In this paper, a simple, yet effective method for TV
commercials detection is proposed, that exploits the presence or
absence, in the screen, of the broadcaster logo (or TV channel logo).
The approach is based on a digital on-screen graphics (DoG)
detection and classification mechanism, targeting to detect and
distinguish TV channel logos from other types of DoGs. No pre-built
database is required, as the proposed solution is able to gather its
own collection of DoGs from the broadcasted videos. A continuous
update and control of the DoGs database is performed, thus allowing
to conclude about the nature of each DoG and to classify each video
segment as Regular Program or Commercial Block. For the used test
video dataset, corresponding to recordings from three Portuguese
TV channels, a minimum accuracy of 93,9% on commercials
detection was achieved; furthermore, the measured processing time
suggests that the proposed solution should enable real-time (i.e.,
while recording) detection of commercial blocks.