Search

Histogram equalization is one of the methods used to increase the contrast of digital images. In the ideal histogram equalization methods, the contrast are stretched by preserving the image similarity between input and output images, making the distributions in the histogram uniform. In this study conducted in the frequency domain, a new image contrast stretching method with Haar Wavelet Transform (HWT) in Low Dynamic Range is proposed. In this method, using the Probability Mass Function (PMF), a suppression process is applied to the noisy frequencies in the high pass channel of the frequencies. Subsequent frequency transformations provide a feasible reduction in the dynamic range of histogram frequencies. This process in the frequency domain improves the image quality by increasing the standard deviation in the image. In experimental studies over a benchmarking dataset, the proposed method is compared with conventional methods and promising results are obtained. In the experimental studies, Peak Signal to Noise Ratio (PSNR), Mean Square Error (MSE), Structural Similarity Index Measurement (SSIM) and Contrast value, which are among the image quality evaluation metrics, are used. In this proposed approach, image quality is evaluated both qualitatively and quantitatively assessments, and successful results are obtained., Histogram eşitleme dijital görüntülerin kontrastını artırmak için kullanılan yöntemlerden biridir. İdeal histogram eşitleme yöntemlerinde, girdi ve çıktı arasındaki görüntü benzerliğini koruyarak histogramdaki dağılımları tekdüze hale getirerek kontrast gerilmektedir. Frekans alanında yapılan bu çalışmada, Düşük Dinamik Aralığında değiştirilmiş ayrık Haar Dalgacık Dönüşümü ile yeni bir görüntü kontrast germe yöntemi önerilmiştir. Bu yöntemde Olasılık Kütle Fonksiyonunu ile frekansların yüksek geçiş kanalında gürültülü frekanslara bir baskılama işlemi gerçekleştirilmiştir. Daha sonra yapılan frekans dönüşümlerinde histogram frekansların dinamik aralıklarında önemli bir azalma sağlanmıştır. Frekans alanındaki bu işlem görüntüde standart sapmanın artmasını sağlayarak görüntü kalitesinin iyileşmesini sağlar. Kıyaslamalı bir veri seti üzerinde yapılan deneysel çalışmalarda, önerilen yöntem konvansiyonel metotlarla kıyaslanmış ve umut verici sonuçlar elde edilmiştir. Görüntü kalitesi değerlendirme metriklerinden Tepe Sinyal Gürültü Oranı (PSNR), Ortalama Kare Hata (MSE), Yapısal Benzerlik Endeks Ölçütü (SSIM) ve Kontrast değeri deneysel çalışmalarda kullanılmıştır. Önerilen bu yaklaşım ile elde edilen sonuçlar diğer algoritmaların sonuçları ile kıyaslandığında hem kalitatif hem de kantitatif açıdan başarılı bulunmuştur.

Bu çalışmada tam spektrum görüntü üzerinden ray kusurlarının belirlenmesine yönelik peridinamik tabanlı bulanık mantık algoritması geliştirilmiştir. Örnek uygulama için tam spektrum ray kusuru görüntüsünün, gri skala tanım kümesinde Gaussian maskelemesi gerçekleştirilmiş, görüntünün farklı boyut ve mertebelerden türev alma işlemleri peridinamik sayısal türev alma metodu kullanılarak elde edilmitşir. Alınan türevler bulanık mantık sisteminde değerlendiriltikten sonra görüntünün RGB ve HSV skalaları üzerinde üç temel dalga boyu dönüşümü yapılmıştır. Bulanık mantık çıktısı üzerine tatbik edilen en etkin üç dalga boyu ile ortak noktaların tespiti gerçekleştirilerek ray kusurlarının daha belirgin bir şekilde ortaya çıkması sağlanmıştır. Ayrıca üç farklı tam spektrum örnek ray kusur görüntüsü ile geliştirilen algoritma test edilip, literatürdeki mevcut kenar bulma algoritmalarına göre daha iyi sonuç verdiği tespit edilmiştir., In this study, a peridinamic based fuzzy logic algorithm has been developed for the detection of rail defects on a full spectrum image. For the sample application, Gaussian masking of the full spectrum ray defect image in the gray scale definition set was performed and the derivation operations of the image from different directions and orders were obtained by using the peridynamic numerical derivation method. Three different wavelength transformations were performed on the RGB and HSV scales of the image. By identifying the common points with the three most effective wavelengths applied on the fuzzy logic output, the rail defects are more pronounced. In addition, the algorithm developed with three different full spectrum sample ray defect images was tested and found to be better than the existing edge detection algorithms in the literature.

Modern teknolojik gelişmeler, görüntü, ses, video ve metin gibi multimedya verilerinin yasadışı kopyalarının üretilmesini çok kolay hale getirmiştir. Sayısal damgalama teknikleri, telif hakkı korumalı veri dosyalarının yasadışı kullanımını önlemek amacıyla telif hakkı bilgilerinin korunacak ortama sayısal olarak gömülmesini amaçlar. Bu çalışmada, biyometrik verilerle sayısal görüntünün damgalanması yöntemi ele alınmıştır. Geleneksel damga yerine biyometrik özelliklerin kullanımı damgalama şemasının güvenliğini artırır. Bu çalışmada daha güvenli olması ve kopyalanmasının daha zor olması sebebiyle iris biyometrik verisi tercih edilmiştir. Kişinin göz görüntüsünden üretilen iris şablonu ayrık kosinüs dönüşümü ile ikili koda dönüştürülür. Bu ikili kod, orijinal görüntüye yeniden dağıtılmış değişmez ayrık dalgacık dönüşümü ve Schur ayrıştırma yardımıyla gömülür. İkili kodun gömülmesi için kullanılan bloklar standart sapma değerlerine bağlı olarak seçilir. Deneysel sonuçlar önerilen yöntemin algısal kalitesinin yüksek olduğunu ortaya koymuştur. Ayrıca farklı görüntü işleme atakları ve geometrik ataklar karşısında dayanıklılığı oldukça yüksek bulunmuştur., Modern technological advances has made it very easy to produce illegal copies of multimedia data such as image, audio, video, or text. Digital watermarking techniques aim to digitally embed copyright information into the media to be protected in an attempt to prevent illegal use of copyright protected data files. In this study, the method of digital image watermarking with biometric data is discussed. The use of biometric features instead of traditional watermarks increases the security of the watermarking scheme. Iris biometric data is preferred in this study because it is more secured and more difficult to copy. The iris template produced from the eye image of the person is converted into binary code via discrete cosine transform. This binary code is embedded into the cover image with the help of the redistributed invariant discrete wavelet transform and Schur decomposition. The blocks used for embedding binary code are selected based on their standard deviation values. Experimental results revealed that the perceptual quality of the proposed method is high. In addition, it is found to be highly robust against different types of image processing attacks as well as geometrical attacks.Keywords: Biometric Watermarking, Redistributed Invariant Discrete Wavelet Transform, Schur Decomposition, Robustness

For the segmentation of brain vessels from MRA images, brain tissue is used in the head, eye, skull, etc. must be separated from the structures. For this reason, studies are carried out for the segmentation of brain tissue. In this study, the method that automatically segregates brain tissue from magnetic resonance angiography images taken with time of flight (TOF) technique is presented. The method in the study consists of five steps. First of all, the tip contrast values in the image are filtered by anisotropic diffusion filtering method. Parameters of anisotropic diffusion method are determined automatically by the natural image quality evaluator method. Sudden density transitions are detected by applying LoG edge detection filter on the filtered image. It is made ready for image analysis by applying etching on the image with density transitions. According to the conditions determined in image analysis, brain tissue is obtained separated from other head structures. As a result of this study, an easy-to-apply, fast-delivering, high-accuracy automatic algorithm has been introduced.

26th IEEE Signal Processing and Communications Applications Conference (2018 : İzmir; Türkiye), Due to the imperfect physical arrangement of camera sensors, spectral bands of ground observation satellite images are usually shifted relative to each other. In order to address this issue, we propose a computationally simple band registration method which is based on Dynamic Time Warping (DTW) and Discrete Wavelet Transform (DWT) algorithms. This method has been tested on 10 frames of GOKTURK-2 images and compared to a Scale-Invariant Feature Transform (SIFT) based method. In terms of quality, the proposed method have yielded very close results compared to SIFT. © 2018 IEEE., Yer gözlem uyduları tarafından çekilen görüntülerin spektral bantları arasında, mükemmel olmayan kamera sensör diziliminden kaynaklanan bazı kaymalar meydana gelebilmektedir. Bu çalışmada, bantlar arasındaki bu tip kaymaları düzeltmek için Dinamik Zaman Bükme (DZB) ve Ayrık Dalgacık Dönüşümü (ADD) algoritmalarına dayalı ve hesaplamsal olarak basit bir bant eşleştime yöntemi önerilmektedir. Bu yöntem 10 kare GÖKTÜRK-2 görüntüsü üzerinde test edilmiş ve Scale-Invariant Feature Transform (SIFT) tabanlı bir yöntemle karşılaştırılmıştır. Önerilen yöntemden elde edilen sonuçların SIFT’e oldukça yakın kalitede oldugu gözlemlenmiştir., Aselsan,et al.,Huawei,IEEE Signal Processing Society,IEEE Turkey Section,Netas

BirLP-kosimpletik manifoldunda kontak pseudo-slant altmanifoldların geodeziklikdurumları için yeni sonuçlar gösterildi. Bir alt manifoldun kontak pseudo-slantolması için gerek ve yeter şartlar verildi. Kontak pseudo-slant çarpımkarakterize edildi ve kontak pseudo-slant altmanifoldun kontak pseudo-slantçarpım olması için gerek ve yeter şartlar verildi. Ayrıca, konuyu açıklamak içinhemen hemen parakontak metrik manifold örneği incelendi., Newresults are shown for the geodesicsituations of contact pseudo-slantsubmanifolds in a LP-cosymplectic manifold. Necessary and sufficient conditionsfor a submanifold to be contact pseudo-slant are given. The contact pseudo-slant product ischaracterized and necessary and sufficient conditions for a contact pseudo-slant submanifold to be thepseudo-slant product is given. Also, an example of a contact pseudo-slantsubmanifold is investigate in an almost paracontact metric manifold to explainthe subject.

26th IEEE Signal Processing and Communications Applications Conference (SIU) -- MAY 02-05, 2018 -- Izmir, TURKEY WOS: 000511448500023 In this article, we have proposed a new hybrid clustering algorithm based on Fuzzy C Means (FCM) and Whale Optimization Algorithm (WOA) using chaotic map. The random selection of the initial cluster centers in the FCM algorithm is a disadvantage for the algorithm. To reduce this disadvantage, the BOA algorithm, which is improved performance with chaotic maps, is used to optimize the initial cluster centers. The proposed hybrid algorithm is tested for four different sets of data and better results are obtained from the FCM algorithm. IEEE, Huawei, Aselsan, NETAS, IEEE Turkey Sect, IEEE Signal Proc Soc, IEEE Commun Soc, ViSRATEK, Adresgezgini, Rohde & Schwarz, Integrated Syst & Syst Design, Atilim Univ, Havelsan, Izmir Katip Celebi Univ

26th IEEE Signal Processing and Communications Applications Conference (SIU) -- MAY 02-05, 2018 -- Izmir, TURKEY WOS: 000511448500024 In this work, we propose a hybrid clustering algorithm that integrates Fuzzy C-Means (FCM) and Whale Optimization Algorithm (WOA) using the Chebshev distance function. The FCM algorithm uses Euclidean distance to measure the similarity between the data. To avoid the existing disadvantages of the Euclidean distance, all distances in the FCM algorithm is calculated with the Chebsyhev distance function. The BOA algorithm is used to optimize the initial cluster centers. The proposed hybrid algorithm is tested with three different sets of data selected from UCI Machine Learning Repository database. As a result, it is seen that the clustering performance of the proposed algorithm is much better than the FCM algorithm. IEEE, Huawei, Aselsan, NETAS, IEEE Turkey Sect, IEEE Signal Proc Soc, IEEE Commun Soc, ViSRATEK, Adresgezgini, Rohde & Schwarz, Integrated Syst & Syst Design, Atilim Univ, Havelsan, Izmir Katip Celebi Univ

Date of Conference: 2-5 May 2018 Compressive spectral imaging is a rapidly growing area yielding higher performance novel spectral imagers than conventional ones. Inspired by compressed sensing theory, compressive spectral imagers aim to reconstruct the spectral images from compressive measurements using sparse signal recovery algorithms. In this paper, first, the image formation model and a sparsity-based reconstruction approach are presented for compressive photon-sieve spectral imager. Then the reconstruction performance of the approach is analyzed using different sparsity priors. In the system, a coded aperture is used for modulation and a photon-sieve for dispersion. In the measurements, coded and blurred images of spectral bands are superimposed. Simulation results show promising image reconstruction performance from these compressive measurements.

25th Signal Processing and Communications Applications Conference (SIU) -- MAY 15-18, 2017 -- Antalya, TURKEY, WOS: 000413813100056, In recent years, breast cancers have been diagnosed by several imaging modalities. One of these imaging methods is the diffuse optical tomography (DOT) system. In the presented study, in-vitro data were acquired from a breast phantom using the our designed DOT system. Breast phantoms were reconstructed in three dimensions (3D) using Truncated Conjugate Gradient (TCG) reconstruction algorithm. Then, image processing methods were applied to the raw images. Image segmentation technique was used to distinguish between "object" and "backgrounds". Later, 3D Gaussian filtering was applied to smooth the image and reduce noise. Finally, a three dimensional interpolation was applied to the noise free images in order to obtain a more realistic 3D image of the tissue phantoms. Thus, we have modified most popular image processing methods to apply the DOT images. As a result, we obtained user friendly, fast, accurate and user independent images processing methods for the DOT systems., Turk Telekom, Arcelik A S, Aselsan, ARGENIT, HAVELSAN, NETAS, Adresgezgini, IEEE Turkey Sect, AVCR Informat Technologies, Cisco, i2i Syst, Integrated Syst & Syst Design, ENOVAS, FiGES Engn, MS Spektral, Istanbul Teknik Univ

In this paper, Discrete-wavelet transform (DWT) based algorithm are used for speech signal denoising. Both hard and soft thresholding are used for denoising. Analysis are done on noisy speech signal corrupted by AGWN(Additive gaussian white noise) noise. Simulation and results are performed by using SignalLab application which developed for this research. MSE and SNR is used to compare results. Qt/C++ IDE is used to develop the application. This application developed for not only analyzing speech signal, but also can be used in an embedded system or a real-time applications. According to the obtained results, DWT is a good method for speech denoising when parameters are selected correctly.

24th Signal Processing and Communication Application Conference (SIU) -- MAY 16-19, 2016 -- Zonguldak, TURKEY, WOS: 000391250900110, In this paper, 11 different power quality disturbances were automatically detected by using statistical features with wavelet transform and norm entropy techniques. The best of the created features were selected with forward selection algorithm. Performance of classification algorithms, Support Vector Machines (SVM), Multi Layer Perceptron (MLP), k Nearest Neighbor (KNN) and random subspace KNN (Sub-KNN) which is an ensemble method, were examined. Consequently, the best classification accuracy of 99.3% was achieved by using Sub-KNN and it was appeared that compared to other methods, this algorithm was more robust against the noise., IEEE, Bulent Ecevit Univ, Dept Elect & Elect Engn, Bulent Ecevit Univ, Dept Biomed Engn, Bulent Ecevit Univ, Dept Comp Engn

24th Signal Processing and Communication Application Conference (SIU) -- MAY 16-19, 2016 -- Zonguldak, TURKEY, WOS: 000391250900327, Different filter methods to reduce noises occurred during image capture have been developed. One of the most effective image filters is diffusion filter. However, the major drawback of conventional diffusion filter is user-dependent. While noises are reduced with conductance coefficient arbitrarily selected, edge pixels are perceived such as noise. In this study, a novel anisotropic diffusion filter using similarity values obtained with the distance of each pixel to its neighbors has been proposed. Initially, similarity values of all image pixels are computed, and then they are used as conductance coefficients in diffusion filter. The mentioned value above is user dependent for conventional diffusion and it is constant for all pixels. On the other hand, it is made adaptive and eliminated user intervention with suggested approach. Developed method has been tested with different noise variances of images and experimental results have been compared with conventional diffusion filter., IEEE, Bulent Ecevit Univ, Dept Elect & Elect Engn, Bulent Ecevit Univ, Dept Biomed Engn, Bulent Ecevit Univ, Dept Comp Engn

This study proposes an improved method based on a nonlinear parametric model for intensity profile to increase the edge localization accuracy in medical images gathered from different imaging modalities. The edge model consists of three parameters associated with an edge point. It also takes into account local background intensity, noise, and blurring. The Marquardt-Levenberg algorithm is used to estimate the parameters because of its accuracy and good convergence rate. Performance of the proposed method is tested quantitatively by comparing the results with those of the well-known active contour method on synthetic vessel images. Qualitative comparisons on real MRI, coronary angiography, CT, ultrasound, and retinal images showed that the proposed method accurately estimates edges in medical images.

Histogram equalization is an extensively used method for image contrast adjustment using the image’s histogram. The method increases the contrast of the regions of an image that are represented by low contrast values. Through this adjustment, the image brightness values can be better distributed on the histogram. The moving low contrast regions to high contrasts after a gain process increases the visibility of unobtrusive regions and thus, allows better interpretation of the image. Histogram equalization method is used as a priority procedure in many disciplines because of its low computation time and it produces effective results. The method is frequently used as an aid to interpretation of the earth science image maps in recent years. In the frame of this study, the histogram equalization method is applied to theoretical and to a real archaeological total magnetic field datasets. In image maps, the method produces successful results in increasing the visibility of the low amplitude regions resulting from the structures that present deep or low magnetization characteristics, Histogram eşitleme, görüntü histogramını kullanarak görüntü karşıtlığının ayarlanması için yaygın olarak kullanılan bir yöntemdir. Yöntem, bir görüntüde düşük karşıtlık değerleriyle betimlenen bölümlerin karşıtlık değerlerini arttırır. Bu düzeltme sayesinde görüntüdeki parlaklık değerleri görüntü histogramı üzerinde daha iyi bir dağılım sergiler. Düşük yerel karşıtlıktaki bölümlerin bir kazanç işlemi sonrası yüksek zıtlık değerlerine taşınması göze çarpmayan bölgelerin görünürlüğünün artmasına ve böylelikle görüntünün daha iyi yorumlanmasına olanak sağlar. Histogram eşitleme yöntemi, hesaplama zamanının düşük olması ve etkili sonuçlar üretmesi nedeniyle birçok bilim dalında öncel bir işlem olarak kullanılmaktadır. Yöntem, son yıllarda yerbilimleri verilerine ait görüntü haritalarının yorumlanmasına yardımcı olarak sıklıkla kullanılmaktadır. Bu çalışma kapsamında, histogram eşitleme yöntemi kuramsal ve arkeolojik bir alanda toplanan toplam manyetik alan veri kümelerine uygulanmıştır. Görüntü haritalarında, derin veya düşük mıknatıslanma özelliği gösteren yapılardan kaynaklanan düşük genlikli bölgelerin görünürlüğünün arttırılmasında yöntemin başarılı sonuçlar verdiği görülmüştür

The probabilistic neural network (PNN) is a special type of radial basis neural network used mainly for classification problems. Due to the size of the network after training, this type of network is usually used for problems with a small-sized training dataset. In this paper, a new training algorithm is presented for use with large training databases. Application to the handwritten digit database shows that the reduced PNN performs better than the standard PNN for all of the studied cases with a big gain in size and processing speed. This new type of neural network can be used easily for problems with large training databases like biometrics and data mining applications. An extension of the network is possible for new training samples and/or classes without retraining.

In the present paper, we study hemi-slant submanifolds of a locally product Riemannian manifold. We prove that the anti-invariant distribution involved in the definition of hemi-slant submanifold is integrable and give some applications of this result. We get a necessary and sufficient condition for a proper hemi-slant submanifold to be a hemi-slant product. We also study these types of submanifolds with parallel canonical structures. Moreover, we give two characterization theorems for the totally umbilical proper hemi-slant submanifolds. Finally, we obtain a basic inequality involving Ricci curvature and the squared mean curvature of a hemi-slant submanifold of a certain type of locally product Riemannian manifolds.

Date of Conference: 16-19 May 2015 Conference Name: 23nd Signal Processing and Communications Applications Conference, SIU 2015 A new multi-resolution super-pixel based algorithm is proposed to track cell size, count and motion in Mesenchymal Stem Cells (MSCs) images. Multi-resolution super-pixels are obtained by placing varying density seeds on the image. The density of the seeds are determined according to the local high frequency components of the MSCs image. In this way a multi-resolution super-pixels decomposition of the image is obtained. A second contribution of the paper is novel decision rule for merging similar neighboring super-pixels. An algorithm based on well known wavelet decomposition is developed and applied to the histograms of neighboring super pixels to exploit similarity. The proposed algorithm is experimentally shown to be successful in segmenting and tracking cells in MSCs images.

In this study, a filter which uses the Gaussian function and selects its variance according to local properties of image without user intervention has been designed in order to eliminate the noise. Consequently, a gradient adaptive image filter that estimates variance and creates different kernel for each pixel in image has been obtained.

Medical Technologies National Conference (TIPTEKNO) -- OCT 15-18, 2015 -- Bodrum, TURKEY WOS: 000380505200014 Image clustering, the separation process of the image to the clusters which have pixels with similar properties is one of the most important sections of the image processing procedures. In this study we developed a novel image clustering algorithm by combining Differential Search algorithm (DS) which is a new optimization algorithm and Fuzzy Clustering Algorithm (FCM). Combining procedure was realized by minimizing the objective function of FCM by DS. As for testing purposes 3 brain MRI images were clustered by the new algorithm and the classical FCM. Both of the algorithms were compared in terms of objective function values. According to the results, the new algorithm out performs the classical FCM algorithm in terms of image clustering.

Image denoising is an important task in image processing. Various filtering techniques has been developed for image denoising but the algorithms have specific limitations. The main feature of a good image denoising technique is that it will remove noise while preserving structure of image. In this paper, we present a new image denoising method based on averaging. The denoising performance is measured according to PSNR (Peak Signal-to-Nosie Ratio) metric by employing experiments. Experimental results show that the proposed method removes noise significantly and preserve image edges.

In this presented study, it is compared that the wavelet and wavelet package decomposition for image denoising. The results shows that increasing of decomposition depth does not affect significantly PSNR rate. In the examination made on standart test images, it is concluded that although it does not show the same performance for every image, wavelet decomposition is superior to wavelet package decomposition. It is decided that it is more suitable to use denoising method based on wavelet decomposition when considering processing load as well., Sunulan bu çaly?mada, imgelerden gürültü temizlemesi için dalgacyk ve dalgacyk paket ayry?tyrma yöntemleri kar?yla?tyrylmy?tyr. Sonuçlar, her iki yöntem için de ayry?tyrma derinli?inin artyrylmasynyn PSNR oranyny ciddi bir ?ekilde etkilemedi?ini göstermi?tir. Standart test imgeleri üzerinde yapylan incelemede, her imge için her ne kadar ayny performansy gösteremese de dalgacyk paket ayry?tyrmasyna göre üstün oldu?u sonucuna varylmy?tyr. Y?lem yükü de dikkate alyndy?ynda, dalgacyk ayry?tyrmasyna dayaly gürültüden temizleme yönteminin kullanylmasynyn daha do?ru olaca?y sonucu çykarylmy?tyr.

This paper present a study on medical image retrieval method based on Gabor filters. After filtering the images using Gabor wavelet, texture features vectors are constructed by computing the mean and the standard deviation of the filtered image. The system construct the feature vector, computes the distance between the query image to all images in the database, through the comparison of their feature vectors, and retrieve then most similar images regarding this kind of feature vector. The results are promising that the retrieval efficiency is higher than %85 in general. The focus of this paper is to show the utility of Gabor wavelet transform on medical image characterization and its suability for image indexing and retrieval., Oldukça yararly doku analizi sa?layan Gabor dalgacyk dönü?ümüne, görüntü i?leme çaly?malarynda yaygyn bir ?ekilde ba?vurulmaktadyr. Bu makale, Gabor filtrelerine dayanan medikal görüntü eri?im metodu üzerine bir çaly?ma sunmaktadyr. Gabor dalgacy?yny kullanarak görüntüleri filtreledikten sonra filtrelenmi? görüntünün ortalamasy ve standart sapmasy hesaplanarak doku öznitelik vektörleri olu?turulur. Sistem; öznitelik vektörünü olu?turur, özniteliklerinin kar?yla?tyrylmasy yoluyla sorgu görüntüsünün veri tabanyndaki bütün görüntülere olan uzakly?yny hesaplar ve bu tür öznitelik vektörü ile ilgili en benzer görüntülere eri?ir. Genel olarak medikal görüntüler üzerinde çaly?yldy?y için Gabor dalgacy?ynyn yöneliminin, öznitelikler üzerinde kayda de?er bir etkisi yoktur. Eri?imin verimlili?inin genel olarak %85'den fazla olmasy ümit vericidir. Bu çaly?manyn odak noktasy, gabor dalgacyk dönü?ümünün medikal görüntüler üzerindeki yararlyly?yny ve görüntü dizinleme ve eri?imine uygunlu?unu göstermektir.

In this work, face recognition performances of eigenface and Fisherface algorithms in different face subspaces obtained from Discrete Cosine Transform (DCT) coefficients and original images were evaluated. DCT coefficients obtained in block structure were classified into subbands according to their energy levels. Eigenface and Fisherface subspaces were determined for four subbands which have the highest energy. Best recognition rate was achieved by the DC coefficients. Face recognition performances of the faces with Gaussian and salt&paper noises with different variance and intensity were tested. Best results were achieved by the horizontal and vertical medium frequency coefficient-subbands, Bu çalışmada öz yüz ve Fisher yüz algoritmaları ile hem Ayrık Kosinüs Dönüşümü (AKD) alt bant katsayılarından hem de orijinal görüntülerden elde edilen farklı yüz alt uzaylarında yüz tanıma başarımları ölçülmüştür. Blok olarak elde edilen AKD katsayıları enerji düzeylerine göre alt bantlara sınıflandırılmış ve enerji düzeyi en yüksek dört alt bant için öz yüz ve Fisher yüz alt uzayları oluşturulmuştur. En iyi tanıma sonucu AKD DC bileşenleri ile yapılan testlerde elde edilmiştir. Görüntülere Gauss ve tuz&biber gürültüleri belirli varyans ve yoğunluklarla eklenmiş ve yüz tanıma performansları ölçülmüştür. Bu testlerde en iyi tanıma sonucu AKD yatay ve düşey orta frekans bileşenleri ile elde edilmiştir.

We describe our efforts to determine the most important track registration parameters for the suitable use of CN-85 track detectors in the registration of low energy protons.

A large number of approaches for texture analysis have been suggested for the purpose of texture classification. Recently, wavelet frames were proposed for texture features extraction. In this study, non-subsampled wavelet frame transform was used for feature extraction of 16 textures from a set of Brodatz' album by means of various wavelet families. Texture classification was accomplished by artificial neural network with a fast adaptive backpropagation algorithm. A new pyramidal-windowing algorithm is proposed, which forms randomly rotated texture windows of variable sizes texture windows for training a neural networks classifier, and perfect classification results were obtained.

In this article, we establish inequalities between the Ricci curvature and the squared mean curvature, and also between the k-Ricci curvature and the scalar curvature for a slant, semi-slant and bi-slant submanifold in a cosymplectic space form of constant j-sectional curvature with arbitrary codimension.

In this paper, we introduce the notion of a slant lightlike submanifold of an indefinite Kenmotsu manifold. We provide a non-trivial example and obtain necessary and sufficient conditions for the existence of a slant lightlike submanifold. Also, we give an example of a minimal slant lightlike submanifold of R29 and prove some characterization theorems.

In this paper, a new face recognition technique is introduced based on the gray-level co-occurrence matrix (GLCM). GLCM represents the distributions of the intensities and the information about relative positions of neighboring pixels of an image. We proposed two methods to extract feature vectors using GLCM for face classification. The first method extracts the well-known Haralick features from the GLCM, and the second method directly uses GLCM by converting the matrix into a vector that can be used in the classification process. The results demonstrate that the second method, which uses GLCM directly, is superior to the first method that uses the feature vector containing the statistical Haralick features in both nearest neighbor and neural networks classifiers. The proposed GLCM based face recognition system not only outperforms well-known techniques such as principal component analysis and linear discriminant analysis, but also has comparable performance with local binary patterns and Gabor wavelets.

In this paper we introduce the notions of semi-slant and bi-slant submanifolds of an almost contact 3-structure manifold. We give some examples and characterization theorems about these submanifolds. Moreover, the distributions of semi-slant submanifolds of 3-cosymplectic and 3-Sasakian manifolds are studied. ϕi)i=1;2;3 and the vector elds should be slant or invariant with respect to all of the ϕi's. Therefore, it is a generalization of invariant, anti-invariant, slant, semi-slant, and bi-slant submanifolds in almost contact metric 3-structures and we denote them by 3- semi-slant and 3-bi-slant submanifolds. Following the approaches of (3, 13), we characterized 3-bi-slant and 3-semi-slant submanifolds and studied geometric properties of distributions of these submanifolds where the ambient manifolds are 3-Sasakian or 3-cosymplectic. It should be noted that, in the denition of semi-slant

In this paper some B.-Y. Chen inequalities for slant submanifolds in quaternionic space forms are established.

An interesting class of submanifolds of almost Hermitian manifolds (\tilde M,\tilde g,J) is the class of slant submanifolds. Slant submanifolds were introduced by the first author in [6] as submanifolds M of (\tilde M,\tilde g,J) such that, for any nonzero vector X \in TpM, p \in M, the angle q(X) between JX and the tangent space TpM is independent of the choice of p\in M and X \in TpM. The first results on slant submanifolds were summarized in the book [7]. Since then slant submanifolds have been studied by many geometers. Many nice results on slant submanifolds have been obtained during the last two decades. The main purpose of this paper is to study pointwise slant submanifolds in almost Hermitian manifolds which extends slant submanifolds in a very natural way. Several basic results in this respect are proved in this paper.

In this article, a software for image indexing and retrieval is presented. The classification proposed here is based on the dominant color(s) of the images. The process consists in assigning a colorimetric profile to the image in HLS space (Hue, Lightness, Saturation). First, the definition of hue is done thanks to a fuzzy representation to take into account the non-uniformity of colors distribution. And then, lightness and saturation are represented through linguistic qualifiers also defined in a fuzzy way. Finally, the profile is built through fuzzy functions representing the membership degree of the image to different classes. In order to improve the performances i.e. to define more accurate profiles, we propose to consider zones of pixels, instead of pixels individually. Those zones can be constructed thanks to an edge detection algorithm. A sample of pixels is chosen inside a zone to determine the color of the zone. According to the detected dominant colors, such a software may be used to classify indoor/outdoor images for example, or harmonious/disharmonious images...

Date of Conference: 24-26 April 2013 In this article, image histogram thresholding is carried out using the likelihood of a mixture of Gaussians. In the proposed approach, a prob ability density function (PDF) of the histogram is computed using Gaussian kernel density estimation in an iterative manner. The threshold is found by iteratively computing a mixture of Gaussians for the two clusters. This process is aborted when the current bin is assigned to a different cluster than its predecessor. The method does not envolve an exhaustive search. Visual examples of our segmentation versus Otsu's thresholding method are presented. © 2013 IEEE.

Date of Conference: 20-22 April 2011 Conference Name: 19th Signal Processing and Communications Applications Conference, SIU 2011 In this paper, a new histogram based auto-focusing method for thermal cameras is proposed. This proposed method is realized by FPGA (Field Programmable Gate Array) and DSP (Digital Signal Processor) working together and simultaneously. HF (High Frequency) component, obtained from real-time image flow by FPGA and DSP is used for auto-focusing process. Proposed method is able to determine the focus direction from the HF component produced in the process of histogram equalization by FPGA, unlike Fourier transform and pixel differenve based methods in the literature. With this superiority, proposed method requires no extra calculation for thermal cameras for which histogram equalization is necessary. Analysis show that proposed method is successful on the simulations and scanning thermal cameras.

In this paper, a new color transform for image compression is introduced. Weights of the color transform are determined using the histogram of an image, making it image-specific. The compression efficiency of the transform is demonstrated using the JPEG image coding scheme. The suggested transformation results in better PSNR values than original JPEG for a given compression level when tested on 15 commonly used test images.

Local invariants are powerful tools to recognize objects in an image under transformations. In this study, we propose a novel method to detect residential regions by using local SIFT (Scale Invariant Feature Transform) features in satellite images. For this purpose, we construct graphs using image features as nodes and their neighborhoods as edges. We then match the extracted graphs with the ones in our database. Our method is able to detect various residential regions. ©2008 IEEE. 2008 IEEE 16th Signal Processing, Communication and Applications Conference, SIU -- 20 April 2008 through 22 April 2008 -- Aydin -- 74111

In this paper, an automatic segmentation algorithm for color image was presented. The similarity percents of neighboring pixels were calculated by using mathematical approximation based on color similarity measurement. In the proposed method does not require any prior knowledge of the number of regions existing in the image. The methodology uses a rule based segmentation concept to reduce the computational burden required for the other methods. The simulation results indicate that the proposed algorithm yields segmented color scale image of perfect accuracy and the required computer time is reasonable, Bu makalede, renkli görüntüler için geliştirilen otomatik ayrıştırma algoritması sunulmuştur. Komşu piksellerin benzerlik yüzdeleri renk benzerlik ölçümü tabanlı matematiksel yaklaşım kullanılarak hesaplanmıştır. Önerilen yöntem, görüntüde muhtemel var olan bölge sayısı hakkında herhangi bir ön bilgiye ihtiyaç duymamaktadır. Böylece diğer metotlara göre ihtiyaç duyulan hesaplama yükü azaltılmıştır. Yapılan benzetim ve uygulamalarla önerilen yöntemin performansı test edilmiştir.

In this study, the linear and kernelized temporal predictability methods which have been used to separate audio mixtures have been applied to blind source separation of linear and nonlinear image mixtures and the statistical performance of the algorithms have been simulated using different algorithm parameters like kernel degree, kernel type, image mass size and nonlinear mixing function. The data sets are chosen to simulate the effect or corruption of images with writing or scribbles.

Date of Conference: 17-19 April 2006 We propose a new Set Partitioning In Hierarchical Trees (SPIHT) based mesh compression framework. The 3D mesh is first transformed to 2D images on a regular grid structure. Then, this image-like representation is wavelet transformed and SPIHT is applied on the wavelet domain data. The method is progressive because the resolution of the reconstructed mesh can be changed by varying the length of the one-dimensional data stream created by SPIHT algorithm. Nearly perfect reconstruction is possible if all of the data stream is received. © 2006 IEEE.

Bu çalışmada, yapay sinir ağları (YSA) destekli zeki bir otomatik parmakizi tanıma sistemi başarıyla geliştirilmiş ve sunulmuştur. Sistem tasarlanırken işlemler adım adım yapılmıştır. Öncelikle bir parmakizi okuyucu yardımıyla alınan parmakizi resimleri sayısala çevrilmiştir. Resimler küçük parçalara bölünerek üzerinde işlem yapılacak alan arkaplandan ayrılmıştır. Gri seviye resimlerden referans noktalar elde edilmiştir. Parmakizi temizleme ve iyileştirme için YSA modeli geliştirilmiş, iyi sonuç veren momentumlu geriyayılım öğrenme algoritması kullanılarak bu model eğitilmiştir. Temizlenip iyileştirilen resimlere bölgesel ikili dönüşüm uygulanmış ve daha sonra siyah beyaz renkten oluşan ikili resim inceltilmiştir. İnceltilmiş resim üzerinde özellik noktaları olarak adlandırılan uç ve çatal noktalar ve bunlarla ilgili gerekli parametreler bulunmuş ve yalancı özellik noktaları elenmiştir. Son olarak karşılaştırma algoritması belirlenip karşılaştırma işlemi yapılmıştır. Sunulan çalışmada, belirtilen tüm adımlar başarıyla tamamlanmış ve bu işlemlerin kolaylıkla yapılabilmesi için Delphi programlama ortamında bir yazılım geliştirilmiştir. Hem tanıma, hem de onaylama/doğrulama modunda çalışabilen sistem, 100 parmakizi resminin bulunduğu bir veritabanında test edilmiş ve başarılı sonuçlar elde edilmiştir., This work presents an intelligent automatic fingerprint identification and verification system based on Artificial Neural Networks (ANNs). In this work, the design processes of the system have been presented step by step. Fingerprints were first converted into digital images using a specific hardware. They were then processed by a computer. Fingerprint images were divided into grid blocks, and these blocks were classified as image area and background. An effective algorithm was used to detect the fingerprint singularities from gray level fingerprint images. In order to improve the performance of the system, fingerprint image enhancement was performed by using ANN. The adaptive backpropagation with momentum learning algorithm was used to train the ANN models. Binary images were obtained from the enhancement images using a regional binarization algorithm. Binary images were converted to thinned images. Ridge endings and ridge bifurcations of the fingerprints (minutiae) were extracted. A postprocessing algorithm was used to eliminate false minutiae patterns and the fingerprint matching process was finally applied. In order to automatise the system, a software for fingerprint identification and verification was developed in Delphi. The system developed in this work was tested 100 fingerprint images for identification and verification; it achieves the task with high accuracy. It is assumed that the developed system can be used in many security applications.

In this paper we have applied 2D Gabor filtering to gravity and magnetic anomalies in estimation of discontinuities. Gabor filtering is an effective separation method compared to others having steerable and frequency parameter properties. We have found new faults using Gabor filtering for gravity and magnetic anomalies of Marmara Sea. © 2005 IEEE.

Bu çalışmada, katsayıları F cisminde iki değişkenli halkalar üzerinde cyclic kodların görüntüleri tespit edilmiştir. Belirli koşullar altında iki değişkenli özel bir halka tanımlanmıştır. Bu halkalar üzerinde cyclic kodların Gray görüntüleri incelenmiştir. Gray dönüşümü yoluyla bu halkalar üzerindeki kodlar ve bir değişkenli sonlu zincir halkası üzerindeki kodlar elde edilmiştir. Sonlu zincir halkasından sonlu bir cisme başka Gray dönüşümü tanımlanmış ve sonra cyclic kodların görüntüleri elde edilmiştir. n uzunluğundaki R üzerinde bir cyclic kodun Gray görüntüsü bulunmuştur, In this study, images of cyclic codes in two variable rings with coefficient field F are detected. p are detected. A special ring in two variables is defined under certain conditions. The Gray images of the cyclic codes over this ring are investigated. Relations between the codes over this ring and the codes over a finite chain ring in one variable are obtained via a Gray map. Another Gray map from the finite chain ring to a finite field is defined and then the images of cyclic codes are obtained.It is obtained that the Gray image of a cyclic code over R with length n