摘要
针对目前马铃薯图像病虫害检测方法识别精度低、效率不高等问题。这里提出了一种主成分分析的特征加权融合自适应算法和改进支持向量机相结合的马铃薯病虫害快速检测方法。主成分分析的特征加权融合自适应算法完成特征块选择、主分量提取、加权和融合。融合后的特征采用决策树的思想,通过改进的支持向量机逐级分类。通过试验进行对比分析。结果表明,与传统的检测方法相比,该方法具有更高的检测精度和更短的执行时间。在病虫害、3类病害和10类虫害的检测准确率分别为98.45%、97.33%和98.00%,运行时间分别为4.81s、3.74s和4.65s。该检测方法为图像病虫害快速检测技术的发展提供了理论方法和依据。
In view of the problems of low recognition accuracy and low efficiency of the current potato image pest detection methods.A fast detection method of potato diseases and insect pests is proposed,which combines the feature weighted fusion adaptive algorithm of principal component analysis and the improved support vector machine.The feature weighted fusion adaptive algorithm of principal component analysis completes feature block selection,principal component extraction,weighting and fusion.The fused feature adopts the idea of decision tree,and is classified step by step through the improved support vector machine.Comparative analysis is carried out through experiments.The results show that the proposed method has higher detection accuracy and shorter execution time than traditional detection methods.The detection accuracy rates of pests,3 types of diseases and 10 types of pests are 98.45%,97.33%and 98.00%,respectively,and the running time is 4.81s,3.74s and 4.65s,respectively.The detection method provides a theoretical method and basis for the development of image pest and disease rapid detection technology.
作者
李英辉
王晓寰
赵翠俭
LI Ying-hui;WANG Xiao-huan;ZHAO Cui-jian(Shijiazhuang Vocational and Technical College,Hebei Shijiazhuang 050081,China;Yanshan University,Hebei Qi‐nhuangdao 066004,China;Shijiazhuang College,Hebei Shijiazhuang 050035,China)
出处
《机械设计与制造》
北大核心
2024年第8期54-58,共5页
Machinery Design & Manufacture
基金
国家科学基金面上项目(52077191)。
关键词
病虫害检测
马铃薯图像
主成分分析
支持向量机
快速检测
Pest Detection
Potato Image
Principal Component Analysis
Support Vector Machine
Rapid Detection