摘要
精准识别作物病害的前提是精确提取其特征。为此,利用计算机图像处理技术,研究了黄瓜霜霉病的特征提取过程。使用高精度的光学设备直接进行数字图像采集,再运用图像预处理方法对图像数据进行优化,去除干扰,精确定位病斑部位;分析黄瓜霜霉病的表现形式,从统计量、颜色和形状3个方面分离出特征参数值15个。这些特征值将作为识别该病害的主要依据,大大提高了病害的识别精度。
The precise of identification of crop diseases is to extract its characteristics accurately. The study used com- puter image-processing techniques, study on the characteristic extraction process of cucumber downy mildew. Directly using a high precision optical equipment for digital image acquisition, then use image preproeessing method for optimizing image data, removing interference, pinpoint lesion site. Analyzed the manifestations of cucumber downy mildew, 15 parameter values were got from statistics, color and shape. These values will serve as the main basis for identifying the disease, greatly improving the disease recognition accuracy.
出处
《农机化研究》
北大核心
2014年第2期179-182,187,共5页
Journal of Agricultural Mechanization Research
基金
吉林省世行贷款农产品质量安全项目(2011-Z20)
吉林农业大学青年启动基金项目(201126)
关键词
黄瓜病害
数字图像处理
特征提取
作物病害识别
cucumber diseases
digital Image processing
feature extraction
plant disease recognition