HOW Finger Print Post Processing Works

Minutia extraction and post processing

This report consists of how we did minutia extraction and remove the false minutia in our project. Block diagram shows how finger print reorganization system works.

To extract the minutia we read many algorithms to extract the minutia but crossing number algorithm give the good result.

Crossing Number (CN) method for feature detection and extraction from the thinned image is shown in diagram. In the CN method, the extraction of the features is done through the scanning of the 3 x 3 matrix of each pixel in the thinned image. The CN value is then calculated from half the sum of the differences between pairs of adjacent pixels in the eight-neighborhood as shown in Equation.

The eight neighboring pixels of P are scanned in clockwise direction as given below

P2	P3	P4
P1	P2	P5
P8	P7	P6

If CN value is 1 then it is minutia if CN value is 3 then it is bifurcation. I make bit change in my code that I remove 0.5 from the formula and double the counts like if CN value is 2 it is minutia and if CN value is 6 it is bifurcation.

Now to remove the false minutia I make a function (false_mi_det ()) to remove false minutia. First I compare the image with the mask image to remove the false minutias which are at the edge of the image. First I locate the minutia then take the neighbor values and multiply with the values of the mask image at that location. After multiplication values are equal to the values of the taken values then it is the right minutia otherwise false.

To remove the false minutia form the image first locate the minutia then make a neighbored matrix and find the connected pixel with the minutia and assign then one after that count the number of transitions 0 to 1 or 1 to 0 at the edge of the matrix. For right minutia transitions should be two and for bifurcation transitions should be six other wise remove that minutia or bifurcation.

                                                    Valid Bifurcation           Valid Minutia

If two minutias are within given rage then that can be a broken line so if two minutias are so close then remove these minutia and if minutia is close to the bifurcation then it can be a spike so remove that minutia and bifurcation.

After implement this all in Matlab results are

If someone require further information indox me

Finger Print from Imran Baig

Finger print post processing from Imran Baig