Automated Kinship Recognition through Facial Feature Matching

With technological advancement, computer vision is now applied in every domain for the identification and recognition of objects, humans and many more. Similarly, in the past decades, numerous applications have been developed by using facial feature matching techniques for authentication, verification and recognition. The human brain works with the concept of pattern matching; therefore, face is one of the common and easiest constituents to recognize each other. Irrespective of environmental conditions, face recognition is the quickest and most dynamic approach used by humans. This chapter presents an automatic kinship recognition by extracting the familial traits features available in facial images. A kinship recognition system is a process to train a machine to verify the relation between kin and non-kin faces and to identify or classify exact type of relation. One of the methods for family verification, accurately, is performed via DNA pattern matching, but it is time taking, not suitable for mass population screening and a very expensive process. An automatic kinship system can be applicable for various applications such as genealogical research, controlling child trafficking, adopting children and finding the missing family member, forensics and many others. Due to the lack of a public database, this area is still unexplored and challenging. Still, deep learning and computer vision techniques can resolve and give better outcomes. Hence, this research includes computer vision and deep learning methods for kinship recognition along with previous research, applications and future directions.