species.This research follows an investigation on butterfly classification using image processing with Deep Learning Methods.Butterflies can be classified by their external morphological characters(structure) and genital characters.There are many types of butterflies and the research on classification of butterfly species is of great significance in practical work such as environmental protection and control of agriculture and forest pests.They are, however, difficult to recognize due to several varieties and patterns and this means there is a need to group based on type for easier recognition.For that, some of the Machine Learning methodologies like Traditional machine learning, Deep learning and Transfer learning are used by training and testing on a butterfly dataset and determine the optimal solution.This application can detect the category of a butterfly by either capturing a real time picture of a butterfly or choosing one picture from gallery. I.