作者: Roszaharah Yaacob , Chok Dong Ooi , Haidi Ibrahim , Nik Fakhuruddin Nik Hassan , Puwira Jaya Othman
DOI: 10.1155/2019/5128062
关键词:
摘要: Palmprint has become one of the biometric modalities that can be used for personal identification. This modality contains critical identification features such as minutiae, ridges, wrinkles, and creases. In this research, feature from creases will our focus. Feature is a special salient palmprint. It worth noting currently, creases-based still not common. we proposed method to extract crease two regions. The first region interest (ROI) in hypothenar region, whereas another ROI interdigital region. To speed up extraction, most processes involved are based on processing image been downsampled by using factor 10. segmentations through thresholding, morphological operations, usage Hough line transform. Based 101 palmprint input images, experimental results show successfully extracts ROIs both achieved an average sensitivity, specificity, accuracy 0.8159, 0.9975, 0.9951, respectively.