Önsen Toygar danışmanlığındaki tezler
17 tez · Eastern Mediterranean University
Fruit Classification using Global and Local Descriptors
Recognizing different kinds of food such as vegetables and fruits is a recurrent task in supermarkets where the cashier must be able to point out not only the species of a particular fruit but also its variety which will determine its price. The use of barcodes has mostly ended this problem for packaged products but given that consumers want to pick their produce, they cannot be packaged, and thus must be weighted. A common solution to this problem is issuing codes for each kind of fruit/vegetable; which have problems given that the memorization is hard, leading to errors in pricing. In view of this, attention for classification and matching of these foods were carried out using global and local descriptors. In this thesis, global descriptors such as Principal Component Analysis (PCA), Histograms of Oriented Gradients (HOG) and local descriptors such as Local Binary Patterns (LBP), Binarized Statistical Image Features (BSIF) are implemented in order to classify fruits. Experiments are conducted on two datasets from Fruits_360 database and TropicalFruits database. Experimental results obtained with global and local descriptors are presented as a comparative analysis on fruit classification on the aforementioned datasets. Among all descriptors, BSIF results are better than the other algorithms employed with 70.06% and 75.00% on the aforementioned datasets, respectively. On the other hand, LBP algorithm achieved 61.11% and 75.00% recognition rate while HOG results are 37.96% and 58.33% and PCA results are 42.90% and 45.83% on both datasets, respectively. The results show that local descriptors achieve better performance compared to the performance of the global descriptors for fruit classification. Keywords: Fruit classification, Global Descriptors, Local Descriptors, PCA, HOG, LBP, BSIF.
Identification of Plant Diseases Using Gray-Level and Color-Based Features of Leaf Images
The detection of plant diseases is a vital factor in agricultural production worldwide, which if ignored, can lead to tremendous losses of plant products and revenue. Farmers and researchers from many centuries ago have learnt to identify some plant disease manually by inspection, but presently, technological development have advanced cultivation to an industrial scale, therefore detection of plant diseases has also become a great issue of concern as the farmers may be unable to identify the diseases, their point of origin or even the infected plants early enough. This can lead to a disease outbreak. Early detection of plant diseases can immensely reduce or avoid massive potential losses as it will provide the opportunity for active and cautionary measures. In view of the aforementioned issue, carrying out researches on different ways and methods to curb this problem is a vital necessity. This thesis study employs the application of computer vision and image processing techniques for plant disease identification. Principal Component Analysis (PCA), Local Binary Patterns (LBP) and Completed Local Binary Patterns (CLBP) feature extraction methods are used for the extraction of texture-based and appearance-based image features. Disease symptoms are analyzed and identified from four different plant leaves to evaluate the performance of the proposed method. We propose a method that incorporates Feature-Level Fusion of the gray level features and color based features using PCA and LBP methods to create a robust system. The proposed method has proven to be more robust compared to the individual systems using LBP and CLBP. Experiments are conducted on PlantVillage dataset due to its diversified collection of plant leaves. Furthermore, two classifiers are used for classification purposes namely k-Nearest Neighbor (k-NN) and Support Vector Machine (SVM). At the end of the empirical evaluations, a comparative study is presented. Keywords: plant disease identification, leaf images, color spaces, Feature-Level Fusion, feature extraction, texture-based features, appearance-based features.
Texture Classification Using Texture-based Feature Extraction Algorithms
Texture is one of the significant characteristics used in identifying objects of interest or regions in an image. Texture is an important characteristic of surface property in visual scenes and is a power cue in visual perception. The real applications of texture classification are remote sensing, medical imaging, industrial inspection and pattern recognition. Texture images are highly affected by rotation and illumination. Extracting texture features that are rotation-invariant and insensitive to illumination with high classification accuracy is still a challenge. Texture analysis has been a popular area of study in computer vision for decades. In this thesis, six texture-based feature extractors that may perform variously to rotation and illumination are used namely Local Binary Patterns (LBP), Complete Local Binary Patterns (CLBP), Segmentation-based Fractal Texture Analysis (SFTA), Histogram of Oriented Gradients (HOG), Rotation Invariant Histogram of Oriented Gradients (RIHOG) and Haralick feature extractor. They are implemented and tested on three benchmark texture databases, such as The Columbia-Utrecht Database (CUReT), University of Oulu Texture database (OUTex) and Textured Surfaces Database. For feature matching, two classifiers are used namely Naive Bayes and Support Vector Machines (SVM). A comparative study is presented at the end of the experimental evaluations on texture classification.
Gender Classification Using Local Binary Patterns and its Variants
Many social interactions and services are dependent on gender today, so, gender classification is appearing as an active research area. Most of the existing studies are based on face images acquired under controlled conditions. In our work, we used different databases such as FERET, AR and ORL for controlled conditions and Labeled Faces in the Wild (LFW) database as real-life faces for uncontrolled conditions. Local Binary Patterns (LBP) and its variants such as Uniform LBP, Completed LBP and Rotation - Invariant LBP are employed to describe faces by extracting features from the region of interests. Manhattan distance measure is used to compare difference between test and training images for gender recognition. Based on the results reported as the state-of-the-art, we have achieved satisfactory results. Keywords: Gender recognition, feature extraction, Local Binary Patterns (LBP)
Fusion of Hand-crafted Descriptors with CNN-based Features for Facial Age Estimation
Age estimation from facial images is an important application of biometrics. In contrast to other facial variations like occlusions, illumination, misalignment and facial expressions, ageing variation is affected by human genes, environment, lifestyle and health which make age estimation a challenging task. In this thesis, we propose three new age estimation systems for automatic facial age estimation. These systems utilize different type of feature descriptors, varying from hand-crafted ones to automatically learned features and combine them in different level of information fusion. In the first proposed system, an integration of different feature extraction algorithms is utilized. This integration is performed by using two-level fusion of features and scores with the help of feature-level and score-level fusion techniques. In our proposed method, the advantage of using different types of features such as biologically-inspired features, texture-based features and appearance-based features is used. Feature-level fusion of biologically-inspired and texture-based methods is integrated into the proposed method and their combination is fused with an appearance-based method using score-level fusion. The second proposed system exploits multi-stage features from a trained Convolutional Neural Network (CNN), and precisely combines these features with a selection of age-related hand-crafted features. This method utilizes a decision-level fusion of estimated ages by two different approaches; the first one uses feature-level fusion of different hand-crafted local feature descriptors for wrinkle, skin and facial iv component while the second one uses score-level fusion of different feature layers of a CNN for age estimation. In the third system, we propose a new architecture of deep neural networks namely Directed Acyclic Graph Convolutional Neural Networks (DAG-CNNs) for age estimation which automatically combine multi-stage features from different layers of a CNN. This system is constructed by adding multi-scale output connections to an underlying backbone from two well-known deep learning architectures, namely VGG-16 and GoogLeNet. DAG-CNNs not only fuse the feature extraction and classification stages of the age estimation into a single automated learning procedure, but also utilize multi-scale features and perform score-level fusion of multiple classifiers automatically. Experiments on the publicly available Morph-II and FG-NET databases prove the effectiveness of our novel method.
Palmprint and Face Biometrics for Person Authentication Using Color Images
Multimodal biometric systems aim to improve the recognition accuracy by minimizing the limitations of unimodal systems. There exist many unimodal biometric traits used for person authentication such as face, iris, fingerprint, palmprint, voice, etc. In this thesis, a study on face and palmprint authentication using color images to construct an effective multimodal system is presented. Face recognition has been studied extensively in the biometrics community. Color face recognition approaches also exist in the literature in the last decades. On the other hand, automated palmprint identification using palmprint images has been extensively studied in the literature. However, most of the palmprint identification approaches exploit the gray-level images and there has been very little efforts to improve the palmprint identification using color information. Therefore, the use of color information from palmprint and face images using RGB, YCbCr and HSV color space representations are studied in this thesis. The experiments are conducted on publicly available color face and color palmprint databases. The results are presented on RGB, YCbCr and HSV color spaces. The effect of different color spaces on face and palmprint authentication is presented at the end of the thesis. Keywords: Multimodal biometric systems, unimodal biometric traits, face biometrics, palmprint biometrics, color spaces, gray-level images.
Fusion of Palmprint, Palm Vein and Dorsal Hand Vein for Personal Identification
Security is one of the major concerns of human beings in the 21st century. Many forensic and governmental sections now have trusted biometric systems to provide high levels of security for them. Lots of researchers have also worked on many different biometric modalities to both ensure the security and the convenience of the end-users. Nowadays, concerning the magnificent potentials of hand based biometrics, they are a trending choice for a wide range of applications since it is commonly accepted by the society and is not considered to be intrusive while it can offer plenty of features that are abundant to identify humans on a large scale. This thesis uses three different hand-based biometric modalities, namely palmprint, palm vein, and dorsal hand vein to create a secure, efficient, and accurate multimodal hand-based biometric system. Additionally, four different feature extraction methods, namely Principal Component Analysis (PCA), Local Binary Patterns (LBP), Scale Invariant Feature Transforms (SIFT) and Speeded-Up Robust Features (SURF), are exploited to perform person identification. Experiments are conducted on the CASIA palmprint database, Tongji palm vein database, and Bosphorus dorsal vein database. Unimodal and multimodal experimental results are presented on all databases. Moreover, we propose a new multimodal method on palmprint, palm vein, and dorsal hand vein biometrics employing Feature-Level Fusion and Decision-Level Fusion techniques. Finally, the results are presented on six different datasets obtained from the aforementioned palmprint, palm vein, and dorsal vein databases. Keywords: Person Identification, Biometrics, Palmprint Biometrics, Palm Vein Biometrics, Dorsal Vein Biometrics, Information Fusion.
Iris Anti-Spoofing Using Image Quality Measures
Spoof detection is a critical issue for the recognition of iris because it reduces the risk of forging iris recognition systems. The most relevant iris spoofing attacks reported in previous studies follows one of the three trends: photo attacks, contact-lens attacks or artificial-eye attacks. Spoofing attacks have prompted the biometric research community to learn more about the threat posed by these kinds of attacks on iris, fingerprint and face biometric systems. In this thesis, various Image Quality Assessment techniques to detect fake and real iris images presented to biometric systems were used. In this context, full reference image quality assessment measures such as Error Sensitivity Measures, Structural Similarity Measures and Information Theoretic Measures are implemented to distinguish fake and real iris images. Full-reference Image Quality Measures are also concatenated using feature-level fusion strategy. We propose to fuse twenty one full-reference image quality measures for iris anti-spoofing against print-attacks, contact-lens attacks and artificial-eye attacks. In order to evaluate the performance of the proposed iris anti-spoofing method using feature-level fusion of Image Quality Assessment techniques, two publicly available databases, namely CASIA and IIITD, were used. A comparative analysis of the performance of these Image Quality Assessment metrics is performed towards the completion of the thesis on various iris spoofing datasets of the aforementioned iris spoofing databases.
Distinguishing Identical Twins Using Facial Images and Various Feature Extractors
Recognizing identical twins is considered as one of the most critical challenges in biometric systems due to the shortage of uniqueness and distinction between the identical twins. The lack of discriminative features could be compensated using different sources of information. In this thesis, two different hybrid approaches using three biometric traits namely frontal face, profile face and ear are proposed and implemented to distinguish identical twins. The proposed strategies are particularly based on feature-level fusion, score-level fusion and decision-level fusion. Both proposed approaches are evaluated using identical twins and non-twins individuals. In the proposed method 1, frontal face is employed together with three feature extraction algorithms namely Principal Component Analysis, Histogram of Oriented Gradients and Local Binary Patterns. Fusion in this approach is conducted by all the aforementioned fusion techniques and different challenges are considered such as illumination, expression and ageing using ND-Twins-2009-2010 and FERET databases. The lowest Equal Error Rates of identical twins recognition that are achieved using the proposed method are 2.07% for natural expression, 0.0% for smiling expression and 2.2% for controlled illumination compared to 4.5%, 4.2% and 4.7% Equal Error Rates of the best state-of-the-art algorithm under the same conditions. On the other hand, symmetry challenge of profile face and ear is tested in the proposed approach 2 by using Local Binary Patterns, Local Phase Quantization and Binarized Statistical Image Features feature extraction algorithms. The samples of both sides of profile face and ear are extracted from ND-Twins-2009-2010 and UBEAR databases. In this approach, the extent of symmetry of left and right sides of each trait is measured in order to be used for recognition purposes. Finally, symmetry experiments using multimodal biometric traits are implemented and compared with our proposed approach which uses feature-level and score-level fusion. The maximum accuracies achieved are 75% for identical twins using ND-Twins-2009- 2010 database; moreover 88.04% and 79.89% for non-twins using ND-Twins-2009- 2010 and UBEAR databases, respectively. Keywords: identical twins, face recognition, ear recognition, score-level fusion, feature-level fusion, decision-level fusion, multimodal biometrics.
A comparative study on palmprint recognition
Palmprint recognition uses the palm of a person as a biometric for identifying or verifying the human beings. The palmprint contains a number of distinctive features such as principal lines, wrinkles, ridges and minutiae. Therefore, it is appropriate to use feature extraction techniques in order to extract line, texture, statistics and multiple representations. This thesis presents a comparative study on palmprint recognition using different approaches to extract palmprint features. Appearance-based approaches such as Principal Component Analysis, statistical approaches such as Local Binary Patterns, transform-based approaches such as Discrete Cosine Transform and other approaches such as Log-Gabor filters have been investigated and evaluated on PolyU palmprint database. The experimental results on both right and left palmprint databases demonstrate that Local Binary Patterns approach is a good texture descriptor which achieves the best recognition accuracy compared to other methods. Keywords: Log-Gabor, Discrete Cosine Transform, Local Binary Patterns, Principal Component Analysis
Impulse Noise Removal Using Unbiased Weighted Mean Filter
Digital imaging technology has provided countless opportunities for human visual applications and many scientific disciplines such as astronomy and microbiology. Digital images are subject to various noise due to environmental factors or faults in hardware. One type of noise — impulse noise — manifests itself with the highest or the lowest intensity value in the dynamic range during digitization process. Impulse noise involves high frequency components which are undesirable. Therefore, it is vital to restore contaminated digital images before utilizing them in various applications. In this thesis, we have investigated Nonlinear Fixed-Valued Impulse (salt-and-pepper) Noise removal methods. Restoration of a contaminated image is composed of two stages. These are noise detection and restoration. The performance of various state-ofthe- art impulse noise removal methods are empirically compared for these two stages. For detection, misclassification and false-alarm rates are used for objective measurement. Restoration capabilities are compared in terms of Peak Signal-to-Noise Ratio (PSNR), Structural Similarity (SSIM) and Mean Absolute Error (MAE). We have also identified a common problem among impulse noise removal methods, namely, spatial bias. Asymmetric distribution of corruption prevents an equal contribution of uncorrupted pixels in the filtering window from a spatial perspective, effectively yielding a biased estimation of the original intensity value. In order to eliminate spatial bias, we have proposed Unbiased Weighted Mean Filter (UWMF). UWMF eliminates spatial bias by recalibrating pixel weights based on the positional distribution of corrupted pixels in the filtering window. Recalibrated weights reflect the spatial properties of corruption and compensate the missing contribution. iii We have demonstrated that elimination of spatial bias improves restoration quality in terms of objective measurements (PSNR, SSIM and MAE). In addition, unbiased restoration results with least amount of disturbance in the edges and smooth regions. Keywords: Impulse Noise Removal, Nonlinear Filters,Weighted Mean Filters, Median Filters
Facial Age Classification Using Geometric Ratios and Wrinkle Analysis
Age group classification is the process of automatically determining an individual’s age range based on features extracted from facial image. It plays an important role in many real-life applications such as age specific human computer interaction, forensic art, access control and surveillance monitoring, person identification, data mining and organization, and cosmetology. In this thesis, we propose facial age classification approaches based on local and global descriptors extracted through feature selection methods. This thesis proposes two different methods on facial age classification. The first proposed method is a novel and efficient age group classification approach that combines holistic and local features extracted from facial images. These combined features are used to classify subjects into several age groups in two key stages. First, geometric features of each face are extracted to construct a global facial feature. Support Vector Classifier (SVC) is used to classify the facial images into several age groups using computed facial feature ratios. Then, local facial features are extracted utilizing subpattern-based Local Binary Patterns (LBP) to classify adults. These combined features are used to classify subjects into six major age groups. The superiority of subpattern-based LBP over Principal Component Analysis (PCA) and Subspace Linear Discriminant Analysis (subspace LDA)techniques is presented. The second proposed method presents geometric feature-based model for age group classification of facial images. The feature extraction is performed considering significance of the effects that age has on facial anthropometry. In this context, Particle Swarm Optimization (PSO) technique is used to find optimized subset of geometric features. Age Classification on these features is evaluated using SVC. Wrinkle feature analysis is also applied to classify adult images. The facial images are categorized into seven major age groups. The effectiveness and accuracy of the proposed age classification are demonstrated with the experiments that are conducted on two publicly available databases namely Face and Gesture Recognition Research Network (FGNET) and Iranian Face Database (IFDB). The experimental results show significant improvement of the proposed methods compared to the state-of-the-art models. Keywords: Age group classification, feature extraction, Local Binary Patterns, Particle Swarm Optimization.
Efficient Multimodal Biometric Systems Using Face and Palmprint
Multimodal biometric systems aim to improve the recognition accuracy by minimizing the limitations of unimodal systems. Fusion of two or more biometric modalities provides a robust recognition system against the distortions of individual modalities by combining the strengths of single biometrics. This thesis proposes different fusion approaches using two biometric systems namely face and palmprint biometrics. These fusion strategies are particularly based on feature level fusion and score level fusion. In this thesis, face and palmprint biometrics are employed to obtain a robust recognition system using different feature extraction methods, score normalization and different fusion techniques in three different proposed schemes. In order to extract face and palmprint features, local and global feature extractors are used separately on unimodal systems. Then fusion of the extracted features of these modalities is performed on different sets of face and palmprint databases. Local Binary Patterns (LBP) is used as a local feature extraction method to obtain efficient texture descriptors and then Log Gabor, Principal Component Analysis (PCA) and subspace Linear Discriminant Analysis (LDA) are used as global feature extraction methods. In order to increase the performance of multimodal recognition systems, feature selection is performed using Backtracking Search Algorithm (BSA) to select an optimal subset of face and palmprint features. Hence, computation time and feature dimension are considerably reduced while obtaining the higher level of performance. Then, match score level fusion and feature level fusion are performed to show the effectiveness and accuracy of the proposed methods. In score level fusion, face and palmprint scores are normalized using tanh normalization and matching scores are fused using Sum Rule method. The proposed approaches are evaluated on a developed virtual multimodal database combining FERET face and PolyU palmprint databases. In addition, a large database is composed by combining different face databases such as ORL, Essex and extended Yale-B database to evaluate the performance of the proposed method against the existing state-of-the-art methods. The results demonstrate a significant improvement compared with unimodal identifiers and the proposed approaches significantly outperform other face-palmprint multimodal systems. Furthermore, we propose an anti-spoofing approach which utilizes both texture-based methods and image quality assessments (IQA) in order to distinguish between real and fake biometric traits. In the proposed multi-attack protection method, well-known full-reference objective measurements are used to evaluate image quality including, Peak Signal-to-Noise Ratio (PSNR), Structural Similarity (SSIM), Mean Squared Error (MSE), Normalized Cross-Correlation (NXC), Maximum Difference (MD), Normalized Absolute Error (NAE) and Average Difference (AD). The three types of feature extraction approaches namely Local Binary Patterns (LBP), Difference of Gaussians (DoG) and Histograms of Oriented Gradients (HOG) are employed as texture-based methods to perform spoof detection in order to detect texture patterns such as print failures, and overall image blur to detect attacks. A palmprint spoof database made by printed palmprint photographs using the camera to evaluate the ability of different palmprint spoof detection algorithms was constructed. We present the results of both face and palmprint spoof detection methods using two public-domain face spoof databases (Idiap Research Institute’s PRINT-ATTACK and REPLAY-ATTACK databases) and our own palmprint spoof database. Keywords: multimodal biometrics, face recognition, palmprint recognition, feature level fusion, match score level fusion, Backtracking Search Algorithm, spoofing, face spoofing detection, palmprint spoofing detection, print-attack, replay-attack
Plant Disease Classification Using Texture-Based Methods through Leaf Images
Plant products have been a major source of food for animals, raw materials for industry and source of revenues to governments. In view of this, careful attention is needed for quality and quantity of plant products. Biotic and abiotic factors contribute immensely in hampering agricultural produce. In this research, computer vision techniques such as texture-based algorithms namely Histogram of Oriented Gradients (HOG), Local Binary Patterns (LBP) and Binarized Statistical Image Features (BSIF) are employed in plant disease identification and classification. Nine different popular plant species are used with symptoms on leaf images to extract features to develop a novel system. We propose an approach that employs Decision-Level Fusion which is used to incorporate different algorithms’ strengths for a robust and more accurate system. The proposed method is also compared with Scale Invariant Feature Transform (SIFT) and its derivatives such as Dense Scale Invariant Feature Transform (DSIFT) and Pyramid Histogram of Visual Words (PHOW). The experiments are conducted on PlantVillage database that includes healthy and infected plant leaf images of tomato, apple, cherry, corn, grape, peach, pepper, potato and strawberry plants. Consequently, the diverse nature of the database and the high accuracy of the proposed system show that Decision-Level Fusion of texture-based features extracted from plant leaves are good in detecting and classifying plant diseases. Keywords: Plant disease identification, Computer vision, Texture-based features, Decision-Level Fusion
Person Recognition Through Profiler Faces Using Ear Biometrics
Recent studies in biometric systems have shown that the ear biometric is a reliable biometric for human recognition and among a lot of biometric traits it has achieved satisfying results for human recognition. In this thesis, 2D ear recognition approach based on the fusion of ear and tragus (small outer part of ear) using score-level fusion strategy is proposed. An attempt to overcome the effect of challenges such as partial occlusion, pose variation and weak illumination is done since the accuracy of ear recognition may be reduced if one or more of these challenges are available. In this thesis, the effect of the aforementioned challenges is estimated separately, and many samples of ear that are affected by two different challenges at the same time are also considered. The tragus is used as a biometric trait because it is often free from occlusion; it also provides discriminative features even in different poses and illuminations. The features are extracted using Local Binary Patterns (LBP) and the evaluation has been done on four datasets, namely USTB-1, USTB-2, USTB-3 and UBEAR. It has been observed that the fusion of ear and tragus can improve the recognition performance compared to the use of ear or tragus systems individually. Experimental results show that the proposed approach 1 enhances the recognition rates by fusion of parts that are non-occluded such as tragus in cases of partial occlusion, pose variation and weak illumination. It is observed that the proposed approach 1 that uses score-level fusion strategy performs better than feature-level fusion methods. Additionally, the proposed approach 1 performs better than most of the state-of-the-art ear recognition systems. Experimental results on three datasets show that the proposed approach 1 is robust and effective since it gives better results than the other matching algorithms under different ear challenges. The maximum accuracies achieved are 100% (under partial occlusion), 97.4% (under weak illumination), 100% (under pose variation), 97.5% (under real occlusion) for USTB-set1, USTB-set2, USTB-set3, UBEAR database, respectively. On the other hand, this study aims to measure the efficiency of ear and profile face modalities in human recognition under identification and verification modes. In order to obtain a robust multimodal recognition system using different feature extraction methods, we propose to fuse these traits with all possible binary combinations of left ear, left profile face, right ear and right profile face. Fusion is implemented by score-level fusion and decision-level fusion techniques in the proposed approach 2. Additionally, feature-level fusion is used for comparison. All experiments in this approach are implemented on the UBEAR database. Local Binary Patterns, Local Phase Quantization and Binarized Statistical Image Features approaches are used for feature extraction process in proposed approach 2. Images under different challenge such as illumination variation, pose variation and blurring are tested. Ear and profile face images from UBEAR database are used in the experiments. The experimental results show that the proposed approach 2 is more accurate and reliable than using ear or profile face images separately. The performance of the proposed approach 2 in terms of recognition rate is 100%, and in terms of Equal Error Rates is 1.9%.
Color-Based Face Recognition with Different Color Spaces and Image Quality Assessment
Image quality is a critical issue for the recognition of faces because i.t reduces the risk of forging face recognition systems. The most relevant face spoofing attacks reported in previous studies follow o.ne of the three trends: mobile attacks, high-def attacks, or print attacks. Spoofing attacks have prompted the biometric research community to learn more about the threat posed by these kinds of attacks on many biometric traits such as face, fingerprint, iris, etc. In this thesis, various Image Quality Assessment techniques are used to detect image quality. Fake and real face images presented to biometric systems can also be detected by analyzing the image quality. In this context, No-Reference Image Quality Assessment measures such as Distortion Specific Measures (JQI, HLFI), Training Based Measure (BIQI) and Natural Scene Statistic Measure (NIQE) a.re implemented to analyze the quality of the face images. Three color spaces are employed to check the quality of images under various conditions. RGB, HSV and YCbCr color spaces are implemented for each of their channels separately and then the channel outputs are concatenated for each color space. The facial features are extracted using Principal Component Analysis (PCA), Local Binary Patterns (LBP) and Color Local Binary Patterns (ColorLBP) feature extraction methods for face recognition experiments. Moreover, we propose a general face recognition algorithm for low, medium and high quality face images. The experimental results are demonstrated on three publicly available face databases, namely Replay Attack, Faces94, and ColorFERET. Face recognition rates on all databases with all color spaces are presented using three aforementioned feature extraction methods. Finally, the proposed method results are demonstrated and compared with the existing systems. The experimental results are successful and encouraging for the proposed method. Keywords: Face recognition, color spaces, feature extraction, image quality assessment.
Animal Classification Using Appearance-Based, Model-Based and Texture-Based Methods
Animal recognition and their classification have become increasingly popular area in pattern recognition and computer vision. Distinguishing between images of different animals and different species as means of classifying animals is an easy task for humans. However, it is difficult to distinguish animal species automatically even in simple cases such as distinguishing cats and dogs. Animal bodies easily get disfigured, they can appear in images in a way that they self-occlude and often the background in such images could be very complex and noisy. Furthermore, just as all objects in an image, illumination may vary from image to image, the dimension and viewpoints may also differ. There has been attempts to carry out animal recognition from images but this problem has gained not enough attention. In the literature, there are some visual animal biometrics algorithms categorizing specific animal species such as zebra, elephant, chimpanzees, tiger, whales, pet animals like dogs, etc. Moreover, there are a few publicly available animal face databases including the facial images of different animals such as LHI Animal Faces, HiT and KTH Databases. In this thesis, we studied on different animal face images to classify different animal species. We implemented appearance-based, texture-based and model-based feature extraction methods to categorize animals from their faces and a comparative study is performed at the end of the thesis. Keywords: Animal classification, visual animal biometrics, appearance-based methods, model-based methods, texture-based methods.