Hasan Sakir Bilge

Adaptive multi-element synthetic aperture imaging with motion and phase aberration correction

Multi-element synthetic aperture techniques employing subaperture processing over successive firing steps can produce good image quality with simple front-end hardware but are susceptible to motion and phase aberration artifacts. We explore correlation processing using fully common spatial frequencies of overlapping subapertures to adapt beamforming for motion and phase aberrations. Signals derived from the subset of elements representing common spatial frequencies exhibit significantly higher correlation coefficients than those from signals computed using the entire subaperture. In addition, the correlation coefficient decreases linearly with subaperture separation for complete subaperture signals, but remains nearly constant with subaperture separation if only common spatial frequencies are used. Adaptive multi-element synthetic aperture imaging with correlation processing using fully common spatial frequencies is tested on experimental RF data acquired from a diffuse scattering phantom using a 3.5 MHz, 128-element transducer array. The results indicate that common spatial frequencies can be used efficiently for correlation processing to correct motion and phase aberration for adaptive multi-element synthetic aperture imaging.

Motion estimation using common spatial frequencies in synthetic aperture imaging

Correlation processing using fully common spatial frequencies of overlapping subapertures is explored for motion estimation between consecutive excitations in multi-element synthetic aperture imaging. Signals derived from the subset of elements representing common spatial frequencies exhibit significantly higher correlation coefficients than those from signals computed using the entire subaperture. In addition, the correlation coefficient decreases linearly with subaperture separation for complete subaperture signals, but remains nearly constant with subaperture separation if only common spatial frequencies are used. Correlation processing using fully common spatial frequencies is tested on experimental rf data acquired from a diffuse scattering phantom using a 3.5 MHz, 128-element transducer array. The results indicate that common spatial frequencies can be used for correlation processing to minimize motion artifacts in synthetic aperture imaging.

Face Recognition with Discriminating 3D DCT Coefficients

In this study, we investigate the use of a 3D discrete cosine transform (DCT) for 3D face recognition and present a novel 3D DCT-based feature extraction method with the selection of discriminating coefficients. We apply a 3D DCT on the voxel data, and use transform coefficients as features. Then the most discriminating 3D transform coefficients are selected with the proportion of variance, sequential floating forward selection and sequential floating backward selection methods. After feature selection, the linear discriminant analysis is applied on reduced sized feature vectors. We compare the results of different feature selection methods and show that a hybrid feature selection method has the best performance both in terms of time and recognition. Our experimental results verify that the discriminating DCT coefficients increase the face recognition rate more than the low-indexed coefficients do. On the other hand, the discriminating coefficients have only an energy level of 1.58%, too low when compared with the total energy of low-indexed coefficients. This fact shows that the discriminating coefficients are not the most energetic ones. With these coefficients, a recognition rate of 99.25% is achieved and this result is compared with other methods tested on a 3D RMA face database.

Content based image retrieval with sparse representations and local feature descriptors : A comparative study

Abstract Content Based Image Retrieval (CBIR) has been widely studied in the last two decades. Unlike text based image retrieval techniques, visual properties of images are used to obtain high level semantic information in CBIR. There is a gap between low level features and high level semantic information. This is called semantic gap and it is the most important problem in CBIR. The visual properties were extracted from low level features such as color, shape, texture and spatial information in early days. Local Feature Descriptors (LFDs) are more successful to increase performance of CBIR system. Then, a semantic bridge is built with high level semantic information. Sparse Representations (SRs) have become popular to achieve this aim in the last years. In this study, CBIR models that use LFDs and SRs in literature are investigated in detail. The SRs and LFD extraction algorithms are tested and compared within a CBIR framework for different scenarios. Scale Invariant Feature Transform (SIFT), Speeded-Up Robust Features (SURF), Histograms of Oriented Gradients (HoG), Local Binary Pattern (LBP) and Local Ternary Pattern (LTP) are used to extract LFDs from images. Random Features, K-Means and K-Singular Value Decomposition (K-SVD) algorithms are used for dictionary learning and Orthogonal Matching Pursuit (OMP), Homotopy, Lasso, Elastic Net, Parallel Coordinate Descent (PCD) and Separable Surrogate Function (SSF) are used for coefficient learning. Finally, three methods recently proposed in literature (Online Dictionary Learning (ODL), Locality-constrained Linear Coding (LLC) and Feature-based Sparse Representation (FBSR)) are also tested and compared with our framework results. All test results are presented and discussed. As a conclusion, the most successful approach in our framework is to use LLC for Coil20 data set and FBSR for Corel1000 data set. We obtain 89% and 58% Mean Average Precision (MAP) for Coil20 and Corel1000, respectively.

Feature extraction and discriminating feature selection for 3D face recognition

This paper presents a 3D face recognition method. In this method, 3D Discrete Cosine Transform (DCT) is used to extract features. Before the feature extraction, faces are aligned with respect to nose tip and then registered two times: according to average nose and average face. Then the coefficients of 3D transformation are calculated. The most discriminating 3D transform coefficients are selected as the feature vector where the ratio of between-class variance and within-class variance is used for discriminant coefficient selection. The results show that the most energetic features, low frequency components, are not the most discriminating features. The method was also modified based on 3D Discrete Fourier Transform (DFT) for feature selection as regarding real and complex DFT coefficients as independent features. Discriminating features were matched by using the Nearest Neighbor classifier. Recognition experiments were realized on 3D RMA face database. The proposed method yileds a recognition rate above 99% for 3D DCT based features.

3D Face Decomposition and Region Selection Against Expression Variations

3D face recognition exploits shape information as well as texture information in 2D systems. The use of whole 3D face is sensitive to some undesired situations like expression variations. To overcome this problem, we investigate a new approach that decomposes the whole 3D face into sub-regions and independently extracts features from each sub-region. 3D DCT is applied to each sub-region and most discriminating DCT coefficients are selected. The nose region gives the most contribution to the list of discriminating coefficients. Furthermore, a better recognition rate is achieved by only using the nose region. The highest recognition score in our experiments is 98.97% where rank-one recognition rates are considered. The results of the proposed approach are compared to other methods that use FRGC v2 database.

Subarray delta-sigma beamforming for ultrasonic imaging

We present a beamforming architecture based on subarray processing with non-uniform oversampling 1-bit delta-sigma (/spl Delta//spl Sigma/) modulation. The subarray processing combines conventional phased array and synthetic aperture approaches to form a large aperture using small subarrays thus reducing active channel count. /spl Delta//spl Sigma/-based beamforming improves the efficiency of front-end processing further: oversampling permits precise delaying and single bit data processing simplifies beamforming operation. To reduce the number of firings we use a low beam density associated with the subarray size, and then increase the beam density by lateral interpolation prior to coherent beam summation. Our experimental test results show that the proposed scheme provides high-resolution beamforming while simplifying the front-end.

Symmetry analysis for 2D images by using DCT coefficients

In this study, we proposed a new method to align symmetric signals by using symmetry property of Discrete Cosine Transform (DCT), which is widely used in signal compression and pattern recognition. For the symmetric signals, the energy is concentrated in the even indexed DCT coefficients. Using this property, we defined a symmetry measure. In this measure, ratio of the energy in even indexed coefficients to total energy gives the symmetry value for the signal. When the symmetry values of the rotated image at different angles become maximum, it means that the image is aligned according to its symmetry axis. We use 2D face data in experimental studies, because of its well-known symmetric property. This symmetry measure can also be adapted to any dimensional signals. The proposed method is tested on texture and shape data of the 3D face which is taken from FRGC database. Using the proposed method which exploits face symmetry, it is shown that the alignment resolution better than 1° can be achieved.

Face Recognition by Using 3D Discrete Cosine Transform

In this study, a new method is presented for face recognition by using DCT-faces that are calculated by the 3 dimensional discrete cosine transform (3D DCT). In this method, firstly, face images in the database are collected on top of each other and then 3D DCT is applied to this 3D structure. Then 2D Inverse DCT of each coefficient layer is calculated separately. After this operation, DCT-faces are found which are very similar to eigenfaces. By taking the inner product between DCT-faces and original face images, feature vectors are calculated to be used in face recognition. Experimental results show that, the face recognition ratio of DCT-faces is close to that of eigenface method. Moreover, the proposed method has a lower time complexity and lower memory complexity.

A new classification method by using Lorentzian distance metric

In this study, we propose a new algorithm which works in Lorentzian space with a similar sense in the k-NN method. We exploit the distance metric of Lorentzian space in classification problem. It is a special metric which may give a zero distance for far points. To take best benefit from structural and other properties of the Lorentzian space, a special projection over the data sets is applied. By this projection, basic geometrical operations are used; namely translation (shifting), compression and rotation. Our new algorithm does classification according to the nearest neighbor in Lorentzian space. The usability and validity of the proposed classification method is tested by some public data sets such as WHOLE, VERTEBRAL, RELAX, ECOLI. The results are compared with results of well-known classical classification methods such as kNN, LDA, SVM and Bayes. As a result, our proposed algorithm produces more successful results.