One solution for detecting human faces in real time environment

Abstract

In this paper we present a new, fast and robust method, which is suitable for detecting human faces in real time. In order to speed up processing, the application must have some prior knowledge about the scene, so it only investigates the area in which movement is detected. The implemented method uses a histogram based algorithm to find the possible face areas. After finding the areas, a skin color segmentation and a face feature extraction method is applied to rank the decisions. Depending on the given rank, the system can decide which candidate can take place in further processing, for example in face recognition. This solution provides efficient human face recognition even in a fast changing scene.