Infrared and visible image fusion algorithm based on split-attention residual networks
在红外和可见光图像融合算法中, 图像信息的丢失始终是制约融合图像质量提升的关键问题, 为此, 提出了一种基于拆分注意力残差网络的红外和可见光图像融合算法, 使用带有拆分注意力模块的深层残差网络拓展感受野和提高跨通道信息融合能力, 运用平滑最大值单元函数作为激活函数进一步提升网络性能; 特征提取后运用零相位分量分析和归一化算法得到融合权重后完成图像融合。实验结果表明, 融合后的图像细节丰富, 边缘锐利; 在峰值信噪比、结构相似性指数度量和基于梯度的融合性能等指标上与经典的6种算法相比均有不同程度提升。...
Main Authors: | QIAN Kun, LI Tianyu, LI Zhe, CHEN Meishan |
---|---|
Format: | Article |
Language: | zho |
Published: |
EDP Sciences
2022-12-01
|
Series: | Xibei Gongye Daxue Xuebao |
Subjects: | |
Online Access: | https://www.jnwpu.org/articles/jnwpu/full_html/2022/06/jnwpu2022406p1404/jnwpu2022406p1404.html |
Similar Items
-
一种基于残差网络的改进网络流量分类算法
by: 陆煜斌, et al.
Published: (2021-01-01) -
Research on sonar image few-shot classification based on deep learning
by: CHEN Yule, et al.
Published: (2022-08-01) -
Predicting trajectory of drogue based on multi-head convolutional long-short-term memory network
by: WU Cihang, et al.
Published: (2022-06-01) -
Several combinatorial bijections about compositions without 2's of positive integers(关于正整数不含分部量2的有序分拆的几个组合双射)
by: GUOYuhong(郭育红), et al.
Published: (2017-05-01) -
一种基于聚合流的端到端统计时延界的新算法
by: 赵一A, et al.
Published: (2012-01-01)