The goal of my Phd work is to propose a general approach for surveillance video indexing and retrieval. Based on the hypothesis that videos are preprocessed by an external video analysis module, this approach is composed of two phases:indexing phase and retrieval phase.

Fig 1. Architecture of the proposed approach

Indexing phase:
In order to profit from the output of various video analysis modules, a general data model consisting of two main concepts objects and events, is proposed.
Firstly, two new key blob detection methods in the object representation task choose for each detected object a set of key blobs associated with a weight. (cf. Fig 2) 

Fig 2. Representative blobs for a detected and tracked person

Retrieval phase:
The retrieval phase starts with a user query and is composed of 4 tasks. In the formulation task, user expresses his query in a new rich query language. This query is then analyzed by the syntax parsing task. A new matching method based on EMD (Earth Movers Distance) in the matching task aims at retrieving effectively relevant results [2]. Two proposed methods in the relevance feedback task allow to interact with the user in order to improve retrieved results[3].