An Unified Approach: Palmprint Recognition using Fused Local and Global Characteristics

The palmprint recognition model has been attracting researchers in the last few years. Among the mage concentrations, most of the researchers focused on local feature extraction model and recognition model, or a few focused on global feature extraction model for classification purposes. In this model, the main focus was on fusion of both the local and global features of palmprint image. Various traditional methods were used to extract both the global and local features of palmprint images. The feature selection algorithm selects the optimal features from a list to include in fusion. Hausdorff distance method is used for matching purpose. For this model, two different datasets (IITD and Tongji palmprint) were used. The accuracy of the dataset was 98.91% and 98.25%, respectively. Both datasets are contactless.