基于深度学习的复杂背景下目标检测研究
作者:
作者单位:

1.西北工业大学 航天学院;2.航空工业西安航空计算技术研究所

作者简介:

通讯作者:

中图分类号:

TP391

基金项目:

国家自然科学基金项目(面上项目,重点项目,重大项目)


Research on Target Detection under Complex Background Based on Deep Learning
Author:
Affiliation:

School of Astronautics,Northwestern Polytechnical University,Xi '' an AVIC Computing Technique Research Institute,Xi '' an

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:

    目标检测是计算机视觉领域的重要研究方向。传统的目标检测方法在特征设计上花费了大量时间,且手工设计的特征对于目标多样性的问题并没有好的鲁棒性,深度学习技术逐渐成为近年来计算机视觉领域的突破口。本文对现有的基础神经网络进行研究,采用经典卷积神经网络VGGNet作为基础网络,添加部分深层网络,结合SSD(Single Shot MultiBox Detector)算法构建网络框架。针对模型训练中出现的正负样本不均衡问题,根据困难样本挖掘原理,在原有的损失函数中引入调制因子,将背景部分视为简单样本,减小背景损失在置信损失中的占比,使得模型收敛更快速,模型训练更充分,从而提高了复杂背景下的目标检测精度。同时,通过构建特征金字塔和融合多层特征图的方式,实现对低层特征图的语义信息融合增强,以提高对小目标检测的精度,从而提高整体的检测精度。仿真实验结果表明,本文提出的目标检测算法(FF-SSD,feature fusion based SSD)在复杂背景下对各种目标均可取得较高的检测精度。

    Abstract:

    Object detection is an important research in the field of computer vision. Traditional target detection methods spend a lot of time on feature extraction, and manual feature is not robust to the problem of diverse targets. Deep learning technology has gradually become a breakthrough in computer vision in recent years. By using classical convolutional neural network VGGNet as basic network, a novel network framework for target detection is built by adding some deep networks in combining with SSD (Single Shot Multi-Box Detector) algorithm. Aiming at the problem of sample imbalance during the training of the model, original loss function is modified according to the principle of hard example mining. Namely, the background is regarded as a simple sample and a modulation factor is introduced to reduce the proportion of background loss to the confidence loss, which makes the model be trained more fully and converge faster, and the target detection accuracy under the complex background is promoted as a result. Meanwhile, for poor detection effect to small targets of SSD algorithm, the feature pyramid is constructed according to the feature maps extracted from each convolutional layer. Appropriate feature maps are selected and fused to form a new feature map for the prediction. The semantic information fusion is strengthened to enhance the detection accuracy of small targets in order to improve the overall detection accuracy. Experimental results show that the proposed target detection algorithm(FF-SSD,feature fusion based SSD) can achieve high detection accuracy for all kinds of targets in complex background.

    参考文献
    相似文献
    引证文献
引用本文
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:2021-04-21
  • 最后修改日期:2021-07-22
  • 录用日期:2021-07-30
  • 在线发布日期: 2021-09-01
  • 出版日期: