Conference Paper

Face Detection in Low-Resolution Color Images.

DOI: 10.1007/978-3-642-13772-3_46 In proceeding of: Image Analysis and Recognition, 7th International Conference, ICIAR 2010, Póvoa de Varzim, Portugal, June 21-23, 2010. Proceedings, Part I
Source: DBLP

ABSTRACT Most face detection methods require high or medium resolution face images to attain satisfactory results. However, in many
surveillance applications, where there is a need to image wide fields of view, faces cover just a few pixels, which makes
their detection difficult. Despite its importance, little work has been aimed at providing reliable detection at these low
resolutions. In this work, we study the relationship between resolution and the automatic face detection rate with the Modified
Census Transform, one of the most successful algorithms for face detection presented to date, and propose a new Color Census
Transform that provides significantly better results than the original when applied to low-resolution color images.

  • [Show abstract] [Hide abstract]
    ABSTRACT: This work presents a novel approach for pedestrian head localization and head pose estimation in single images. The presented method addresses an environment of low resolution gray-value images taken from a moving camera with large variations in illumination and object appearance. The proposed algorithms are based on normalized detection confidence values of separate, pose associated classifiers. Those classifiers are trained using a modified one vs. all framework that tolerates outliers appearing in continuous head pose classes. Experiments on a large set of real world data show very good head localization and head pose estimation results even on the smallest considered head size of 7×7 pixels. These results can be obtained in a probabilistic form, which make them of a great value for pedestrian path prediction and risk assessment systems within video-based driver assistance systems or many other applications.
    Pattern Recognition - 33rd DAGM Symposium, Frankfurt/Main, Germany, August 31 - September 2, 2011. Proceedings; 01/2011

Full-text (2 Sources)

Available from
Jun 5, 2014