A robust eyelid and eyelash removal method and a local binarization based feature extraction technique for iris recognition system

Biometric systems are becoming popular in replacing conventional security systems. Out of several, iris recognition systems are proving to be highly successful due to their high accuracy and low error rate. This paper presents a new and simple method of extracting features from the iris pattern hence avoiding most of the heavy computations and complexity observed in existing algorithms. In addition to this, robust techniques for eyelid and eyelash detection have been proposed. The proposed detection and encoding scheme gives satisfactory results when tested on several existing databases and outperforms existing techniques when compared to them.