Önsen Toygar

Multiple classifier implementation of a divide-and-conquer approach using appearance-based statistical methods for face recognition

Abstract This paper presents a multiple classifier system for the face recognition problem-based on a novel divide-and-conquer approach using appearance-based statistical methods, namely principal component analysis (PCA), linear discriminant analysis (LDA) and independent component analysis (ICA). A facial image is divided into a number of horizontal segments and the associated local features are extracted using a particular statistical method. Using a simple distance measure and an appropriate classifier combination method, facial images are successfully classified. The standard FERET database and the FERET evaluation methodology are used in all experimental evaluations. Computational and storage space efficiencies and experimental recognition performance of the proposed approach indicate that significant achievements are obtained compared to individual classifiers.

Geometric leaf classification

Abstract In this paper, we propose a novel method including segmentation, a combination of new and well-known feature extraction and classification methods to classify plant leaves. The aim of the proposed features is to distinguish leaf margins, which cannot be distinguished using commonly used geometric features. Additionally, Linear Discriminant Classifier is used for classification, therefore using features that are noisy for some leaf types does not reduce the performance of the system. The proposed system outperforms the well-known geometric methods that are used for leaf classification.

Fusion of face and iris biometrics using local and global feature extraction methods

Fusion of multiple biometrics combines the strengths of unimodal biometrics to achieve improved recognition accuracy. In this study, face and iris biometrics are used to obtain a robust recognition system by using several feature extractors, score normalization and fusion techniques. Global and local feature extractors are used to extract face and iris features separately, and then, the fusion of these modalities is performed on different subsets of face and iris image databases of ORL, FERET, CASIA and UBIRIS. The proposed method uses Local Binary Patterns local feature extractor and subspace Linear Discriminant Analysis global feature extractor on face and iris images, respectively. Face and iris scores are normalized using tanh normalization, and then, Weighted Sum Rule is applied for the fusion of these two modalities. Improved recognition accuracies are achieved compared to the individual systems and multimodal systems using other local or global feature extractors for both modalities.

Selection of optimized features and weights on face-iris fusion using distance images

A novel system for face-iris fusion on distance images is proposed.Optimal features, feature extractors and weights are selected.Backtracking Search Algorithm and Particle Swarm Optimization are used.The proposed system outperforms state-of-the-art face-iris recognition systems.The proposed system is robust against spoofing attacks. The focus of this paper is on proposing new schemes based on score level and feature level fusion to fuse face and iris modalities by employing several global and local feature extraction methods in order to effectively code face and iris modalities. The proposed schemes are examined using different techniques at matching score level and feature level fusion on CASIA Iris Distance database, Print Attack face database, Replay Attack face database and IIIT-Delhi Contact Lens iris database. The proposed schemes involve the consideration of Particle Swarm Optimization (PSO) and Backtracking Search Algorithm (BSA) in order to select optimized features and weights to achieve robust recognition system by reducing the number of features in feature level fusion of the multimodal biometric system and optimizing the weights assigned to the face-iris multimodal biometric system scores in score level fusion step. Additionally, in order to improve face and iris recognition systems and subsequently the recognition of multimodal face-iris biometric system, the proposed methods attempt to correct and align the location of both eyes by measuring the iris rotation angle. Demonstration of the results based on both identification and verification rates clarifies that the proposed fusion schemes obtain a significant improvement over unimodal and other multimodal methods implemented in this study. Furthermore, the robustness of the proposed multimodal schemes is demonstrated against spoof attacks on several face and iris spoofing datasets.

Feature extractor selection for face–iris multimodal recognition

Multimodal biometrics-based systems aim to improve the recognition accuracy of human beings using more than one physical and/or behavioral characteristics of a person. In this paper, different fusion schemes at matching score level and feature level are employed to obtain a robust recognition system using several standard feature extractors. The proposed method involves the consideration of a face–iris multimodal biometric system using score level and feature level fusion. Principal Component Analysis (PCA), subspace Linear Discriminant Analysis (LDA), subpattern-based PCA, modular PCA and Local Binary Patterns (LBP) are global and local feature extraction methods applied on face and iris images. In fact, different feature sets obtained from five local and global feature extraction methods for unimodal iris biometric system are concatenated at feature level fusion called iris feature vector fusion (iris-FVF), while for unimodal face biometric system, LBP is used to achieve efficient texture descriptors. Feature selection is performed using Particle Swarm Optimization (PSO) at feature level fusion step to reduce the dimension of feature vectors for improving the recognition performance. Our proposed method is validated by forming three datasets using ORL, BANCA, FERET face databases and CASIA, UBIRIS iris databases. The results based on recognition performance and ROC analysis demonstrate that the proposed matching score level fusion scheme using Weighted Sum rule, tanh normalization, iris-FVF and facial features extracted by LBP achieves a significant improvement over unimodal and multimodal methods. Support Vector Machine (SVM) and t-norm normalization are also used to improve the recognition performance of the proposed method.

A NEW APPROACH FOR FACE-IRIS MULTIMODAL BIOMETRIC RECOGNITION USING SCORE FUSION

In this paper, a new approach based on score level fusion is presented to obtain a robust recognition system by concatenating face and iris scores of several standard classifiers. The proposed method concatenates face and iris match scores instead of concatenating features as in feature-level fusion. The features from face and iris are extracted using local and global feature extraction methods such as PCA, subspace LDA, spPCA, mPCA and LBP. Transformation-based score fusion and classifier-based score fusion are then involved in the process to obtain, concatenate and classify the matching scores. Different fusion techniques at matching score level, feature level and decision level are compared with the proposed method to emphasize improvement and effectiveness of the proposed method. In order to validate the proposed scheme, a combined database is formed using ORL and BANCA face databases together with CASIA and UBIRIS iris databases. The results based on recognition performance and ROC analysis demonstrate that the proposed score level fusion achieves a significant improvement over unimodal methods and other multimodal face-iris fusion methods.

A weighted mean filter with spatial-bias elimination for impulse noise removal

In this paper, we propose Unbiased Weighted Mean Filter (UWMF) for removing high-density impulse noise. Asymmetric distribution of corrupted pixels in the filtering window creates a spatial-bias towards the center of uncorrupted pixels. UWMF eliminates this bias by recalibrating the contribution factor (weight) of each uncorrupted pixel in such a way that the center shifts back to the center of the filtering window. The restoration process involves three sequential operations while convolving a filtering window over a contaminated image. Noise is detected, weights are recalibrated and the new intensity value is replaced by weighted mean using the recalibrated weights. Compared to the state-of-the-art impulse noise removal methods, UWMF provides superior performance, without requiring a fine-tuning for its parameters, in terms of both objective measurements and subjective assessments.

Recognizing partially occluded irises using subpattern-based approaches

In this study, iris recognition in the presence of partial occlusions is investigated using holistic and subpattern-based approaches. Principal Component Analysis (PCA) and subspace Linear Discriminant Analysis (ssLDA) methods are used as feature extractors to recognize iris images. In order to eliminate the effect of illumination changes, histogram equalization and mean-and-variance normalization techniques are used. The recognition performance of the holistic approaches is compared with the performance of subpattern-based approaches spPCA, mPCA and subpattern-based ssLDA approaches in order to demonstrate the performance differences and similarities between these two types of approaches in the presence of partial occlusions. Various experiments are carried out on CASIA, UPOL and UBIRIS databases to demonstrate the effect of occlusions on iris recognition with holistic and subpattern-based approaches.

Spoof detection on face and palmprint biometrics

Spoofing attacks made by non-real images are a major concern to biometric systems. This paper presents a novel solution for distinguishing between live and forged identities using the fusion of texture-based methods and image quality assessment measures. In our approach, we used LBP and HOG texture descriptors to extract texture information of an image. Additionally, feature space of seven full-reference complementary image quality measures is considered including peak signal-to-noise ratio, structural similarity, mean-squared error, normalized cross-correlation, maximum difference, normalized absolute error and average difference. We built a palmprint spoof database made by printed palmprint photograph of PolyU palmprint database using camera. Experimental results on three public-domain face spoof databases (Idiap Print-Attack, Replay-Attack and MSU MFSD) and palmprint spoof database show that the proposed solution is effective in face and palmprint spoof detection.

A Hybrid Approach for Person Identification Using Palmprint and Face Biometrics

This paper proposes hybrid approaches based on both feature level and score level fusion strategies to provide a robust recognition system against the distortions of individual modalities. In order to compare the proposed schemes, a virtual multimodal database is formed from FERET face and PolyU palmprint databases. The proposed hybrid systems concatenate features extracted by local and global feature extraction methods such as Local Binary Patterns, Log Gabor, Principal Component Analysis and Linear Discriminant Analysis. Match score level fusion is performed in order to show the effectiveness and accuracy of the proposed schemes. The experimental results based on these databases reported a significant improvement of the proposed schemes compared with unimodal systems and other multimodal face–palmprint fusion methods.