Research

Share    

Anomaly understanding visual attention

Objective

To develop a novel technique system for anomaly detection in the natural environment.

Description

The research focuses on small objects detection in the natural environment using the enhanced visual attention based saliency detection technique. The proposed detection scheme is a fusion of two major algorithms, namely the fast edge detection method based on structured random forest and the saliency detection model based on wavelet transform. In order to distinguish between moving and static objects, the enhanced spatial saliency map is fused with temporal saliency map, by which relative velocity of each moving object can be computed. The proposed saliency detection model has been tested on different natural scenes, including satellite image, Singapore’s East Coast in the daytime as well as evening. The results indicated that the fusion of different detection methods has significantly improved the detection performance of the model.



Result

        Enhance visual attention based saliency detection technique for both static and moving objects.



Demonstration Video



Publications

  1. Y. Fang, W. Lin, Z. Chen, C-M Tsai, C-W Lin, “A Video Saliency Detection Model in Compressed Domain”, IEEE Transactions on circuits and systems for video technology, VOL. 24, NO. 1, pp. 27 - 38, 2014.
  2. Y. Yuan, Y. Fang, W. Lin, “Visual Object Tracking by Structure Complexity Coefficients”,IEEE Trans. Multimedia, accepted, 2015.
  3. Y. Yuan, S. Emmanuel, Y. Fang, W. Lin, “Visual Object Tracking based on Backward Model Validation”, IEEE Transactions on Circuits and Systems for Video Technology, VOL. 24, NO. 11, pp. 1898 - 1910, 2014.



Principal Investigators


Associate Prof. Lin Weisi (NTU)

Telephone: 6790 6651
Office: N4-02b-60
Email: wslin@ntu.edu.sg
Mr. Paul Tan (STE)

Telephone: 6660 1052
Office: S1-B4a-03
Email: paultan@stengg.com