||One of the classical problems in computer vision is that of determining whether or not an image contains some specific object, feature or activity. This task can be solved robustly and without effort by a human, but it is still not solved in computer vision for the general case in which there is an arbitrary object in an arbitrary situation.
The study and the analysis of the visual information coming from an image can be tackled through global or local approaches. To global descriptors we preferred the local approach since recent research demonstrated that it leads to a more compact and robust representation of the image even when there are major changes in the object appearance. Thus in our work we concentrated on the use of local features and interest points to determine a representation of the image content by means of its most informative elements.
In this seminar we will first introduce some of the basic concepts related with image description and image representation, and then we will focus on the use of local descriptors for recognizing objects. We will show the results obtained with a prototype system capable to identify different known objects in video sequences.