智能驾驶场景下的中小型障碍物检测方法
DOI:
作者:
作者单位:

新疆大学电气工程学院

作者简介:

通讯作者:

中图分类号:

TP391

基金项目:

国家自然科学基金(62303394);新疆维吾尔自治区自然科学基金(2022D01C694)


Small and Medium-sized Obstacle Detection Methods for Intelligent Driving Scenarios
Author:
Affiliation:

School of Electrical Engineering, Xinjiang University

Fund Project:

  • 摘要
  • |
  • 图/表
  • |
  • 访问统计
  • |
  • 参考文献
  • |
  • 相似文献
  • |
  • 引证文献
  • |
  • 资源附件
  • |
  • 文章评论
    摘要:

    针对智能驾驶场景下路面中小型障碍物易发生漏检、小目标障碍物难检测、模型参数量大等问题,提出了改进YOLOv8n的障碍物目标检测算法。在主干网络中融入分布移位卷积(Distribution Shifting Convolution,DSConv),将浮点运算替换为整数运算,减少了冗余计算量,通过量化和分布移位的方式模仿原始卷积层,维持了准确率;通过添加小目标检测层,更好地捕捉小目标的特征信息,适配小目标的尺度特征;结合SimAM无参数注意力机制,引入SPPF-SimAM模块,提高特征表示的质量与多样性,在不增加参数量的情况下实现了检测精度的提升;通过组合鬼影混洗卷积(Ghost-Shuffle Convolution,GSConv)和VoV-GSCSP模块的方式轻量化颈部特征融合网络,降低了模型的参数量和计算量。实验结果表明,改进后模型的准确率、召回率、平均精度均值相较于原始模型分别提升了1.6%、8.0%、6.2%,参数量降低了6.7%,所提算法有效提升了智能驾驶场景下中小型障碍物的检测精度,并且在检测性能与模型轻量化之间达到较好的平衡。

    Abstract:

    Aiming at the problems such as small and medium-sized obstacles on the road are prone to miss detection, small target obstacles are difficult to detect, and the number of model parameters is large in smart driving scenarios, the obstacle target detection algorithm with improved YOLOv8n is proposed. Distribution Shifting Convolution (DSConv) is incorporated into the backbone network, replacing floating-point operations with integer operations to reduce the amount of redundant computation, and mimicking the original convolutional layer by quantization and distribution shifting to maintain the accuracy; the feature information of small targets is better captured by adding a small-target detection layer to adapting the scale features of small targets; introducing the SPPF-SimAM module by combining the SimAM parameter-free attention mechanism to improve the quality and diversity of feature representations, and realizing the improvement of detection accuracy without increasing the number of parameters; and improving the detection accuracy by combining the Ghost-Shuffle Convolution (GSConv) and VoV-GSCSP module to lighten the neck feature fusion network, which reduces the number of parameters and computation of the model. The experimental results show that the accuracy, recall, and mean average precision of the improved model are improved by 1.6%, 8.0%, and 6.2%, respectively, and the number of parameters is reduced by 6.7% compared with the original model, and the proposed algorithm effectively improves the detection accuracy of small and medium-sized obstacles in smart driving scenarios, and achieves a better balance between the detection performance and the model lightweighting.

    参考文献
    相似文献
    引证文献
引用本文

龙小羽,南新元. 智能驾驶场景下的中小型障碍物检测方法[J]. 科学技术与工程, , ():

复制
分享
文章指标
  • 点击次数:
  • 下载次数:
  • HTML阅读次数:
  • 引用次数:
历史
  • 收稿日期:2024-04-06
  • 最后修改日期:2024-06-30
  • 录用日期:2024-07-09
  • 在线发布日期:
  • 出版日期:
×
亟待确认版面费归属稿件,敬请作者关注
《科学技术与工程》入选维普《中文科技期刊数据库》自然科学类期刊月度下载排行榜TOP10