Fuzzy Logic Based Adaptive Resonance Theory-1 Approach for Offline Signature Verification

Abstract

This paper presents the use fuzzy logic with adaptive resonance theory-1 in signature verification. Fuzzy model is capable of stable learning of recognition categories in response to arbitrary sequences of binary input pattern. The work was carried out on two famous available signature corpuses i.e. MCYT (Online Spanish signatures database) and GPDS (Grupo de Procesado Digital de la se?al). Local binary patterns (LBP) and Gray Level Co-occurrence Matrices (GLCM) features were calculated for robust offline signature verification system. Training and verification was done using fuzzy adaptive resonance theory-1(FART-1). The system is trained and verified for different datasets to increase the accuracy of the classifier. The results thus obtained are robust than other existing techniques. The FAR and FRR for the system is 0.74% and 0.83% respectively.

Offline Handwritten Signature Verification Based on Local Ridges Features and Haar Wavelet Transform

“…The obtained recognition rate was 92%.  In [9], the authors depended on fuzzy logic with adaptive resonance theory-1 for signature verification. They calculated Local binary patterns (LBP) and Gray Level Co-occurrence Matrices (GLCM) as features for the offline system.…”

“…The main focus of this stage is to enhance the image for feature extraction stage [9]. Preprocessing step performs the following operations: A.…”

“… Noise elimination using (5×5) mean filter. The scanned image contains spurious pixels (noise) that must be removed to guarantee accuracy in other processing steps [9]. B. Binary Image: to convert the grayscale image to binary image, a global threshold is selected automatically by taking the average of gray level pixels multiplied with α.…”

“…C. ROI Clipping: it determines which part of the image is rich in features, keeping the beneficial data in the ROI only and ignore the not useful data in the background [13]. D. Ridge Thinning: thinning the ridges to a single pixel width, making the process of feature extraction more efficient [9].…”

“…The used methods in [3,4,5] were explained in related works section. In [9], the authors used Deep Multitask Metric Learning as a classifier and Histogram of Oriented Gradients (HOG) in combination with Discrete Radon Transform as features. The table showed that the proposed method gave better results than the other methods.…”