fingerprint images enhancement ppt
TRANSCRIPT
Analysis of Finger Print Images Using Access Controlled System
Submitted By : Mukta Gupta (00120802809)
Abhishake Gupta (05920802809) B.Tech (ECE-09)
Outline• What is An Access Controlled System
• Biometric Techniques used for Identification
• Benefits of Finger Print for Identification
• Finger Print Recognition Procedure
• Implemented algorithms
• Results
Access Controlled System
• Access: The flow of information between subject and object
• Access Controls: The security features that control how users and systemscommunicate and interact with one another.
• Subject: An active entity that requests access to an object or the data in an object
• Object: A passive entity that contains information
Access Controlled System
• Access: The flow of information between subject and object
• Access Controls: The security features that control how users and systemscommunicate and interact with one another.
• Subject: An active entity that requests access to an object or the data in an object
• Object: A passive entity that contains information
Biometric • Biometrics are used to identify people
based on their biological traits.• Most common biometric systems:
Fingerprint Palm Scan Hand Geometry Iris Scan Signature Dynamics Voice Print Facial Scan Hand Topography
Benefits of Finger Print Images In Identification
• Fingerprint identification is one of the most popular biometric identification technologies which are used in determining one’s identification uniquely.
Fingerprints have a wide variation since no two individual have identical prints.
There is high degree of consistency in fingerprints. A person’s fingerprint may change in scale but not in relative appearance.
Fingerprints are left each time the finger contacts a surface.
Comparative Survey of Finger Prints with Other Biometrics:
• Existence of small and inexpensive fingerprint capture devices
• Existence of fast computing hardware
• Existence of high recognition rate and speed equipment.• The explosive growth of network and Internet transactions
Features of Finger Print• LOOP : shape.• WHORL :ο shape.• HILL : ∆ shape.
Steps Involved in Finger Print Recognition Process
• Finger Print Image Enhancement segmentation Normalization Orientation estimation Frequency estimation Filtering
• Image Binarization • Image thinning • Minutiae extraction
Project Overview:Steps involved in the finger print enhancement
Finger Print Enhancement • Goal – to improve the clarity of the ridge
structure in the recoverable regions and mark unrecoverable regions as too noisy for further processing
• Input – a gray-scale image• Output – a gray-scale or binary image
depending on the algorithm• Effective initial steps - Segmentation,
Normalization, wavelet Transform, Ridge Orientation And Frequency Estimation
Need For Image Enhancement ??
• Performance depends on quality of images• Degradation types – ridges are not continuous,
parallel ridges are not well separated, cuts/creases/bruises
• Leads to problems in minutiae extraction
Step –I : Segmentation [Otsu’s method]
• To separate the foreground and background area• Steps involved
a. Divide the input image into the blocks of size w(8 × 8)b. Compute the mean & standard deviation of each block.c. Recombine the blocks, after comparing probabilistic parameters with a
threshold valueStd > threshold , ForegroundStd < threshold , Background
• Segmentation reduces the burden on next stages…
FOCUSSING ONLY ON THE FOREGROUND REGION
Segmentation: Foreground separated
Input B/W Image Output Segmented image
Step-II : Normalization• A linear and pixel-wise process.• Reduces the differences in the gray-level
values along the ridges and valleys without changing its structure.
Normalized ImageInput Segmented Image Output Normalized image
Step-III: Ridge Orientation Estimation• Orientation estimation is the first of the prerequisites for
fingerprint image filtering.• Ridges form patterns that flow in different directions
• Steps Involved:▫ Calculate Gradient of image▫ Compute Local orientation▫ Convert Oriented image into a continuous vector field▫ Perform Gaussian smoothing
Algorithm implemented Gx & Gy : Gradient of a 16*16 Gaussian filter is
operated on the image. Gxx , Gyy, Gxy : Co-variance of image gradient
is calculated and smoothened Analytic solution of principal direction is
calculated using : denom = sqrt(Gxx.^2 + Gyy.^2) sin2theta = Gxy./denom cos2theta = (Gxx-Gyy)./denom
Gaussian smoothing Intermediate results
Orientation (angle) lies b/w 0-3.14 rad.
Ridge’s OrientationNormalized Image Orientation of Image
Step-IV: Ridge Frequency Estimation
• Ridge distance: The distance from a given ridge to adjacent ridges.
• Ridge frequency: Reciprocal of ridge distance & indicates the number of ridges within a unit length.
• A frequency image F, analogous to the orientation image D, can be defined if the frequency is estimated at discrete positions and arranged into a matrix.
• Steps Involved: Divide image into blocks. Project the gray-level values of all the pixels located inside each
block along a direction orthogonal to the local ridge orientation.
It forms an almost sinusoidal-shape wave with the local minimum points corresponding to the ridges in the fingerprint.
The ridge spacing is calculated by counting the number of pixels between consecutive minima points in the projected waveform.
Ridge & Valley Topography
Algorithm Implemented Mean orientation of each block is calculated (26*26) Block rotated by 90◦ to make ridges vertical. Block cropped to eliminate Invalid regions. Columns summed down to get a projection of gray values. Dilation is performed with the structuring element. Then peaks are calculated using function. Spatial frequency of Ridges is calculated by
maxima points. Intermediate results (for 96 dpi image)
Wavelength: 2-25 pixels Frequency: 0.04 – 0.2 ridges/pixel Median frequency : 0.1333 ridges/pixel
Ridge Frequency Estimation
Normalized Image Frequency Estimation Results
Step-V: Filtering via Gabor Filter• Gabor filter is tuned to specific frequency and
orientation values [obtained from image]• Filter is convolved with the image.• Gabor filter can enhance the ridges in the
direction of local orientation effectively preserving the ridge structures.
It acts as a local band-pass filter with certain
optimal joint localization properties in both the spatial domain and the frequency domain.
What’s Gabor Filter ????????? • A Gabor filter is a linear filter whose impulse
response is defined by a harmonic function multiplied by a Gaussian function.
• It can be viewed as a sinusoidal plane of particular frequency and orientation, modulated by a Gaussian envelope.
• Both Orientation-selective & Frequency-selective properties.
• It preserves the ridge-valley structure of finger-print image.
Gabor filters are self-similar i.e. all can be generated by from one mother by DILATION
& ROTATION.
Gabor Filter…
Gabor Filtering ResultsOriginal Image Filtered Image
Step-VI: Binarization & Thinning
To skeletonize the FP image for minutiae extraction
To obtain an image’s best Performance & Threshold
Binarization: Ridges : 0 Valley : 1
Thinning: Reduces memory for structural information storage. Deletes the unwanted pixels and transforms the image
pattern one pixel thick
Binarization ResultsFiltered Image Binarized Image
Thinning ResultsBinarized Image Thinned Image
Thank you