To diagnose and identify tobacco diseases quickly and effectively,and improve the diagnostic level of tobacco dis⁃eases,the adaptive median filtering technology is used to perform the noise removal of the collected images of tobacco angular leaf spot and wildfire,the fast fuzzy C⁃means clustering algorithm is adopted to segment the lesion,and then the fuzzy recogni⁃tion technology is used for the Auto⁃Recognition of the disease according to the characteristics of color,shape and texture of to⁃bacco spot disease. Test results show the model has high recognition rate for the disease samples,and can meet the actual pro⁃duction requirements.%为快速有效地诊断识别烟叶病害，提高烟草病害诊断水平，首先采用自适应中值滤波技术对采集的角斑病和野火病图像进行去噪处理，然后采用快速模糊C⁃均值聚类算法对病斑进行分割，最后提取烟叶病斑颜色、形状和纹理方面的特征，采用模糊识别技术对病害进行自动识别。实验测试表明，该模型对病害样本进行处理识别的正确率较高，能够满足生产的实际需求。
展开▼
