安徽农业科学2025，Vol.53Issue(19)：207-212,6.DOI:10.3969/j.issn.0517-6611.2025.19.043
基于CBAM-ResNet50的昆虫识别系统的建立
Establishment of Insect Identification System Based on CBAM-ResNet50
摘要
Abstract
[Objectives]This research aims to construct an insect image dataset in a complex environment and propose an insect species recog-nition model based on an improved CBAM-ResNet50 network.[Method]The deep residual network ResNet50 was used as the backbone net-work in this research,and the mixed attention mechanism module CBAM is introduced to enable the model to more accurately extract insect fea-tures from images;and the updated training samples are used to train the classification model until the model reaches iteration and stops.[Re-sult]Compared with the ResNet network,after deep residual learning optimization,this method has further improved the recognition accuracy and speed of insect images.The results showed that this method achieved a recognition accuracy of 97.65%for 50 types of insects,which showed certain improvements in accuracy,precision,recall,and F1 value compared to the original model,and performed better.[Conclusion]This research model is feasible in insect recognition.关键词
昆虫识别/图像/深度学习/ResNet50/注意力机制Key words
Insect recognition/Image/Deep learning/ResNet50/Attention mechanism分类
农业科技引用本文复制引用
刘璇,张玉姣,杨晋宇,曹铭亮,靳德容,刘颍,李欣洋..基于CBAM-ResNet50的昆虫识别系统的建立[J].安徽农业科学,2025,53(19):207-212,6.基金项目
国家自然科学基金(31971651) （31971651）
科技基础资源调查专项(2023FY100301) （2023FY100301）
河北省自然科学基金项目(C2018204154). （C2018204154）
