No description, website, or topics provided.
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
Demo
opencv2.4.9/build
x64
RGB2LAB.h
Saliency.cpp
SaliencyMDC.sln
SaliencyMDC.v11.suo
SaliencyMDC.vcproj
SaliencyMDC.vcxproj
SaliencyMDC.vcxproj.user
imgLib.cpp
imgLib.h
readme.txt
test.cpp

readme.txt

================================================================================================
               Xiaoming Huang, and Yu-Jin Zhang,
               300 FPS Salient Object Detection via Minimum Directional Contrast,
               IEEE Trans. on Image Processing
================================================================================================
* Copyright(c)2015-2017 Xiaoming Huang, <huangxm14@mails.tsinghua.edu.cn>.
* Last update:26-05-2017. 

This is the code of above work. The source code is developed with visual studio 2012 and opencv 2.4.9.
We have test the program on a 64-bit PC with Win7 OS,achieve 300 FPS speed performance on Intel Core i5-4590 CPU @ 3.3 GHz and 8GB RAM.
If you have any problem, please don't hesitate to contact with huangxm14@mails.tsinghua.edu.cn.

Note that we find that time cost of program fluctuate due to some reason, we suggest run 5 times and take average time cost.


1.Directory introduction
--Demo: One demo program
--SaliencyMDC.sln: solution file (visual studio 2012)
--opencv2.4.9:header files and lib of opencv
--test.cpp: test file of our method
--Saliency.cpp: detail implementation of our method

2.Build Project
Open solution file saliencyMDC.sln with VS2012, select Release @ x64 mode, build.

3.Run
You need set directory of input images, e.g., .\Demo\MSRA, then run the program, the final result will be also saved in the directory.