Determining the dry parameter of fingerprint image using clarity score and ridge-valley thickness ratio

This paper proposes a novel procedure to determine the parameter values of dry fingerprint images based on the score of clarity and ridge-valley thickness ratio. The parameters are local clarity scores (LCS), global clarity scores (GCS) and ridge-valley thickness ratio (RVTR). Our analysis started by quantizing fingerprint images into blocks with size of 32×32 pixels. The orientation of each block was perpendicularly calculated to the ridge. The middle of the block along the ridge (two-dimensional vector V ₁ with the size 32×13 pixels) was extracted and transformed into a two-dimensional vertical vector V ₂. Linear regression applied to the one-dimensional vector V ₃ which is the average of vector V ₂ to produce a Determinant Threshold (DT ₁). Less than area of DT ₁ is called a ridge, while the opposite is a valley. The tests carried out by calculating the clarity of the image from the overlapping area of the gray-level distribution of ridge and valley that has been separated. The thickness ratio of ridge to valley was then computed for each block based on gray-level value per block of image in the normal direction toward the ridge. Finally, we found the thickness ratio of ridge to valley for all images from which the average value obtained. The results showed that the dry fingerprint could be obtained when the image parameters have LCS values between 0.0127 to 0.0149, GCS values between 0.0117 to 0.0120, RVTR values greater than 7.75E-05.