Inception residual attention network for remote sensing image super-resolution

杨利霞

Inception residual attention network for remote sensingimage super-resolution

5 M( _; I2 d6 j
; P6 ^# J& N% N& [- l7 k* |: z' n6 i
How to enhance the spatial resolution for a remote sensing image is
8 h* |" o9 T. j8 Y  Y5 san important issue that we face. Many image super-resolution (SR)
techniques have been proposed for this purpose and deep con
, O; Q5 ?0 S& `- b: \/ ?% @$ \) Hvolutional neural network (CNN) is the most effective approach in
% l8 L; O$ u& w/ I6 N- Trecent 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
8 N; e6 j6 _( w& ~# R& o7 @! Rpaper, we propose a network named inception residual attention
" a0 C1 S! O' l+ f$ inetwork (IRAN) to address this problem. Specifically, we propose
a spatial attention module to make the network adaptively learn
8 M- ]8 Q$ p5 V( U' _% Tthe importance of different spatial areas, so as to pay more atten
tion to the areas with high-frequency information. Furthermore, we
present an inception module to fuse local multilevel features, so as
to provide richer information for reconstructing detailed textures. In
, @& Q: q( e1 s. D3 U1 T' U" Worder to evaluate the effectiveness of the proposed method, a large
+ e0 G( {- I; M* M2 a- ynumber of experiments are performed on UCMerced-LandUse data
, I8 S) Z! `+ x& Kset 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.
: N7 g' E2 j4 K/ J9 U" N- M# x. k


+ D6 u6 b. h8 C5 i, u