首页|基于YOLOX-EfficientNet模型的路面病害识别研究

基于YOLOX-EfficientNet模型的路面病害识别研究

扫码查看
本文在YOLOX模型的基础上,以G320部分路段的路面存有病害的图片数据库为基础,融合了EfficientNet网络,建立完备的数据集,分析了路面多种病害的成因及特征,对数据集采用Labelimg标注软件对图片上的病害进行定位、分类,从而建立完备的模型训练、测试的数据集,以实现基于卷积网络的YOLOX-EfficientNet模型对路面病害进行识别、分类、定位。Efficient网络从多个方面对YOLOX模型进行优化,提高YOLOX模型的检测精度。
Automatic Pavement Disease Identification Research Based on YOLOX-EfficientNet Model
Based on the YOLOX model,this paper,based on the picture database of road diseases on some sections of G320,integrated the EfficientNet network to establish a complete data set,analyzed the causes and characteristics of various road diseases,and used Labelimg annotation software to locate and classify the diseases on the pictures in the data set. A complete dataset for model training and testing was thus established,and the YOLOX-EfficientNet model based on convolutional networks was efficientNet to identify,classify and locate pavement diseases. The Efficient network optimizes the YOLOX model from many aspects to improve the detection accuracy of the YOLOX model.

YOLOX-EfficientNetpavement damagetarget detectiondeep learning

陈杜强、廖俊杰、罗玉秋、戴苑彬、刘佳雨

展开 >

梅州市梅江公路事务中心,广东梅州 514000

华东交通大学交通运输工程学院,江西南昌 330013

YOLOX-EfficientNet 路面病害 目标检测 深度学习

2024

交通节能与环保
人民交通出版社股份有限公司,交通运输部公路科学研究院

交通节能与环保

影响因子:0.286
ISSN:1673-6478
年,卷(期):2024.20(5)