Video Segmentation Based on SURF Flow Field

In recent days, the automatic extraction of salient object regions from video data is the most significant for visual analytical solutions. It also brings up more challenges, including pose differences amongst background and foreground objects, motion patterns etc. In this project, a framework of co-segmentation is presented in order to discover or segment the objects in a joint fashion from multiple videos and frames. Feature extraction is one of the main steps in object detection. For this feature extraction, Here introduce a speed up Robust features (SURF) flow descriptor to integrate the object features from the video frames in an optical manner. This novel SURF flow can extract foreground object over the complete video dataset.

[1]  V. Shunmughavel,et al.  Video Object Co-Segmentation , 2016 .

[2]  Andrew Blake,et al.  Cosegmentation of Image Pairs by Histogram Matching - Incorporating a Global Constraint into MRFs , 2006, 2006 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'06).

[3]  Vikas Singh,et al.  An efficient algorithm for Co-segmentation , 2009, 2009 IEEE 12th International Conference on Computer Vision.

[4]  Vikas Singh,et al.  Half-integrality based algorithms for cosegmentation of images , 2009, CVPR.

[5]  King Ngi Ngan,et al.  A Co-Saliency Model of Image Pairs , 2011, IEEE Transactions on Image Processing.

[6]  Joan Serrat,et al.  Video Co-segmentation , 2012, ACCV.

[7]  Mario Fritz,et al.  Multi-class Video Co-segmentation with a Generative Multi-video Model , 2013, 2013 IEEE Conference on Computer Vision and Pattern Recognition.