改进YOLOv5的路面坑洞检测算法OA
Improved Road Pothole Detection Algorithm based on YOLOv5
路面坑洞的检测和识别对于自动驾驶和公路养护至关重要.为了解决在不同光照条件下以及坑洞积水所导致的路面坑洞识别难题,并将其应用于一些资源受限的现场环境中,本文提出了一种改进 YOLOv5 的路面坑洞检测算法.该算法利用 PConv 改进了 C3 模块来实现更快的检测速度,并通过采用动态检测头的不同感知模块来适应路面坑洞的不同特征.同时,将模型的损失函数CIoU改进为SIoU,提高模型的精度和收敛能力.实验的结果表明,改进后的YOLOv5 算法在路面…查看全部>>
Detection and recognition of road potholes are crucial for autonomous driving and highway maintenance.In order to address the challenges of road pothole recognition under different lighting conditions and water accumulation,and to apply it in resource-constrained field environments,this article proposes an improved YOLOv5 algorithm for detecting road potholes.The C3 module was improved using PConv to achieve faster detection speed,and different perception mo…查看全部>>
王耀宗;杨洁
西南林业大学 昆明 650224西南林业大学 昆明 650224
计算机与自动化
路面坑洞检测YOLOv5算法C3模块动态检测头损失函数
Road PotholesYOLOv5 AlgorithmC3 ModuleDynamic Detection HeadLoss Function
《福建电脑》 2024 (1)
16-20,5
评论