重参数化YOLOv8路面病害检测算法

Re-Parameterized YOLOv8 Pavement Disease Detection Algorithm

路面病害检测是保障人民交通安全的重要方式,为了提高路面病害检测的准确率,实现及时、精准的路面病害检测,提出了一种重参数化YOLOv8路面病害检测算法。在主干网络引入CNX2f模块,提高网络对路面病害特征的提取能力,有效解决路面病害特征与背景环境特征易混淆问题;引入RepConv和DBB重参数化模块,增强多尺度特征融合能力,解决路面病害尺度差异较大问题;改进头部采用共享参数结构,并引入RBB重参数模块,解决头部参数冗余问题,并提高特征提取能力;引入SPPF_Avg模块,解决路面病害特征丢失问题,丰富多尺度特征表达。实验结果表明,改进后的路面病害检测网络精度为73.3%、召回率为62.3%、mAP为69.3%,较YOLOv8网络分别提高了2.6、3.0、2.8个百分点,提高了模型的检测效果。

Road disease detection is an important way to ensure people’s traffic safety.In order to improve the accuracy of road disease detection and achieve timely and accurate road disease detection,a pavement disease detection model of re-parameterized YOLOv8 is proposed.First of all,CNX2f module is introduced into the backbone network to improve the ability of the network to extract pavement disease features,and effectively solve the problem that the pavement dis-ease features are easily confused with the background environmental features.Secondly,RepConv and DBB reparameter-ization modules are introduced to enhance the capability of multi-scale feature fusion and solve the problem of large scale difference of pavement diseases.At the same time,the shared parameter structure of the head is improved,and RBB repa-rameterization module is introduced to solve the problem of head parameter redundancy and improve the feature extrac-tion capability.Finally,the SPPF_Avg module is introduced to solve the problem of pavement feature loss and enrich the multi-scale feature expression.The experimental results show that the accuracy of the improved road disease detection net-work is 73.3%,the recall rate is 62.3%and the mAP is 69.3%,which is 2.6,3.0 and 2.8 percentage points higher than that of the YOLOv8 network,and the detection effect of the model is improved.