A new head detection method based on the region shield segmentation in complex background

摘要

Automatic detection of the human head/face from complex backgrounds is one of the most difficult problems in pattern recognition. In this paper, a novel segmentation approach, namely Region Shield Segmentation (RSS), is proposed, which is used to segment the human head from complex backgrounds. The usual segmentation techniques segment the objects directly from backgrounds. RSS, however, shields the object region and processes the backgrounds first. It leaves the object region with a uniform background for further processing. The RSS process can effectively simplify the head/face detection algorithm. The MIT Face Database is used to evaluate the capacity of the proposed method. The results are found to be promising.