Personal Identification Using Palmprint Images
Mrinal Kanti Dhar and Rupak Kanti Dhar (Leading University, Bangladesh); Mohammed J. Islam (Shahjalal University of Science & Technology, Bangladesh)
Biometric identifiers take distinctive characteristics into consideration which are robust parameters for recognition. In this paper, we present a novel and efficient approach for personal identification using texture based palmprint technology. The method has three steps - preprocessing, feature extraction and matching. In preprocessing, we have proposed a new method for ROI extraction using three valley points and two additional points on the hand image. Features are extracted using discrete cosine transform (DCT) and finally five distance classifiers are used to find the recognition rate. CASIA palmprint database is used to evaluate the performance of the model. The experimental results offer 97% recognition rate with total execution time less than 200ms per test, which is suitable for on-line identification system as well. It has been observed that the holistic DCT is little faster than the block-wise DCT for the proposed method. MATLAB R2015a is used to simulate the model.
Journal: International journal of simulation: systems, science & technology V18
Published: Mar 31, 2017