IKP based human authentication using multi-feature fusion.(Conf 2017)
The mainstay of the project is to develop a novel framework to perform multi biometrics by comprehensively combining the left and right finger knuckle pattern images.
We propose a novel framework of combining the knuckle print at the matching score level. IKP recognition methods suffered from finger pose variations, as this lead to false rejections in the matching process. To handle such issues, a peg based imaging setup is developed to acquire the fixed position fingers; otherwise matching accuracy may fall massively due to flexibility in finger.