Biometrie based identification systems have beer widely used due to their reliability. Inner Knuckle Print images contain unique and reliable features for human identification. fr this paper, we propose a personal identification method using the Center Inner Knuckle Prints. The proposed method uses the Neighboring Direction Indicator features along with a perfect alignment and enhancement preprocessing steps that boost the performance compared to state-of-the-art methods. The performance of different feature extraction methods has been investigated using Sfax-Miracle Database, which is composed of low resolution hand images captured by a contactless capture in a free environment to test the effect of alignment and enhancement The effect of prints' fusion at the score level has also been investigated for a multimodal identification system. The result show that the proposed method outperforms state-of-the-are methods considering both Equal Error Rate and Best Identification Rate.
展开▼
