期刊文献+

基于改进的Faster R-CNN图像目标检测方法研究 被引量:6

下载PDF
导出
摘要 随着科技的不断发展,图像充斥在人类生活的任何角落,因此如何提取出图像中包含的信息,完成对图像目标的检测是当前研究的热点问题。针对一阶段目标检测算法可能会对目标图像产生漏检的情况,本文使用MobileNet V2和ResNet50网络对传统的Faster R-CNN的主干特征提取网络进行改进。在公开数据集上的实验结果表明,基于MobileNet V2特征提取的Faster R-CNN网络占用的计算资源和存储资源最少,基于ResNet50特征提取的Faster R-CNN网络的检测效果最优。此外,两种改进的Faster R-CNN网络均能有效克服一阶段目标检测算法中的漏检问题。
作者 代恒军 DAI Hengjun
出处 《信息技术与信息化》 2023年第8期91-94,共4页 Information Technology and Informatization
  • 相关文献

参考文献3

二级参考文献29

  • 1HUANG P F, CA/J, MENG Z J, et al. Novel method of monocular real-time feature point tracking for tethered space robots]-J:. Journal of Aerospace Engineering, 2014, 27(6): 04014039.
  • 2WONG S K. Non-cooperative target recognition in the fre- quency domain J-JJ. IEE Proceedings-Radar, Sonar and Navigation, 2004, 151(2): 77-84.
  • 3RAMIREZ V A, GUTIERREZ S A M, YANEZ R E S. Quadrilateral detection using genetic algorithms[J]. Computaei6ny Sistemas, 2011, 15(2) .. 181-193 .
  • 4CAI J, HUANG P F, WANG D K. Novel dynamic tem- plate matching of visual servoing for tethered space robot [C://2014 4th IEEE International Conference on Informa- tion Science and Technology (ICIST). Piscataway, NJ.. IEEE Press, 2014: 389-392.
  • 5ZITNICK C L, DOLLAR P. Edgebox: Locating object proposals from edges[C]//Computer visi0n-ECCV 2014.Zurich: Springer International Publishing, 2014.. 391- 405.
  • 6HEITZ G, KOLLER D. Learning spatial context: Using stuff to find things[C]//Computer vision-ECCV 2008. Marseille: Springer, 2008: 30-43.
  • 7YANULEVSKAYA V, UIJLINGS J, GEUSEBROEK J M. Salient object detection: From pixels to segments[J]. Image and Vision Computing, 2013, 31(1): 31-42.
  • 8ALEXE B, DESELAERS T, FERRARI V. Measuring the objectness of image windows[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2012, 34 (11) : 2189-2202.
  • 9RAHTU E, KANNALA J, category independent object BLASCHKO M. Learning a detection cascade [ C]//2011 IEEE International Conference on Computer Vision (IC- CV). Piscataway, NJ: IEEE Press, 2011: 1052-1059.
  • 10CHENG M M, ZHANG Z M, LIN W Y, et al. BING: Binarized normed gradients for objectness estimation at 300fps[C]//2011 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Piscataway, NJ: IEEE Press, 2014.

共引文献16

同被引文献45

引证文献6

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部