Test code for SRCNN
所属分类：图形图象
开发工具：matlab
文件大小：6444KB
下载次数：75
上传日期：2017-07-25 20:58:27
上 传 者：
MXN—_flying
说明： SRCNN代码实现。该代码使用三层卷积神经网络，进行图像的超分辨率重建，效果比双三次插值好很多
(The code uses three layer convolutional neural network for image super-resolution reconstruction, the effect is much better than the double three interpolation.)
文件列表:
SRCNN (0, 2014-09-03)
SRCNN\SRCNN.m (1267, 2014-09-02)
SRCNN\Set14 (0, 2014-09-03)
SRCNN\Set14\baboon.bmp (720054, 2013-10-06)
SRCNN\Set14\barbara.bmp (1244214, 2013-10-06)
SRCNN\Set14\bridge.bmp (263222, 2013-10-06)
SRCNN\Set14\coastguard.bmp (304182, 2013-10-06)
SRCNN\Set14\comic.bmp (271526, 2013-10-06)
SRCNN\Set14\face.bmp (228584, 2013-10-06)
SRCNN\Set14\flowers.bmp (543054, 2013-10-06)
SRCNN\Set14\foreman.bmp (304182, 2013-10-06)
SRCNN\Set14\lenna.bmp (786486, 2013-10-06)
SRCNN\Set14\man.bmp (786486, 2013-10-06)
SRCNN\Set14\monarch.bmp (1179702, 2013-10-06)
SRCNN\Set14\pepper.bmp (786486, 2013-10-06)
SRCNN\Set14\ppt3.bmp (1041782, 2013-10-06)
SRCNN\Set14\zebra.bmp (688214, 2013-10-06)
SRCNN\Set5 (0, 2014-09-01)
SRCNN\Set5\baby_GT.bmp (786486, 2013-10-06)
SRCNN\Set5\bird_GT.bmp (248886, 2013-10-06)
SRCNN\Set5\butterfly_GT.bmp (196730, 2013-10-06)
SRCNN\Set5\head_GT.bmp (235254, 2013-10-06)
SRCNN\Set5\woman_GT.bmp (235350, 2013-10-06)
SRCNN\compute_psnr.m (314, 2014-09-02)
SRCNN\demo_SR.m (1827, 2014-09-03)
SRCNN\modcrop.m (279, 2014-03-27)
SRCNN\model (0, 2014-09-02)
SRCNN\model\x2.mat (38625, 2014-09-02)
SRCNN\model\x3.mat (38483, 2014-09-01)
SRCNN\model\x4.mat (38489, 2014-09-01)
SRCNN\shave.m (110, 2013-10-06)
***********************************************************************************************************
***********************************************************************************************************
Matlab demo code for "Learning a Deep Convolutional Network for Image Super-Resolution" (ECCV 2014)
by Chao Dong (ndc.forward@gmail.com)
If you use/adapt our code in your work (either as a stand-alone tool or as a component of any algorithm),
you need to appropriately cite our ECCV 2014 paper.
This code is for academic purpose only. Not for commercial/industrial activities.
NOTE:
The running time reported in the paper is from C++ implementation. This Matlab version is a re-
implementation, and is for the ease of understanding the algorithm. This code is not optimized, and the
speed is not representative. The result can be slightly different from the paper due to transferring
across platforms.
***********************************************************************************************************
***********************************************************************************************************
Usage:
demo_SR.m - demonstrate super-resolution using SRCNN.m
function:
SRCNN.m - realize super resolution given the model parameters
Folders:
Set5 and Set14 - test images (Sec.4 in the paper)
Model - "x2.mat" "x3.mat" and "x4.mat" are model parameters used for upscaling factors 2,3 and 4 seperately
近期下载者：
相关文件：
收藏者：