Staty On Image Preprocessing In Fingerprint Recognition

The computer technology of fingerprint recognition has been applyied in a number of areas. The main application is in criminal investigation, safety inspection, office fingerprint card, car locks, bank safes, and other fields. The recognition algorithm generally consists of two stages.They are image preprocessing and feature matching.The image preprocessing stage is one of the important impact on the efficiency of identifying because feature extraction and matching is completely pretreatment results in the basis. So pretreatment effect will have a direct impact on the reliability of fingerprint profiling and identification has great influence on the results.Aiming at the current problems of fingerprint image preprocessing stage, this paper makes an in-depth study of the fingerprint image preprocessing algorithms. The main objective is to explore new ways to enhance the efficiency and the effect of fingerprint image pretreatment. Firstly, the article makes an in-depth research on the normalization and gray balance and various segmentation methods of the fingerprint image . Secondly, on the basis of classical algorithm of orientation image, the algorithm of orientation image based on the orientation window is put forward ,which makes full use of information of pixel neighborhood .Then the orientation image is applied in Gabor filter to enhance fingerprint image. Finally, a table-look-up thinning algorithm is put forward which uses new eliminate template. The new thinning algorithm solves the rapid thinning algorithm's problem of incomplete thinning and the improved OPTA algorithm's problem of burr, and good effect is obtained.