| 基于卷积神经网络的病虫害可视化监测系统设计 |
| |
| 引用本文： | 沈凯文,李浩伟,韩 进,房若民.基于卷积神经网络的病虫害可视化监测系统设计[J].教育技术导刊,2009,19(9):122-126. |
| |
| 作者姓名： | 沈凯文 李浩伟 韩 进 房若民 |
| |
| 作者单位： | 山东科技大学 计算机科学与工程学院，山东 青岛 266000 |
| |
| 基金项目： | 教育部协同育人项目（201901055015）|山东省教育厅研究生教育质量提升计划项目（SDYAL17034）|山东科技大学优秀教学团队支持计划项目（JXTD20170503） |
| |
| 摘 要： | 病虫害监测是提高农业生产效率和产量的有效措施。提出一个基于卷积神经网络的农作物病虫害智能监测系统。该系统以移动端为媒介实现监测众包化，基于 GIS 平台对相关区域病虫害发展态势进行数据可视化分析，显示病虫害位置与规模，代替人工识别常见农作物病虫害。实验证明，该系统对农作物病虫害常见的 10 个物种与 27 种病虫种类识别率达 95%，病虫害严重程度识别率达 85%。移动端采集地理定位信息误差5~10m，系统还可对变化趋势进行可视化展示，为农作物病虫害识别、防治及决策提供有效方案。
|
| 关 键 词： | 病虫害识别 卷积神经网络 智慧农业 众包 |
| 收稿时间： | 2020-01-11 |
The Design of Visualized Disease and Pest Monitoring System Based on CNN |
| |
| SHEN Kai-wen,LI Hao-wei,HAN Jin,FANG Ruo-min.The Design of Visualized Disease and Pest Monitoring System Based on CNN[J].Introduction of Educational Technology,2009,19(9):122-126. |
| |
| Authors: | SHEN Kai-wen LI Hao-wei HAN Jin FANG Ruo-min |
| |
| Institution: | School of Computer Science and Engineering，Shandong University of Science and Technology，Qingdao 266000，China |
| |
| Abstract: | Pest and disease monitoring is an effective measure to improve agricultural yield and efficiency. This paper proposes a con? volutional neural network which uses mobile terminals as media to monitor crowdsourcing，and the GIS platform can visualize and analyze data on the development of pests and diseases in relevant regions，showing the location and scale of pests and diseases. The manual identification of common crop diseases and insect pests is thus avoided. The experimental results prove that the system recognizes 10 common species of crop pests and diseases，27 species of pests and diseases，the recognition rate is 95%，and the severity recognition rate of pests and diseases is 85%|the error of geolocation information collected on the mobile terminal is within 5~10m. The system can provide visualization of the trends of change，which provides effective solutions for the identification，control and decision-making of crop diseases and insect pests. |
| |
| Keywords: | pest and disease recognition convolutional neural network smart agriculture crowdsourcing |
|
| 点击此处可从《教育技术导刊》浏览原始摘要信息 |
| 点击此处可从《教育技术导刊》下载免费的PDF全文 |
