2D Finger Shape Recognition Based on Local Zero-Order Moment Features

This paper proposed a biometric identification system through 2D fingers´ geometry measurements. The hand image captured from digital camera was first preprocessed to segment hand and its contour out from background. Then the finger tops and valleys were located by tracking out the maximum and minimum radial distances to a wrist reference point. The geometry features of index, middle, ring and little fingers were extracted, which included the length, mean width and normalized local zero-order moment. Experiments show that a high recognition accuracy and efficient matching performance were achieved with the proposed algorithm, which can meet the requirement of real time searching in large hand database.