Method for comparing iris images by the intelligent selection of textured areas
A method for comparing at least two iris images comprises determining M measurements each, of quality level associated with M regions each making up the first and second image. Said measurements are centered on M measurement points, the M measurements of the second image corresponding to the M measurements of the first image by the fact that the M measurement points of the second image correspond to the M measurement points of the first image. The method comprises merging the quality measurements, being obtained by the combination of two corresponding measurements belonging to the two images. The method also comprises selecting N regions exhibiting the N highest quality levels. The method also comprises encoding the two images by using the N selected regions to obtain a binary code for each image. Furthermore, the method comprises comparing the two binary codes to quantify the level of similarity between the two images.
1. A method for comparing at least two iris images comprising a step determining M measurements q i,j (1) representative of the quality level associated with M regions making up the first image, said regions being centered on M measurement points, a step determining M measurements q i,j (2) representative of the quality level associated with M regions making up the second image, said measurements being centered on M measurement points, the M measurements of the second image corresponding to the M measurements of the first image by the fact that the M measurement points of the second image correspond to the M measurement points of the first image, said method comprising: merging the quality measurements q i,j (1) , q i,j (2) , M merged measurements q i,j (f) , being obtained by the combination of two corresponding measurements belonging to the two images;
selecting N regions exhibiting the N highest quality levels;
encoding the two images by using the N selected regions so as to obtain a binary code for each image; and
comparing the two binary codes so as to quantify the level of similarity between the two images.
2. The method according to claim 1 , wherein the two images are segmented and normalized.
3. The method according to claim 1 , wherein the measurement points are distributed uniformly over the iris images.
4. The method according to claim 1 , wherein a merged measurement q i,j (f) associated with a given region is determined by selecting one of the measured values q i,j (1) , q i,j (2) associated with that same region, the value corresponding to the lowest quality level being chosen.
5. The method according to claim 1 , wherein a merged measurement q i,j (f) associated with a given region is determined by weighting the measurements of the two images q i,j (1) and q i,j (2) corresponding to that region.
6. The method according to claim 1 , wherein a merged measurement q i,j (f) associated with a given region is determined by using the following expression:
q i,j (f) =√{square root over ( q i,j (1) ×q i,j (2) )}{square root over ( q i,j (1) ×q i,j (2) )}.
7. The method according to claim 1 , wherein the quality measurements q i,j (1) are stored in a matrix Q 1 , the quality measurements q i,j (2) are stored in a matrix Q 2 , and the merged measurements q i,j (f) are stored in a matrix Qf.
8. The method according to claim 1 , wherein the binary codes are obtained by applying a phase demodulation around N points of application corresponding to the center of the N selected regions.
9. The method according to claim 1 , wherein the comparison step determines a similarity score corresponding to the Hamming distance between the two binary codes.
10. The method according to claim 1 , wherein the result of a measurement is a real number between 0 and 1, the value 1 being reached for a maximum quality of the measured region.
11. The method according to claim 2 , wherein the merging, selection and encoding steps are applied for a number of translated versions of the normalized iris images.