Abdelhameed Ibrahim

Multimodal biometric authentication algorithm using ear and finger knuckle images

Biometrics that use physiological traits such as face, iris, fingerprints, ear, and finger knuckle (FK) for authentication face the problems of noisy sensors data, non-universality, and unacceptable error rates. Multimodal biometric methods use different fusion techniques to avoid such problems. Fusion methods have been proposed in different levels such as feature and classification level. This paper proposes two multimodal biometric authentication methods using ear and FK images. We propose a method based on fusion of images of ear and FK before the feature level, thus there is no information lost. We also propose a multi-level fusion method at image and classification levels. The features are extracted from the fused images using different classifiers and then combine the outputs of the classifiers in the abstract, rank, and score levels of fusion. Experimental results show that the proposed authentication methods increase the recognition rate compared to the state-of-the-art methods.

Spectral imaging method for material classification and inspection of printed circuit boards

We propose a spectral imaging method for material classification and inspection of raw printed circuit boards (PCBs). The method is composed of two steps (1) estimation the PCB surface-spectral reflectances and (2) unsupervised classification of the reflectance data to make the inspection of PCB easy and efficient. First, we develop a spectral imaging system that captures high dynamic range images of a raw PCB with spatially and spectrally high resolutions in the region of visible wavelength. The surface-spectral reflectance is then estimated at every pixel point from multiple spectral images, based on the reflection characteristics of different materials. Second, the surface-spectral reflectance data are classified into several groups, according to the number of PCB materials. We develop an unsupervised classification algorithm incorporating both spectral information and spatial information, based on the Nystrom approximation of the normalized cut method. The initial seeds for the Nystrom procedure are effectively chosen using a guidance module based on the K-means algorithm. Low-dimensional spectral features are efficiently extracted from the original high-dimensional spectral reflectance data. The feasibility of the proposed method is examined in experiments using real PCBs in detail.

Linear discriminant analysis: A detailed tutorial

Linear Discriminant Analysis (LDA) is a very common technique for dimensionality reduction problems as a preprocessing step for machine learning and pattern classification applications. At the same time, it is usually used as a black box, but (sometimes) not well understood. The aim of this paper is to build a solid intuition for what is LDA, and how LDA works, thus enabling readers of all levels be able to get a better understanding of the LDA and to know how to apply this technique in different applications. The paper first gave the basic definitions and steps of how LDA technique works supported with visual explanations of these steps. Moreover, the two methods of computing the LDA space, i.e. class-dependent and class-independent methods, were explained in details. Then, in a step-by-step approach, two numerical examples are demonstrated to show how the LDA space can be calculated in case of the class-dependent and class-independent methods. Furthermore, two of the most common LDA problems (i.e. Small Sample Size (SSS) and non-linearity problems) were highlighted and illustrated, and state-of-the-art solutions to these problems were investigated and explained. Finally, a number of experiments was conducted with different datasets to (1) investigate the effect of the eigenvectors that used in the LDA space on the robustness of the extracted feature for the classification accuracy, and (2) to show when the SSS problem occurs and how it can be addressed.

Ear Recognition Using Block-Based Principal Component Analysis and Decision Fusion

In this paper, we propose a fast and accurate ear recognition system based on principal component analysis (PCA) and fusion at classification and feature levels. Conventional PCA suffers from time and space complexity when dealing with high-dimensional data sets. Our proposed algorithm divides a large image into smaller blocks, and then applies PCA on each block separately, followed by classification using a minimum distance classifier. While the recognition rates on small blocks are lower than that on the whole ear image, combining the outputs of the classifiers is shown to increase the recognition rate. Experimental results confirm that our proposed algorithm is fast and achieves recognition performance superior to that yielded when using whole ear images.

Human Thermal Face Recognition Based on Random Linear Oracle (RLO) Ensembles

This paper proposes a human thermal face recognitionapproach with two variants based on Random linearOracle (RLO) ensembles. For the two approaches, the Segmentation-based Fractal Texture Analysis (SFTA) algorithmwas used for extracting features and the RLO ensembleclassifier was used for recognizing the face from its thermalimage. For the dimensionality reduction, one variant (SFTALDA-RLO) was used the technique of Linear DiscriminantAnalysis (LDA) while the other variant (SFTA-PCA-RLO) wasused the Principal Component Analysis (PCA). The classifiersmodel was built using the RLO classifier during the trainingphase and in the testing phase then this model was usedto identify the unknown sample images. The two variantswere evaluated using the Terravic Facial IR Database and theexperimental results showed that the two variants achieved agood recognition rate at 94.12% which is better than related work.

Invariant representation for spectral reflectance images and its application

Spectral images as well as color images observed from object surfaces are much influenced by various illumination conditions such as shading and specular highlight. Several invariant representations were proposed for these conditions using the standard dichromatic reflection model of dielectric materials. However, these representations are inadequate for other materials like metal. This article proposes an invariant representation that is derived from the standard dichromatic reflection model for dielectric and the extended dichromatic reflection model for metal. We show that a normalized surface-spectral reflectance by the minimum reflectance is invariant to highlights, shading, surface geometry, and illumination intensity. Here the illumination spectrum and the spectral sensitivity functions of the imaging system are measured in a separate way. As an application of the proposed invariant representation, a segmentation algorithm based on the proposed representation is presented for effectively segmenting spectral images of natural scenes and bare circuit boards.

Illumination-invariant representation for natural color images and its application

Illumination factors such as shading, shadow, and specular highlight, observed from object surfaces in a natural scene, affect seriously the appearance and analysis of the color images. This paper proposes an illumination-invariant representation that is derived from the standard dichromatic reflection model for inhomogeneous dielectric and the extended dichromatic reflection model for homogeneous metal. Illumination color is estimated from specular reflection component on inhomogeneous surfaces without using a reference white standard. The overall performance of the proposed representation is examined in experiments using real-world objects including metals and dielectrics in detail. The feasibility of effective edge detection is introduced and compared with the state-of-the-art illumination-invariant methods.

Human Thermal Face Extraction Based on SuperPixel Technique

Face extraction is considered a very important step in developing a recognition system. It is a challenging task as there are different face expressions, rotations, and artifacts including glasses and hats. In this paper, a face extraction model is proposed for thermal IR human face images based on superpixel technique. Superpixels can improve the computational efficiency of algorithms as it reduces hundreds of thousands of pixels to at most a few thousand superpixels. Superpixels in this paper are formulated using the quick-shift method. The Quick-Shift’s superpixels and automatic thresholding using a simple Otsu’s thresholding help to produce good results of extracting faces from the thermal images. To evaluate our approach, 18 persons with 22,784 thermal images were used from the Terravic Facial IR Database. The Experimental results showed that the proposed model was robust against image illumination, face rotations, and different artifacts in many cases compared to the most related work.

Material Classification for Printed Circuit Boards by Spectral Imaging System

This paper presents an approach to a reliable material classification for printed circuit boards (PCBs) by constructing a spectral imaging system. The system works in the whole spectral range [400-700nm] and the high spectral resolution. An algorithm is presented for effectively classifying the surface material on each pixel point into several elements such as substrate, metal, resist, footprint, and paint, based on the surface-spectral reflectance estimated from the spectral imaging data. The proposed approach is an incorporation of spectral reflectance estimation, spectral feature extraction, and image segmentation processes for material classification of raw PCBs. The performance of the proposed method is compared with other methods using the RGB-reflectance based algorithm, the k-means algorithm and the normalized cut algorithm. The experimental results show the superiority of our method in accuracy and computational cost.

Optimized superpixel and AdaBoost classifier for human thermal face recognition

Infrared spectrum-based human recognition systems offer straightforward and robust solutions for achieving an excellent performance in uncontrolled illumination. In this paper, a human thermal face recognition model is proposed. The model consists of four main steps. Firstly, the grey wolf optimization algorithm is used to find optimal superpixel parameters of the quick-shift segmentation method. Then, segmentation-based fractal texture analysis algorithm is used for extracting features and the rough set-based methods are used to select the most discriminative features. Finally, the AdaBoost classifier is employed for the classification process. For evaluating our proposed approach, thermal images from the Terravic Facial infrared dataset were used. The experimental results showed that the proposed approach achieved (1) reasonable segmentation results for the indoor and outdoor thermal images, (2) accuracy of the segmented images better than the non-segmented ones, and (3) the entropy-based feature selection method obtained the best classification accuracy. Generally, the classification accuracy of the proposed model reached to 99% which is better than some of the related work with around 5%.