Face Detection and Recognition using Local Binary Patterns

Abstract

The human face conveys a lot of information about identity and emotional state of a person. Face recognition is an interesting and challenging task and impacts important applications in many areas such as identification for law enforcement, authentication for banking and security system access and personal identification. This paper proposes an Automatic Face Verification system using Local Binary Patterns (LBP). It is a texture based algorithm used for face recognition that describes the texture and shape of digital face image. The digital facial image is first divided into small blocks from which Local Binary Patterns are formed and then concatenated into a single feature vector. This feature vector plays an important role in efficient representation of the face and is used to measure similarities by calculating the distance between feature points.