Inception residual attention network for remote sensing image super-resolution

杨利霞

Inception residual attention network for remote sensingimage super-resolution

  [, M4 D+ o; R; `How to enhance the spatial resolution for a remote sensing image is
an important issue that we face. Many image super-resolution (SR)
; ?, M9 A& z, M3 @; Btechniques have been proposed for this purpose and deep con
$ W9 \5 J: q0 r0 M- M! qvolutional neural network (CNN) is the most effective approach in
recent years. However, we observe that most CNN-based SR meth
ods treat low-frequency areas and high-frequency areas equally,
hence hindering the recovery of high-frequency information. In this
paper, we propose a network named inception residual attention
/ n0 f+ Z$ L4 p: R5 z- i9 b' ]6 Wnetwork (IRAN) to address this problem. Specifically, we propose
a spatial attention module to make the network adaptively learn
the importance of different spatial areas, so as to pay more atten
  W* k9 T5 r2 I7 \tion to the areas with high-frequency information. Furthermore, we
* e: ?) X$ m+ L3 ^; R+ C- lpresent an inception module to fuse local multilevel features, so as
7 F3 m! R# x. H" e* Zto provide richer information for reconstructing detailed textures. In
9 {" P4 N& X: X6 ?6 ?1 P* jorder to evaluate the effectiveness of the proposed method, a large
6 }5 q. O( z- k9 W! v8 Ynumber of experiments are performed on UCMerced-LandUse data
set and the results show that the proposed method is superior to
the current state-of-the-art methods in both visual effects and
+ N, H7 v* h3 D+ _* K( kobjective indicators.


7 Z6 \) P; N: K3 `& o( P
5 ^: [/ s! n+ `5 y4 S2 ~
1 S2 Q/ n$ H+ K5 J0 {7 l+ N