Systems, Man, and Cybernetics

 

Home

• [P818] Foreground Segmentation via Sparse Representation
  Author(s): Bin Shen, Yu-Jin Zhang
  
  

  Abstract
  In this paper, we consider the the problem of foreground segmentation in videos. We propose bag of superpixels to simultaneously model both the foreground and the background. Then we show that an image has a hierarchical structure. Based on this observation, we exploit the discriminative nature of sparse representations to segment the foreground in each frame. Experimental results show that our algorithm can accurately locate the object's position and segment its image support.

  
   

Full Text (PDF)