Ear Localization using Hierarchical Clustering

Ear biometrics has been found to be a good and reliable technique for human recognition. With the initial doubts on uniqueness of the ear, ear biometrics could not attract much attention. But after it has been said that it is almost impossible to find two ears with all the parts identical, ear biometrics has gained its pace. To automate the ear based recognition process, ear in the image is required to be localized automatically. This paper presents a technique for the same. Ear localization in the proposed technique is carried out by using the hierarchical clustering of the edges obtained from the side face image. The technique is tested on a database consisting of 500 side face images of human faces collected at IIT Kanpur. It is found to be giving 94.6 % accuracy