基于图像处理的农作物受害程度检测方法

Detection method for damage level of crop based on image processing

针对农作物叶片受害程度等级大小的问题,研究了一种图像处理的方法,以测量出农作物叶片受害程度的比例大小.首先,通过对样本叶片进行灰度化处理,并对图像中复杂无关的背景进行剔除以获取目标区域;在此基础上,利用最大类间方差法对图像进行分割,并统计分割后二值化图像病斑区与非病斑区像素个数,从而得出农作物受害程度比例以及相应的误差分析;最后,利用多个实例验证了该方法的有效性与可行性.

Aiming at the problem of the damage level of crop,an image processing technology is used to measure the proportion of damage level in crop leaf diseases.First of all,by using image graying method to process the leaf samples,the complex background can be removed to obtain the independent regions.Based on this,by using the maximum variance between two classes analysis algorithm,the pixels from lesion area and non-lesion area can be counted in binary image.Then,the proportion of the damage level of crop is obtained for error analysis.At last,the results of multiple experiments verify the effectiveness and feasibility of the proposed method.