基于SSGB一YOLOv5s的轻量级马铃薯疫病检测方法

打开文本图片集

中图分类号：S6；TP391.41 文献标识码：A 文章编号：2095-5553（2025）07-0211-09

Abstract：Rapidandaccurate identificationofcrops intheearlystageof epidemicdiseaseisanimportantlink toreducethe economicloss of crops.Aiming at theactual production environment，thetraditional image procesingalgorithms can not accurately identifythe leaf surface texture features and determine the typeof disease，and the YOLOv5s model hasa large number of parameters，andthe recognition efect is poor in complex environments and other problems，this paper proposes an integratedand improved method for potato disease detectionandidentification.In this paper，the numberof parameters isreducedbyreplacing thelightweight network ofYOLOv5s，thefusionabilityofdiferent featurelayersof the modelisenhancedbyusing the weightedbidirectional featurepyramid network（BiFPN），andtheattntion mechanism module SimAMis increased，theabilityof theYOLOalgorithmis enhanced toextract thekey information by usingthe GSConvconvolution，andfinally，theregressonaccuracyisimproved byintroducing theSIOUlossfunctionUnder the same experimental conditions，comparedwith the original model of YOLOv5s，YOLOv7—tiny，Faster R—CNN and other models，the accuracy rate，recallrate，and average recognition accuracy of this proposed method reaches 97.7% ， （20 95.9% and 95.4% ， respectively. The proposed algorithm not only improves the accuracy and average accuracy， but also reaches the speed of 144.93 frames/s，which meets the requirement of potato blight detection.

Keywords： potato blight detection;YOLOv5s； loss function； SimAM attention; lightweight network

0 引言

马铃薯别名土豆，是世界七种主要粮食作物之一，同时也是十大热门营养健康食品之一，具有良好的发展前景[12]。（剩余12708字）