摘要
利用高光谱图像技术研究了诊断温室黄瓜病害的方法,以提高诊断的准确性和效率。试验以黄瓜霜霉病、白粉病为研究对象,利用高光谱图像采集系统获取黄瓜病叶的高光谱图像数据,在450～900nm范围内的高光谱图像数据中,选出特征波长下的图像;然后,对该图像进行去除噪声的滤波处理,并提取黄瓜病叶的色度矩纹理特征向量;最后采用支持向量机分类方法对黄瓜病害进行诊断。研究结果表明,采用高光谱图像新技术与线性核函数对黄瓜霜霉病、白粉病的正确诊断率达100%,采用高光谱图像技术可以实现对温室黄瓜病害进行快速、精确的分类诊断。
In order to improve the accuracy and efficiency of diagnosis for cucumber disease in greenhouse,the hyperspectral imaging technology was proposed.The cucumber disease of downy mildew and powdery mildew were adopted as the experimental object.The cucumber disease hyperspectral images data were acquired by a hyperspectral imaging system.The feature wavelength images were selected via analysis between 450 nm and 900 nm wavelength hyperspectral imaging.Then these images were filtered to eliminate noise.And texture features were extracted based on chromaticity monuments of cucumber disease in leaf.Finally,the SVM was used to diagnose cucumber disease.The experimental results show that the accuracy of diagnosing cucumber disease of downy mildew and powdery mildew with hyperspectral imaging technology was 100%.So,the hyperspectral imaging technology can be used to diagnose cucumber disease,which has better features of accuracy and efficiency.
出处
《农业工程学报》
EI
CAS
CSCD
北大核心
2010年第5期202-206,I0006,共6页
Transactions of the Chinese Society of Agricultural Engineering
基金
辽宁省自然科学基金(20052125)
关键词
病害
诊断
支持向量机
高光谱图像
色度矩
黄瓜
disease
diagnosis
support vector machines
hyperspectral imaging
chromaticity monuments
cucumber
