BNCOD'13 Proceedings of the 29th British National conference on Big Data
| Hi-index | 0.00 |
This paper proposes an iDistance based interactive retrieval algorithm to accomplish semantic retrieval in visual surveillance system and improve the retrieval performance. The concept of user is interactively learned by training a SVM classifier from the user feedbacks. The search range of the algorithm is reduced by an iDistance based optimization algorithm. The surveillance objects are characterized by the color, texture and coefficient trajectory features. Experimental results on real scenes demonstrate the effectiveness of the proposed algorithm.