Ahmet Çinar

KINECT NOKTA BULUTU VERİSİNE METRİK TABANLI YENİ BİR SINIFLANDIRMA YAKLAŞIMI

3B nokta bulutu verileri ile nesne siniflandirma arastirma alaninda gelismekte calismalarinda en onemli sorunlardan biridir. Bu yazida 3B nokta bulutu verilerine gore gercek nesnelerin etkin sekilde tespiti icin yeni bir metriksel yontem onermekteyiz. Gercek nesnelerin algilanmasi ortalama kayma (mean-shift) siniflandirma algoritmasina dayali bir yontemle yapilmaktadir. Yontemin verimliligi karsilastirmali resmi 3B verileri ve gercek 3B nokta bulutu verileri ile dogrulanmaktadir. Nokta bulutu verileri ve metrik bilgilerin kombinasyonu bir yazilim cercevesinde uygulanarak siniflandirma asamasinin sonuclarini iyilestirilmektedir. Bu amacla, 3B nokta bulutu verilerinin siniflandirilmasi onemli olcude hata azaltilarak farkli nesneler icine saglam segmentasyon ve ozellik cikarimlari degerlendirilmistir. Ham veri uzerine uygulanan ortalama kayma algoritmasi ile metrik siniflandirma gerceklestirilmis veri uzerine ortalama kayma algoritmasi uygulamasi otomatik olarak karsilastirilarak metrik siniflandirma algoritmasinin dogrulugu degerlendirilmektedir. Sonuclar metrik siniflandirma algoritmasinin basit duzlemsel sekle sahip farkli nesnelerin otomatik olarak siniflandirilmasina verimli bir surece sahip oldugu gostermektedir

A method for local tuning of fuzzy membership functions

In this paper, a new method based on genetic algorithms is proposed for local tuning of fuzzy membership functions. For this purpose, the local adjustment is employed on the initial membership functions. Genetic algorithm is used to investigate discrete points that will be modified on the membership functions Hence, global adjustment does not require and the processing time required for tuning of membership functions is decreased.

Neural Networks Based Mesh Generation Method in 2-D

This paper describes a novel method for mesh generation in 2D by means of feed forward single layer neural networks. Original values of an initially given boundary are represented by finite values instead whole of points. For this aim, b-spline control points are made up to boundary curve. The obtained control points are used as inputs of the single layer feed forward neural network and points which belongs to closed area are obtained from output of neural networks. Obtained points are meshed by proposed method. As application, some mesh samples are given.

Image classification with caffe deep learning framework

Image classification is one of the important problems in the field of machine learning. Deep learning architectures are used in many machine learning applications such as image classification and object detection. The ability to manipulate large image clusters and implement them quickly makes deep learning a popular method in classifying images. This study points out the success of the convolutional neural networks which is the architecture of deep learning, in solving image classification problems. In the study, the convolutional neural network model of the winner of ilsvrc12 competition is implemented. The method distinguishes 1.2 million images with 1000 categories in success. The application is performed with the caffe library, and the image classification process is employed. In the application that uses the speed facility provided by GPU, the test operation is performed by using the images in Caltech-101 dataset.

Gabor wavelet and unsupervised Fuzzy C-means clustering for edge detection of medical images

It is well known that the Gabor wavelet transform (GWT) provides directional information for the analysis of an image. In this paper, we proposed an approach based on the GWT by combining unsupervised Fuzzy c-means (FCM) clustering which provides plays an important role in recognition as a classifier. After enhancing the edge of the input image using GWT, the binary image showing the edge is obtained using FCM clustering and morphological skeletonization. When compared to the Canny method and other conventional method, the proposed method has showed a better performance in terms of detection accuracy for noisy medical images.

Fuzzy blending of materials

In this paper, we propose a novel method for free form surface modeling based on fuzzy logic technique. The presented method is dependent on finding control points of surface. The main objective of method is that; in case of knowing start and final points, a surface is to construct by modeling curves. Firstly, each curve is modeled and then a surface is generated by joining the created curves. As an application, the method is applied to 3D blending and some blended objects are given. Application results show the effectiveness and applicability of the proposed method.

A Method Based on Fuzzy Logic Technique for Smoothing in 2D

This paper describes a novel approach based on fuzzy logic technique for smoothing process on the unstructured meshes in 2D. The proposed method works local on triangulation. Therefore, remeshing does not require. It is known that other smoothing operations such as laplacian smoothing , optimization based smoothing, angle based smoothing, and hybrid smoothing operations work global. That is, in these methods, meshing is achieved, afterwards remeshing is done for smoothing. Whereas, the goal of presented method is to find the best fit location of node while meshing is done. However, remeshing process is avoided.

Blending of 2D Objects Using Fuzzy Logic

In this paper, a method based on fuzzy logic for blending of 2D objects is presented. For this purpose, the curve modeling technique based on control points is used for curve generation where the control points are determined through fuzzy logic technique. As application, 2D object blending operations are implemented.

A GPU-based convolutional neural network approach for image classification

Deep learning obtains successful results in solving many machine learning problems. In this study, image classification process is performed by using Convolutional Neural Network (CNN) which is the most used architecture of deep learning. Image classification is used in a lot of basic field like medicine, education and security. Conditions that correct classification has vital importance may be especially in medicine field. Therefore, improved methods are needed in this issue. Although several algorithms for image classification have been developed over the years, they have not been used with the discovery of Convolutional Neural Networks. Convolutional Neural Networks provide better results than existing methods in the literature due to advantages such as processing by extracting hidden features, allowing parallel processing thanks to parallel structure, and real time operation. Furthermore, we use Convolutional Neural Networks in the proposed method. In this study, the image classification process is performed by using like a LeNet network model. The caffe library, which is often used for deep learning, is utilized. Our method is trained and tested with images of cats and dogs taken from the kaggle dataset. 10.000 tagged data is used for training and 5.000 unlabeled data is used for testing. Owing to Convolutional Neural Networks allow parallel processing, GPU technology has been used. In our method is used GPU technology and classification is evaluated with acceptable accuracy rate and speed performance.

NESNE KAYNAŞTIRMA İŞLEMİ İÇİN SİNİRSEL-BULANIK TEKNİĞİ TABANLI BİR YÖNTEM

Bu makalede, bilgisayar ortaminda nesne kaynastirma islemi icin sinirsel-bulanik mantik teknigini kullanan bir yontem sunulmustur. Klasik nesne kaynastirma yontemlerinde, kaynastirma islemi yapilirken birlesim bolgesi ve birlesim yuzeyi seklinin belli olmasi gerekir. Yapilan calismada kaynastirma isleminde gecis yuzeyi yapay sinir aglari kullanilarak uretilmis ve uretilen gecis yuzeyi uzerindeki kontrol, bulanik mantik teknigi yardimiyla saglanmistir. Bu sekilde tamamen kontrollu bir gecis yuzeyi elde edilmistir. Ileri beslemeli, egiticisiz yapay sinir agi modeli kullanilmis ve agin agirlik degisimi ayarlamasinda bulanik mantik teknigi kullanilmistir. Uygulama olarak yontem cesitli nesnelere uygulanmistir.