期刊文献+

基于差异双分支编码器的多阶段图像融合方法 被引量:1

Multi-stage Image Fusion Method Based on Differential Dual-Branch Encoder
下载PDF
导出
摘要 在现有的红外和可见光图像融合方法中,融合图像中的细节信息丢失严重,视觉效果不佳.针对上述问题,文中提出基于差异双分支编码器的多阶段图像融合方法.通过两支不同结构的编码器提取多模态图像的特征,增强特征的多样性.设计多阶段的融合策略,实现精细化图像融合.首先,在差异双分支编码器中,对两个编码分支提取的差异性特征进行初级融合.然后,在融合阶段,对多模态图像的显著性特征进行中级融合.最后,使用远程横向连接将差异双分支编码器的浅层特征传送给解码器,同时指导融合过程和图像重建.对比实验表明,文中算法可增强融合图像的细节信息,并在视觉效果和客观评价上都较优. In the existing infrared and visible image fusion methods,the details of the fused image are lost seriously and the visual effect is poor.Aiming at the problems,a multi-stage image fusion method based on differential dual-branch encoder is proposed.The features of multi-modal images are extracted by two encoders with different network structures to enhance the diversity of features.A multi-stage fusion strategy is designed to achieve refined image fusion.Firstly,primary fusion is performed on the differential features extracted by the two encoding branches in the differential dual-branch encoder.Then,mid-level fusion on the saliency features of the multi-modal images is conducted in the fusion stage.Finally,the long-range lateral connections are adopted to transmit shallow features of the differential dual-branch encoder implemented to the decoder and guide the fusion process and the image reconstruction simultaneously.Experimental results show the proposed method enhances the detailed information of the fused images and achieves better performance in both visual effect and objective evaluation.
作者 洪雨露 吴小俊 徐天阳 HONG Yulu;WU Xiaojun;XU Tianyang(Jiangsu Provincial Engineering Laboratory of Pattern Recognition and Computing Intelligence,School of Artificial Intelligence and Computer Science,Jiangnan University,Wuxi 214122)
出处 《模式识别与人工智能》 EI CSCD 北大核心 2022年第7期661-670,共10页 Pattern Recognition and Artificial Intelligence
基金 国家自然科学基金项目(No.62020106012,U1836218,62106089) 教育部111项目(No.B12018)资助。
关键词 图像融合 红外图像 可见光图像 卷积神经网络 Image Fusion Infrared Image Visible Image Convolutional Neural Network
  • 相关文献

参考文献6

二级参考文献39

  • 1汤雷,蒋金平,顾培英,胡振华.超声红外法检测混凝土构件试验研究[J].水利学报,2012,43(S1):70-75. 被引量:9
  • 2罗丽,袁真,王珂.基于àtrous-Contourlet变换的遥感图像融合算法[J].模式识别与人工智能,2007,20(2):248-253. 被引量:8
  • 3Gavrila D M. Vision-Based 3-D Tracking of Human in Action. Ph. D Dissertation. Maryland, USA: University of Maryland, 1996.
  • 4Sminchisescu C, Kanaujia A, Li Zhiguo, et al. Conditional Models for Contextual Human Motion Recognition / / Proc of the 10th Inter?national Conference on Computer Vision. Beijing, China, 2005, II: 1808-1815.
  • 5Belongie S, MalikJ, PuzichaJ. Shape Matching and Object Recog?nition Using Shape Context. IEEE Trans on Pattern Analysis and Machin", Intelligence, 2002, 24(4): 509-522.
  • 6Kumar S, Hebert M. Discriminative Random Fields: A Discrimina?tive Framework for Contextual Interaction in Classification / / Proc of the 9th IEEE Conference on Computer Vision. Nice, France, 2003, II: 1150-1157.
  • 7McCallum A, Freitag D, Pereira F. Maximum Entropy Markov Mo?dels for Information Extraction and Segmentation / / Proc of the 17 th International Conference on Machine Learning. Stanford, USA, 2000: 591-598.
  • 8Laptev I, Lindeberg T. Space-Time Interest Points / / Proc of the 9th IEEE Conference on Computer Vision. Nice, France, 2003, I: 432-439.
  • 9Bobick A F, DavisJ W. The Recognition of Human Movement Using Temporal Templates. IEEE Trans on Pattern Analysis and Machine Intelligence, 2001, 23 (3) : 257-267.
  • 10Kellokumpu V, Pietikainen M, HeikkilaJ. Human Activity Recogni?tion Using Sequences of Postures / / Proc of the International Associ?ation for Pattern Recognition Conference on Machine Vision Applica?tions. Tsukuba Science City,Japan, 2005: 570-573.

共引文献43

同被引文献17

引证文献1

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