Fingerprint Image Processing for Automatic Verification

The performance of an automatic fingerprint verification approach relies heavily on the quality of the fingerprint image. Enhancement of the fingerprint image is then a crucial step in automatic fingerprint verification. This paper discusses the fingerprint image processing methods for automatic verification and proposes an adaptive oriented low pass filter to enhance the fingerprint image quality. For automatic fingerprint verification the fingerprint image processing is not aimed at improving the visual appearance of the fingerprint image but aimed at facilitating the subsequent processing. Therefore, the fingerprint image processing method is closely related to the method employed for the subsequent minutiae detection. The proposed approach takes efforts to increase the chances for success of the subsequent processes and at the same time avoid producing undesired side effects such as losing of the original ridge structure information or introducing additional spurious ridge structure information. Some sample results are given to illustrate the performance of the proposed approach.