Automatic Performance Evaluation of Palmprint Recognition

This paper is concerned with the performance evaluation of palmprint recognition systems. We have designed and implemented an automatic system for performance evaluation (ASPE), which adopts the process kernel of palm recognition. We have also proposed a performance evaluation method based on minutiae, which are essentially terminations and bifurcations of the ridge lines that constitute a palmprint pattern. In this paper we define two evaluation indexes: recall and accuracy. The system can display the minutia in the image directly and output the two rates in the EXCEL form. The test result of speed and precise shows that automatic evaluation outperforms manual evaluation dramatically both in terms of efficiency and robustness.