ROI selection for saliency maps

Region-of-interest selection for saliency maps

A detailed discussion of the ROI detection algorithm can be found here, with examples:

http://imageprocessingblog.com/region-of-interest-selection-for-saliency-maps/

This is an implementation of the algorithm described in our paper [1]. The input is any map generated by saliency detection algorithms like Itti-Koch [2] or GBVS [3]. The algorithm outputs a binary mask without requiring a threshold for the saliency map. More details about it are described in our paper.

Please cite our paper if you find it useful.

[1] Bharath, Ramesh, et al. "Scalable scene understanding using saliency-guided object localization." Control and Automation (ICCA), 2013, 10th IEEE International Conference on. IEEE, 2013.
http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=6565074

[2] Itti, Laurent, Christof Koch, and Ernst Niebur. "A model of saliency-based visual attention for rapid scene analysis." Pattern Analysis and Machine Intelligence, IEEE Transactions on 20.11 (1998): 1254-1259.

[3] Harel, Jonathan, Christof Koch, and Pietro Perona. "Graph-based visual saliency." Advances in neural information processing systems. 2006.