Method of improving fingerprint images
View Patent ↗In a method of improving fingerprint images, wherein recorded images are processed on image areas of the fingerprint images by using Gabor filters, the Gabor filters are adapted to characteristic quantities of the image in the relevant image area to be processed.
1. A method of improving fingerprint images, wherein recorded images are processed on image areas of the fingerprint images by using Gabor filters, characterized in that the Gabor filters are adapted to characteristic quantities of the image in the relevant image area to be processed, and characterized in that the adaptation of the image area to be processed is determined by the following characteristic quantities:
mean direction of grooves in the image area,
mean spacing of the grooves in the image area,
change of direction of the grooves in the image area, and
change of spacing of the grooves in the image area.
2. A method as claimed in claim 1 , characterized in that the image areas are square and have a side length at which spacing and direction changes of the grooves are relatively small but comprise several grooves, such that statistical mean values and standard deviations can be formed for the spacings and directions.
3. A method as claimed in claim 1 , characterized in that, of an ellipse which encloses a Gaussian bell curve representing the Fourier transform of the Gabor filter, the length of the main axis directed towards the origin increases as the spacing of the grooves decreases, the length of the other main axis increases as the direction change increases, and the intersection of the main axis is determined by the mean direction and mean frequency of the grooves in the relevant image area.
4. A method as claimed in claim 1 , characterized in that the image areas overlap and before being combined, the overlapping image areas are weighted with a function decreasing towards the edges of the image areas.
5. A method of improving fingerprint images, wherein recorded images are processed on image areas of the fingerprint images by using Gabor filters, characterized in that the Gabor filters are adapted to characteristic quantities of the image in the relevant image area to be processed, and characterized in that the image areas overlap and before being combined, the overlapping image areas are weighted with a function decreasing towards the edges of the image areas.
6. A method as claimed in claim 5 , characterized in that the image areas are square and have a side length at which spacing and direction changes of grooves in the image areas are relatively small but comprise several grooves, such that statistical mean values and standard deviations can be formed for the spacings and directions.
7. A method as claimed in claim 5 , characterized in that the adaptation of the image area to be processed is determined by the following characteristic quantities:
mean direction of grooves in the image area,
mean spacing of the grooves in the image area,
change of direction of the grooves in the image area, and
change of spacing of the grooves in the image area.
8. A method as claimed in claim 7 , characterized in that, of an ellipse which encloses a Gaussian bell curve representing the Fourier transform of the Gabor filter, the length of the main axis directed towards the origin increases as the spacing of the grooves decreases, the length of the other main axis increases as the direction change increases, and the intersection of the main axis is determined by the mean direction and mean frequency of the grooves in the relevant image area.