摘要
为探讨利用图像处理检测烤烟鲜烟叶含水率的方法,采用1 600万像素的数码相机采集处于旺长期的烤烟叶片的图像信息,用烘干法测量叶片的含水率;通过Matlab软件图像处理中的灰度直方图提取叶片图像的灰度均值和熵值特征值,并测量了烤烟叶片的最大长度、最大宽度、伸缩率、湿重等参数。结果表明,利用提取的15组烟叶叶片数据进行多项式曲线拟合,建立了图像灰度均值与烤烟叶片含水率的线性拟合预测模型,模型的相关系数R^2达到了0.796 5。由此表明,利用烟叶图像的灰度均值可以对其含水率进行诊断。
In order to discuss the method of evaluating tobacco leaf water based on image processing, the digital camera with 16 million pixels was used to capture the images of tobacco leaves after being picked from the plants during the vigorous growing stage ,and then the moisture content of the leaves was detected by using drying method.The gray mean and entropy of leaf images were calculated by Matlab, and the max-length, max-width, expansion rate and wet weight of tobacco leaves were measured. The data of 15 groups of tobacco leaves were extracted by polynomial curve fitting, linear fitting prediction model were set up between the leaf water content and the gray mean of leaf image.The R2 of the model was up to 0.796 5. It was concluded that of the ~rav average of leaf ima,,es could be used to predict the moisture content in tobacco leaves.
出处
《现代农业科技》
2017年第9期3-5,共3页
Modern Agricultural Science and Technology
基金
2014年贵州省科学技术厅
兴义民族师范学院联合基金项目(黔科合LH字[2014]7412号)
贵州省高等学校大学生创新创业训练计划项目(201410666010)
关键词
烤烟鲜烟叶
图像处理
叶片含水率
灰度均值
预测模型
flue-cured tobacco fresh leaf
image processing
leaf moisture content
gray mean
prediction model
