Font Size: a A A

Development Of Flower Recognition System Based On Deep Learning

Posted on:2019-12-06Degree:MasterType:Thesis
Country:ChinaCandidate:Y Y LiFull Text:PDF
GTID:2393330572462610Subject:Agriculture
Abstract/Summary:PDF Full Text Request
Plants are a form of life that can be seen everywhere in our daily life.It provides us with the oxygen necessary for life and is closely related to our lives.As a kind of plant,flower,in addition to providing oxygen,has a very strong ornamental value.As an agricultural college with a history of 100 years,our school has a few species of plants on the campus.Flowers as a major highlight of these plants have also attracted widespread attention.This makes the recognition of flowers even more important.In recent years,almost all intelligent scientific researchers have noticed a mysterious technical term-Deep Learming.This term and the cutting-edge technologies it represents are selected by the famous "MIT Tech Review" as the world's top ten breakthrough technologies in 2013.Before this,many information technology giants,including Google,Microsoft,Facebook and other companies,have invested unprecedented attention and strategic resources in this technology,and then high-profile announced the deployment of smart applications.Industry and academia are also sparing no effort to research and explore this technology.This article starts with the common flowers on the campus of Shanxi Agricultural University,and explores flower recognition methods based on deep learning technology.Based on the structure of the convolutional neural network,this method constructs a deep neural network structure that simulates the principle of visual perception,and extracts the essential characteristics of flowers by unsupervised learning to classify flowers.The algorithm proposed in this paper is mainly used to identify common flowers on the campus of Shanxi Agricultural University.The recognition rate can reach 95.6%,and obvious recognition results are achieved.
Keywords/Search Tags:Flower classification, deep learning, Convolution neural network, feature extraction
PDF Full Text Request
Related items