Different illumination, rotation and position invariant palm print segmentation

Joint Event on 5^th International summit on Medical Biology & Bioengineering & 8^th International Conference & Exhibition on Biosensors and Bioelectronics

September 27-28, 2017 Chicago, USA

Ali Mamoud Mayya

Tishreen University, Syria

Posters & Accepted Abstracts: J Bioengineer & Biomedical Sci

Abstract :

In this paper, we introduce an innovative method for palm print segmentation to obtain the lifelines, points between fingers and the lines of fingers. Illumination variations are removed and the position of hand that is either left or right is corrected. Rotation correction is also performed to unify the rotation angle of hand. In the next stage, three segmentation operations are done. The first is the lifelines extraction depending on canny detector and morphological operations. The second is the extraction of points between fingers using masks derived from position of fingers. The third is the lines of fingers which are extracted depending on points between fingers and edges of Canny. The algorithm is applied on CASIA palm print database which consists of (2496) palm print images corresponding to (312) individuals and it achieved segmentation rate (99%).