Systems, Man, and Cybernetics

 

Home

• [P746] Efficient Entropy-based Features Selection for Image Retrieval
  Author(s): Yo-Ping Huang, Tsun-Wei Chang, Frode Eika Sandnes
  
  

  Abstract
  Information retrieval techniques should provide users a shortcut to search desired information quickly. There is no rule of thumb to explicitly express inexperienced users' query in retrieving images from certain ecological databases by high level semantics. An entropy-based feature selection strategy is proposed to find interesting images from databases in this paper. While retrieving the bird information from an ecological database, six visual features are designed for users to recall their memory on the just watched bird. Then, the query is formulated in terms of the chosen features. This implies that users' impressions or perception on the target are transformed into numerical data for retrieving target. The proposed method is tested in a real world bird database and the experimental results demonstrate the effectiveness of the presented work.

  
   

Full Text (PDF)