Real-Time Visual Recognition of Ob jects and Scenes Using P-Channel Matching

Felsberg, M., Hedborg, J.
SCIA07
Proc. 15th Scandinavian Conference on Image Analysis
Volume 4522, Pages 908-917
2007

Abstract

In this paper we propose a new approach to real-time view-based object recogniti on and scene registration. Object recognition is an important sub-task in many a pplications, as e.g., robotics, retrieval, and surveillance. Scene registration is particularly useful for identifying camera views in databases or video sequen ces. All of these applications require a fast recognition process and the possib ility to extend the database with new material, i.e., to update the recognition system online.

The method that we propose is based on P-channels, a special kind of information representation which combines advantages of histograms and local linear models. Our approach is motivated by its similarity to information representation in bi ological systems but its main advantage is its robustness against common distort ions as clutter and occlusion. The recognition algorithm extracts a number of ba sic, intensity invariant image features, encodes them into P-channels, and compa res the query P-channels to a set of prototype P-channels in a database. The alg orithm is applied in a cross-validation experiment on the COIL database, resulti ng in nearly ideal ROC curves. Furthermore, results from scene registration with a fish-eye camera are presented.

Full Paper

Portable document format file PDF

Bibtex entry

@inproceedings{fh07,
        Author = {Felsberg, M. and Hedborg, J.},
        Booktitle = {Proc. 15th Scandinavian Conference on Image Analysis},
        Date-Added = {2007-08-07 11:14:41 +0200},
        Date-Modified = {2007-08-07 11:14:41 +0200},
        Pages = {908--917},
        Series = {LNCS},
        Title = {Real-Time Visual Recognition of Ob jects and Scenes 
                 Using P-Channel Matching},
        Volume = {4522},
        Year = {2007}}