基于决策树与卷积神经网络的害虫识别算法

Pest Detection Algorithm Based on Decision Tree and Convolution Neural Network

为了推动农业自动化的发展,以常见农业害虫为例,设计了一种基于决策树与卷积神经网络的害虫识别算法.其结合了无人机、ZigBee通信协议、数据处理服务器、机载报警器、害虫消除执行机构、Boosting算法、OpenCV开源库和TensorFlow开源库.经验证,本算法具有更高的害虫识别检出能力.

To promote the development of agricultural automation,this study,taking the common agricultural pests as samples,designed a pest detection algorithm based on decision tree and convolution neural network.This algorithm which combines UAV,ZigBee communication protocol,data processing server,airborne alarm and pest elimination executing mechanism,Boosting algorithm,OpenCV open source library,and TensorFlow,has been proved to have higher ability of pest identification and detection.