This toolbox is from the upcoming paper "Saliency of magnetic tile surface defects" .
The 14 models in this toolbox are: ITTI and BMS based visual attention mechanism; FT, LC and HC based on global color rarity; AC and MSS} based on local color rarity; SR, Rudinac and PHOT based on frequency domain analysis; RC , SF and GMR based on region or superpixel contrast; and MBP based on the fast Minimum Barrier distance transformation.
The models are fast in running, easy to transplant and denpended on opencv only.
The version of the Opencv we used is Opencv3.1, and we compile it with Vs2013 on a windows 10 x64 computer.
If you need some images to test, you can download from https://github.com/abin24/Magnetic-tile-defect-datasets.
If you want to use the Mcue method in our paper, you can run the McueSalTest.cpp file.
References of this toolbox can be found in our paper.
It is welcomed to use our toolbox. And if it is used in your research, please cite our paper.
If you have any questions, please contact me by
huangyibin2014@ia.ac.cn