Search

Contactless palm vein recognition plays a significant role in biometric application because of its high stability, non-intrusive, flexibility and unique nature. Thus, different neural approaches were proposed to identify and segment the vein from the contactless palm image. But the traditional techniques face challenging issues in vein tracking and segmentation. Thus, a novel hybrid optimized deep network named Monkey-based Elman Neural Vein Recognition Framework was developed in this article. First, the dataset is pre-processed, and the palm region is extracted. Then, the extracted features are matched with the saved ground truth features. Further, the veins are tracked and segmented in the classification phase. The spider monkey fitness function is integrated into the developed model, which tracks and segments the vein from the palm image. The presented work was implemented, and the results are estimated for the palm image dataset. Furthermore, the results are verified with a comparative analysis. The highest accuracy score for the Contactless Palm-Vein Recognition by the proposed model is 99.76, and the lowest error rate is 0.0089%. Hence, the comparative analysis shows that the developed model earned better outcomes than the existing approaches. [ABSTRACT FROM AUTHOR]