22-06-2013, 03:36 PM
Two-Stage Enhancement for Low Quality Fingerprint Images in Spatial and Frequency Domain
[attachment=55879]
Abstract
Fingerprint verification systems for content protection in the human–machine systems, is very popular. Due to the complex input contexts, input fingerprint images always exist with poor ridges and valley contrast ridges. Thus they become low quality fingerprint images. Usually, low quality fingerprint images are enhanced by one stage in either the spatial or the frequency domain. However, the enhanced performances are not satisfactory because of the complicated ridge structures that are affected by unusual input contexts. In this paper, we propose an effective two-stage enhancement scheme in both the spatial domain followed by the frequency domain by learning from the input images. We first enhance the fingerprint image in the spatial domain with a spatial ridge-compensation filter to remedy the ridge areas and enhance the contrast of the local ridges. Second stage filter, i.e., a frequency band pass filter that is separable in the radial- and angular-frequency domains, is employed with the help of the first step. The parameters of the band-pass filters are learnt from both the original image and the first-stage enhanced image instead of acquiring from the input image solely.
INTRODUCTION
Biometrics is described as the science of recognizing an individual based on his or her physical or behavioural traits, is beginning to gain acceptance as a legitimate method for the determination of an individual’s identity [1]. Biometric systems have now been deployed in commercial, civilian, and forensic applications as a means of establishing identity in the human–machine systems, cybernetics, and computational intelligence [2], [3]. In [2], the authors described an innovative multimodal biometric identification system that is based on fingerprints and iris in a reliable and dependable way, according to a specific-target application for security applications. In [3], the authors provided an overview of biometrics and discussed some of the salient research issues that need to be addressed for making biometric technology an effective tool. Among all the biometric indicators, fingerprints are one of the highest levels of reliability and have been extensively used by forensic experts in criminal investigations [4].
RELATED WORKS ON FINGERPRINT ENHANCEMENT
Spatial-Domain Filtering
The spatial-domain techniques involve spatial convolution of the image with filter masks, which is simple for operation. For computational reasons, such masks must be small in the spatial extent.
Contextual filters were used for fingerprint image enhancement. They used an anisotropic smoothing kernel whose major axis was oriented parallel to the ridges. For efficiency, the filter was pre-computed in 16 directions. The filter increased the ridge contrast in a direction perpendicular to the ridges, while it performed smoothing in the direction of the ridges.
Anisotropic filters based on structure-adaptive filtering were also used. The proposed filter has band-pass filter characteristics and is effective in removing noise, while preserving the local ridge frequency of the fingerprint image.
CONCLUSION
An effective two-stage enhancement scheme in both the spatial domain and the frequency domain for low quality fingerprint images which is based upon learning from the images has been proposed in this paper. The first-stage enhancement scheme has been designed to use the contextual information of the local ridges to connect or separate the ridges. Based on this spatial filtering, the broken ridges will be connected and the merged ridges will be well separated, thus the fingerprint ridges can be remedied and recovered well. In the second stage enhancement the filter is separable in the radial and angular domains, respectively. The parameters are determined from the original image and the enhanced images of the first stage instead of acquiring from the original image alone.
[attachment=55879]
Abstract
Fingerprint verification systems for content protection in the human–machine systems, is very popular. Due to the complex input contexts, input fingerprint images always exist with poor ridges and valley contrast ridges. Thus they become low quality fingerprint images. Usually, low quality fingerprint images are enhanced by one stage in either the spatial or the frequency domain. However, the enhanced performances are not satisfactory because of the complicated ridge structures that are affected by unusual input contexts. In this paper, we propose an effective two-stage enhancement scheme in both the spatial domain followed by the frequency domain by learning from the input images. We first enhance the fingerprint image in the spatial domain with a spatial ridge-compensation filter to remedy the ridge areas and enhance the contrast of the local ridges. Second stage filter, i.e., a frequency band pass filter that is separable in the radial- and angular-frequency domains, is employed with the help of the first step. The parameters of the band-pass filters are learnt from both the original image and the first-stage enhanced image instead of acquiring from the input image solely.
INTRODUCTION
Biometrics is described as the science of recognizing an individual based on his or her physical or behavioural traits, is beginning to gain acceptance as a legitimate method for the determination of an individual’s identity [1]. Biometric systems have now been deployed in commercial, civilian, and forensic applications as a means of establishing identity in the human–machine systems, cybernetics, and computational intelligence [2], [3]. In [2], the authors described an innovative multimodal biometric identification system that is based on fingerprints and iris in a reliable and dependable way, according to a specific-target application for security applications. In [3], the authors provided an overview of biometrics and discussed some of the salient research issues that need to be addressed for making biometric technology an effective tool. Among all the biometric indicators, fingerprints are one of the highest levels of reliability and have been extensively used by forensic experts in criminal investigations [4].
RELATED WORKS ON FINGERPRINT ENHANCEMENT
Spatial-Domain Filtering
The spatial-domain techniques involve spatial convolution of the image with filter masks, which is simple for operation. For computational reasons, such masks must be small in the spatial extent.
Contextual filters were used for fingerprint image enhancement. They used an anisotropic smoothing kernel whose major axis was oriented parallel to the ridges. For efficiency, the filter was pre-computed in 16 directions. The filter increased the ridge contrast in a direction perpendicular to the ridges, while it performed smoothing in the direction of the ridges.
Anisotropic filters based on structure-adaptive filtering were also used. The proposed filter has band-pass filter characteristics and is effective in removing noise, while preserving the local ridge frequency of the fingerprint image.
CONCLUSION
An effective two-stage enhancement scheme in both the spatial domain and the frequency domain for low quality fingerprint images which is based upon learning from the images has been proposed in this paper. The first-stage enhancement scheme has been designed to use the contextual information of the local ridges to connect or separate the ridges. Based on this spatial filtering, the broken ridges will be connected and the merged ridges will be well separated, thus the fingerprint ridges can be remedied and recovered well. In the second stage enhancement the filter is separable in the radial and angular domains, respectively. The parameters are determined from the original image and the enhanced images of the first stage instead of acquiring from the original image alone.