基于机器视觉技术的辣椒果实炭疽病病害分级方法研究

打开文本图片集

关键词：辣椒（CapsicumannuumL.）果实；机器视觉技术；炭疽病;病害分级

中图分类号：TP391.41;S436.418 文献标识码：A

文章编号：0439-8114（2025）08-0017-07

DOI：10.14088/j.cnki.issn0439-8114.2025.08.003

Research on anthracnose disease grading method for pepper fruits based on machine vision technology

ZOUWei，YUEYan-bin，LILi-jie，CHENWei-rong，HANWei，ZHUCun-zhou （Guizhou Agricultural Science and Technology Information Institute，Guiyang 55ooo6，China）

Abstract：Toaddress theissesofstrongsubjectivityandlowdetectioneficiencyintraditionalpepper（CapsicumannuumL）disease gradingmethods，thisudyproposedamachinvision-basedsemanticsegmentationmodelforautomatedrapidgradinganddentificationofanthracnose-infectedpepperfruits.Undercontroledenclosedenvironments，sunlightwassimulated，andimagesof healthyfruitsandfourdiseaseseveritylevelsacrossdiferentpeppervarietieswerecollcted.Principalcomponentanalysisasemployedtoreduceredundantimagefeatures，extracting threekeycolorcomponents（Cb，Cr，R）withacumulativecontributionrateof 95 % .Model1（Decision Tree），model2（Naive Bayes），model3（SVM），and model4（KNN）were trained.Model1（Decision Tree）demonstratedthesortesttraining tieandhighestprecision，stablisingitastheoptialpredictiomodelforanthracosediseasegrading.Itrequiredlowcomputationalresourcesandoccupied minimalmemory，facilitating future edgedeployment.Model1 achieved precision rates of 90.3%～98.2% for pepper fruits and 75.3% ～80.7% for disease spots.Its recall rate for anthracnose disease grading was 73.3%～93.3% ，with the recall rate for healthy peppers（Level O） exceeding 90.0% . The prediction results of model 1 showedighconsistencywithmanualanotationsacrossalldiseaselevels，verifingitsapplicabiltinautomateddiseaseoitoring systems as a replacement for manual visual grading methods.

Key words ： pepper（Capsicum annuum L.） fruit；machine vision technology；anthracnose；disease grading

辣椒（CapsicumannuumL.）作为中国广泛种植的蔬菜作物，种植面积及产量常年位居世界首位。（剩余7845字）