Fingerprint Image Segmentation Method Based on Gray Vision Parameter

In order to make the fingerprint image pre-processing algorithm be more consistent with human visual cognitive processes, we accorded to visual perception model to calculate visual experience gray value, and calculate the parameters of the fingerprint image segmentation, achieved a new method. First, the fingerprint image is segmented using the method called OSTU with the physical gray value; Second, using Gray Vision Limens Function (GVLF), the Gray Vision Parameter(GVP) is figured out based on the physical gray value; Third, the converted fingerprint image is segmented also using the OSTU with the GVP; Last, the binary fingerprint image is done also using the OSTU method. The experimental result shows that this four-step segmentation method improves the effective of the poor quality fingerprint image, such as “Peeling”, “Black back ground”, “Tint”, and “Break”. Compared with the classical segmentation method, OSTU, our method could take more accurate segmentation and information from corrupted part of the poor fingerprint images and gets the more complete fingerprint images.