A New 2D Pulse Domain Transform (PDT) Feature Extraction Technique for Fingerprint Biometrics

Abstract

Feature extraction is a technique to extract derived value which is informative and non-redundant. There are many techniques have been reported in the literature on its usage application on signal and image. Some of the common techniques of feature extraction are Fast Fourier Transform, Wavelet Transform, Histogram of Oriented Gradients (HOG), Speeded Up Robust Features (SURF), Local Binary Patterns (LBP), Haar Wavelets, and Color Histograms. Recently, a new feature extraction technique based on Pulse Width Modulation (PWM) has been proposed. PWM is widely used for speed control technique and for the application in feature extraction. PDT is a feature extraction technique that derived from PWM. Initially, PDT applies on electrocardiogram (ECG) signal and draw out a non-invertible output. This output is presented as pulse waveform which is generated by the amplitude’s comparison of ECG signal with a periodic triangular waveform to get the intersection points. The time location of the intersection points are then used to be the transition state for output pulse to raise, fall or preserve the status quo. Since the PDT concept is only applied to 1D signal, it is interesting if the technique can be extending to 2D signal. This work presents a 2D PDT feature extraction technique. The concept of 2D PDT also based on the principle of PWM which is to process the incoming image (2D signal) with the aid of a 2D triangular waveform to form an output of 2D pulse waveform. Furthermore, the 2D triangular waveform is set up based on the desired threshold for its frequency and amplitude, namely, these thresholds are modulated in accordance with the size of the image. Afterward, the superposition process is proceeding between the processed image and modulated 2D triangular waveform to generate the intersection points for further forming of the output of 2D pulse waveform. The technique is then tested on fingerprint biometric and shows its ability to generate unique feature from the fingerprint images. This uniqueness takes shape caused by the inconstancy of the frequency and amplitude of pulse waveform. More than that, it able to provide a quite high accuracy rate and Area Under the Receiver Operating Characteristic curve (AUROC or AUC) reading. The accuracy is about 87% on average and the AUC is about 0.86 on average. In accordance with the result, 2D PDT is proven that it can perform as a good feature extraction technique with no less favorable than other outstanding technique