Inception residual attention network for remote sensing image super-resolution

杨利霞

Inception residual attention network for remote sensingimage super-resolution

8 n1 t; ^: z- V2 @7 c' x0 d
/ K$ o: _$ N* p$ vHow to enhance the spatial resolution for a remote sensing image is
an important issue that we face. Many image super-resolution (SR)
4 }6 f  u$ D4 F* X2 Gtechniques have been proposed for this purpose and deep con
volutional neural network (CNN) is the most effective approach in
recent years. However, we observe that most CNN-based SR meth
6 {0 W* L) }# B( }' tods treat low-frequency areas and high-frequency areas equally,
' R! A% ]) j$ e$ V( chence hindering the recovery of high-frequency information. In this
paper, we propose a network named inception residual attention
# x! H5 u. ^  K+ C; ^network (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
" K  O; B6 r. {- E/ Ztion to the areas with high-frequency information. Furthermore, we
present an inception module to fuse local multilevel features, so as
3 T) H6 F8 }+ o0 C& ~# I4 f9 Mto provide richer information for reconstructing detailed textures. In
order to evaluate the effectiveness of the proposed method, a large
; \5 [( \1 e" e- J) G5 v" E9 @number of experiments are performed on UCMerced-LandUse data
0 X6 ]' A$ a' J. V0 J/ [9 P5 |set and the results show that the proposed method is superior to
the current state-of-the-art methods in both visual effects and
objective indicators.

  M0 ]& o1 w3 T( h3 L! y