摘要
为快速诊断识别玉米叶部病害,及时采取防治措施,提高玉米产量,本文将图像处理技术和BP神经网络算法引入到玉米叶部病害识别诊断中。对田间采集的玉米叶部病害样本图像进行背景去除、灰度化处理、阈值分割、噪声去除等预处理操作,实现叶部病害图像的分割;通过提取病害图像颜色特征和形状特征的17个参量,作为BP神经网络的输入,实现了玉米叶部常见六种病害的分类识别。实验结果表明,6种玉米叶部病害的平均识别率为93.4%,取得较好的识别效果,具有一定的实用价值。
In order to quickly diagnose and identify the disease of corn leaf,timely prevention and control measures are taken to increase corn yield.This paper introduces image processing technology and BP neural network algorithm into the identification and diagnosis of corn leaf disease.The pre-processing operations such as background removal,grayscale processing,threshold segmentation and noise removal were performed on the image of the corn leaf disease samples collected in the field to realize the segmentation of the leaf disease image;17 parameters were extracted by extracting the color features and shape features of the disease image.As the input of BP neural network,the classification and identification of six common diseases in corn leaves are realized.The experimental results show that the average recognition rate of 6 kinds of corn leaf diseases is 93.4%,which has a good recognition effect and has certain practical value.
作者
张开兴
吕高龙
贾浩
赵秀艳
刘贤喜
Zhang Kaixing;Lv Gaolong;Jia Hao;Zhao Xiuyan;Liu Xianxi(College of Mechanical and Electrical Engineering,Shandong Agricultural University,Tai’an,271018,China;College of Information Science and Engineering,Shandong Agricultural University,Tai’an,271018,China)
出处
《中国农机化学报》
北大核心
2019年第8期122-126,共5页
Journal of Chinese Agricultural Mechanization
基金
山东省农业重大应用技术创新项目
山东省“双一流”奖补资金项目(SYL2017XTTD14)
关键词
玉米
病害识别
图像处理
BP神经网络
corn
disease identification
image processing
BP neural network
