摘要
针对粮虫检测中因虫体尺寸小、身体颜色与环境接近、藏匿环境复杂而导致的误检和漏检等问题,提出一种基于深度学习技术的改进YOLOv5粮虫检测算法。该算法通过改进Mosaic模块和特征尺度模块,增强对小目标的感知能力;通过改进卷积模块,在不影响检测速度的同时,增加了检测精度。试验结果表明,相比原算法,改进算法的检测时间降低了4.3 ms,提升了3.2%的检测精度。改进算法解决了粮虫检测中存在的误检、漏检等问题,为进一步研究高效的粮虫检测方法提供了理论参考。
To address the problems of false detection and missed detection caused by the small size of grain pest,close body color and environment,and complex hiding environment in grain pest detection,an improved YOLOv5 grain pest detection algorithm based on deep learning technology is proposed.This algorithm uses improved Mosaic method,modified feature scale,to enhance the perception of small targets.By improving convolutional attention module,the detection accuracy is increased without affecting the detection speed.Experimental results show that the improved YOLOv5 algorithm reduces the detection time by 4.3 ms and improves the detection accuracy by 3.2%compared with the original algorithm,solves the problems of false detection and leakage detection,and has better detection performance.It provides a theoretical reference for further research on efficient grain pest detection methods.
作者
刘明亮
刘晓鹏
周劲
杨红军
尹强
张永林
LIU Mingliang;LIU Xiaopeng;ZHOU Jing;YANG Hongjun;YIN Qiang;ZHANG Yonglin(School of Mechanical Engineering,Wuhan Polytechnic University,Wuhan 430023,China;School of Animal Science and Nutritional Engineering,Wuhan Polytechnic University,Wuhan 430023,China;School of Electrical and Electronic Engineering,Wuhan Polytechnic University,Wuhan 430023,China)
出处
《武汉轻工大学学报》
CAS
2023年第3期34-40,共7页
Journal of Wuhan Polytechnic University
基金
湖北省技术创新专项重大项目(2022BEC054)。
