干旱地区农业研究2013，Vol.31Issue(1)：95-100,6.
基于图像处理的玉米叶片含水率诊断方法研究
Research on methods of diagnosing crop water-deficiency based on image processing
摘要
Abstract
Digital image processing techniques were used to evaluate crop water stress, by cultivating 90 plants of com with different irrigation amounts in a greenhouse. The Canon KUS110 digital camera with 12.1 million pixels was used to capture the images of com leaves after being picked from the plants during the heading stage, and then the moisture content of the leaves was detected by using drying method. The eigenvalues of mean, kurtosis, variance, skew degree , energy and entropy were calculated by using grey histogram of leaf images. The data extracted from the leaves of 20 samples were used to set up the linear regression model showing the relationship between the mean and the leaf moisture, and the other 20 samples were used to verify the model. The standard deviation of the validation results was 0.021. It was concluded that of the eigenvalue of mean of leaf images could be used to predict the moisture content in com leaves.关键词
玉米/叶片含水率/诊断/图像处理/颜色特征提取/线性回归Key words
corn/ leaf water deficiency/ diagnosing/ image processing/ colour feature extraction/ linear regression分类
农业科技引用本文复制引用
徐腾飞,韩文霆,孙瑜..基于图像处理的玉米叶片含水率诊断方法研究[J].干旱地区农业研究,2013,31(1):95-100,6.基金项目
"十二五"国家科技支撑计划课题(2011BAD29B08) （2011BAD29B08）
国家教育部、外专局111项目(B12007) （B12007）
