面向无人驾驶场景下的道路多目标检测算法
A MULTI-TARGET DETECTION ALGORITHM OF ROAD FOR UNMANNED DRIVING SCENE
牛文杰 1伊力哈木·亚尔买买提1
作者信息
- 1. 新疆大学电气工程学院 新疆乌鲁木齐 830046
- 折叠
摘要
针对无人驾驶场景下目标检测算法误检率高的问题,设计一种改进YOLOv3的多目标检测算法.该文在原始特征提取网络Darknet53中引入分组卷积核替换标准卷积核,降低了卷积操作的计算量;改进原始YOLOv3的特征融合方法,使不同尺度的特征层融合更加充分,对遮挡目标和小目标的检测效果有明显提升;构建CIoU位置损失函数,提示网络收敛效果.实验结果表明,改进的YOLOv3算法平均精确度提高了 1.71％,误检率降低了 12％,明显优于原始算法.
Abstract
Aimed at the problem of high false detection rates of object detection in unmanned driving scene,a multi-target detection algorithm with improved YOLOv3 is proposed.The groups convolution kernel was introduced into the original feature network Darknet53 to replace the original convolution kernel,which reduced the complexity of convolution operation.The original feature fusion was improved to make the fusion of different scales more fully,and it improved the detection effect of occluded and small targets.The CIoU loss function was constructed to make the network convergence better.Experimental results show that the average accuracy of the improved YOLOv3 algorithm is increased by 1.71％,and the false detection rate is reduced by 12％,which is significantly better than the YOLOv3 algorithm.
关键词
无人驾驶/多目标检测/分组卷积/YOLOv3/CIoU损失函数Key words
Driverless/Multi-target detection/Group convolution/YOLOv3/CIoU loss function引用本文复制引用
基金项目
国家自然科学基金项目(61866037)
国家自然科学基金项目(61462082)
出版年
2024
国家科技期刊平台
NSTL
万方数据