基于深度学习的害虫识别系统设计与实现
Design and implementation of pest identification system based on deep learning
周文俊 1王国印1
作者信息
- 1. 湖北工程学院数学与统计学院,孝感 432000
- 折叠
摘要
设计基于深度学习的害虫识别系统,完成了系统界面设计与实现,后端系统设计与实现,识别算法实现等基本功能.该系统在架构上使用前后端分离的设计模式,其中主要使用HTML、Flask、PyTorch等组件,详细地使用HTML实现前端服务,使用Python+Flask提供后端服务,使用PyTorch实现识别算法.测试结果表明:该系统基本能够完成害虫识别的需求,同时在自建的害虫识别数据集上,系统的识别准确率能达到90.052%.
Abstract
Designing a deep learning based pest identification system,and completes basic functions such as system interface design and implementation,back-end system design and implementation,recognition algorithm implementation and other basic functions.This system uses the design pattern of separation of front and back ends in the system architecture,which mainly uses HTML,Flask,PyTorch and other components.In detail,HTML is used to implement front-end services,Python+Flask is used to provide back-end services,and PyTorch is used to implement the recognition algorithm.The test results show that this system can basically meet the needs of pest identification.At the same time,on the self-built pest identification data set,the system's identifi-cation accuracy can reach 90.052%.
关键词
深度学习/前后端分离/害虫识别系统Key words
deep learning/front-end and back-end separation/pest identification system引用本文复制引用
出版年
2024
NETL
NSTL
万方数据