8 sonuçlar
Arama Sonuçları
Listeleniyor 1 - 8 / 8
Yayın GIS aided vulnerability assessment for roads(Springer Science and Business Media B.V., 2022-04-21) Çalışkan, Berna; Atahan, Ali Osman; Kesten, Ali SercanRoad networks are vulnerable to natural disasters such as floods, earthquakes and forest fires which can adversely affect the travel on the network. However, not all road links equally affect the travel conditions in a given network; typically some links are more critical to the network functioning than the others. The first stage of study involves the investigation of geological conditions. Image classification used for extracting information classes from ‘Geological Map of Istanbul area’ image file. The resulting raster layer used to create thematic map. A reclassification was performed for lithologic types. The second stage involves analyzing topological situation. A slope map prepared and classified according to percentage of slope values. The third phase is the analysis and interpretation of the accumulated data to establish suitable and applicable road vulnerability scores. The information in the source data for each vulnerability factor are classified into three different vulnerability scores: +2 (considerably increases vulnerability), +1 (increases vulnerability) and 0 (does not increase vulnerability) by using a vulnerability score table. The study area was categorized into three different traffic analysis zones as: (1) least favorable area; (2) favorable area; (3) most favorable area. Vulnerability values obtained to measure serviceability of critical links in dense urban road networks and applies them to the case of ‘Beyoğlu’ region. Thematic layers were prepared using the Geographic Information System (GIS), and they were then combined to produce the serviceability of road links in the ‘Beyoğlu’ region. Consequently, A site specific vulnerability index is proposed, considering the serviceability of road links. A conceptual flowchart of the GIS processing steps taken to obtain the vulnerability index is illustrated.Yayın A haar classifier based call number detection and counting method for library books(IEEE, 2018-12-06) Kanburoğlu, Ali Buğra; Tek, Faik BorayCounting and organization of books in libraries is a routine and time-consuming task The task gets more complicated by misplaced books in shelves. In order to solve these problems, we propose an automated visual call number (book-id) detection and counting system in this paper. The method employs a Haar feature-based classifier from OpenCV library and cloud-based OCR system to decode characters from images. To develop and test the method, we have acquired and organized a dataset of 1000 book call numbers. The proposed method has been tested on 20 bookshelves images that contain 233 call numbers, which resulted in a true detection rate of 96% and false detection rate of 1.75 per image. For OCR step, the number of false recognized characters per call number was 0.76.Yayın Compression of the biomedical images using quadtree-based partitioned universally classified energy and pattern blocks(Springer London, 2019-03-15) Gezer, Murat; Gargari, Sepideh Nahavandi; Güz, Ümit; Gürkan, HakanIn this work, an efficient low bit rate image coding/compression method based on the quadtree-based partitioned universally classified energy and pattern building blocks (QB-UCEPB) is introduced. The proposed method combines low bit rate robustness and variable-sized quantization benefits of the well-known classified energy and pattern blocks (CEPB) method and quadtree-based (QB) partitioning technique, respectively. In the new method, first, the QB-UCEPB is constructed in the form of variable length block size thanks to the quadtree-based partitioning rather than fixed block size partitioning which was employed in the conventional CEPB method. The QB-UCEPB is then placed to the transmitter side as well as receiver side of the communication channel as a universal codebook manner. Every quadtree-based partitioned block of the input image is encoded using three quantities: image block scaling coefficient, the index number of the QB-UCEB and the index number of the QB-UCPB. These quantities are sent from the transmitter part to the receiver part through the communication channel. Then, the quadtree-based partitioned input image blocks are reconstructed in the receiver part using a decoding algorithm, which exploits the mathematical model that is proposed. Experimental results show that using the new method, the computational complexity of the classical CEPB is substantially reduced. Furthermore, higher compression ratios, PSNR and SSIM levels are achieved even at low bit rates compared to the classical CEPB and conventional methods such as SPIHT, EZW and JPEG2000Yayın Convolutional attention network for MRI-based Alzheimer's disease classification and its interpretability analysis(IEEE, 2021-09-17) Türkan, Yasemin; Tek, Faik BorayNeuroimaging techniques, such as Magnetic Resonance Imaging (MRI) and Positron Emission Tomography (PET), help to identify Alzheimer's disease (AD). These techniques generate large-scale, high-dimensional, multimodal neuroimaging data, which is time-consuming and difficult to interpret and classify. Therefore, interest in deep learning approaches for the classification of 3D structural MRI brain scans has grown rapidly. In this research study, we improved the 3D VGG model proposed by Korolev et al. [2]. We increased the filters in the 3D convolutional layers and then added an attention mechanism for better classification. We compared the performance of the proposed approaches for the classification of Alzheimer's disease versus mild cognitive impairments and normal cohorts on the Alzheimer's Disease Neuroimaging Initiative (ADNI) dataset. We observed that both the accuracy and area under curve results improved with the proposed models. However, deep neural networks are black boxes that produce predictions that require further explanation for medical usage. We compared the 3D-data interpretation capabilities of the proposed models using four different interpretability methods: Occlusion, 3D Ultrametric Contour Map, 3D Gradient-Weighted Class Activation Mapping, and SHapley Additive explanations (SHAP). We observed that explanation results differed in different network models and data classes.Yayın Adaptive convolution kernel for artificial neural networks(Academic Press Inc., 2021-02) Tek, Faik Boray; Çam, İlker; Karlı, DenizMany deep neural networks are built by using stacked convolutional layers of fixed and single size (often 3 × 3) kernels. This paper describes a method for learning the size of convolutional kernels to provide varying size kernels in a single layer. The method utilizes a differentiable, and therefore backpropagation-trainable Gaussian envelope which can grow or shrink in a base grid. Our experiments compared the proposed adaptive layers to ordinary convolution layers in a simple two-layer network, a deeper residual network, and a U-Net architecture. The results in the popular image classification datasets such as MNIST, MNIST-CLUTTERED, CIFAR-10, Fashion, and ‘‘Faces in the Wild’’ showed that the adaptive kernels can provide statistically significant improvements on ordinary convolution kernels. A segmentation experiment in the Oxford-Pets dataset demonstrated that replacing ordinary convolution layers in a U-shaped network with 7 × 7 adaptive layers can improve its learning performance and ability to generalize.Yayın Wavelet-based image compression by hierarchical quantization indexing(IEEE, 2009) Ateş, Hasan Fehmi; Tamer, EnginIn this paper, we introduce the quantization index hierarchy, which is used for efficient coding of quantized wavelet coefficients. A hierarchical classification map is defined in each wavelet subband, which describes the quantized data through a series of index classes. Going from bottom to the top of the tree, neighboring coefficients are combined to form classes that represent some statistics of the quantization indices of these coefficients. Higher levels of the tree are constructed iteratively by repeating this class assignment to partition the coefficients into larger subsets. The class assignments are optimized using a rate-distortion cost analysis. The optimized tree is coded hierarchically from top to bottom by coding the class membership information at each level of the tree. Despite its simplicity, the algorithm produces PSNR results that are competitive with the state-of-art coders in literature.Yayın Retinal disease classification using optical coherence tomography angiography images(Institute of Electrical and Electronics Engineers Inc., 2024) Aydın, Ömer Faruk; Nazlı, Muhammet Serdar; Tek, Faik Boray; Turkan, YaseminOptical Coherence Tomography Angiography (OCTA) is a non-invasive imaging modality widely used for the detailed visualization of retinal microvasculature, which is crucial for diagnosing and monitoring various retinal diseases. However, manual interpretation of OCTA images is labor-intensive and prone to variability, highlighting the need for automated classification methods. This study presents an aproach that utilizes transfer learning to classify OCTA images into different retinal disease categories, including age-related macular degeneration (AMD) and diapethic retinopathy (DR). We used the OCTA-500 dataset [1], the largest publicly available retinal dataset that contains images from 500 subjects with diverse retinal conditions. To address the class imbalance, we employed k-fold cross-validation and grouped various other conditions under the 'OTHERS' class. Additionally, we compared the performance of the ResNet50 model with OCTA inputs to that of the ResNet50 and RetFound (Vision Transformer) models with OCT inputs to assess the efficiency of OCTA in retinal condition classification. In the three-class (AMD, D R, Normal) classification, ResNet50-OCTA o utperformed ResNet50-OCT, but slightly underperformed compared to RetFound-OCT, which was pretrained on a large OCT dataset. In the four-class (AMD, DR, Normal, Others) classification, ResNet50-OCTA and RetFound-OCT achieved similar classification a ccuracies. This study establishes a baseline for retinal condition classification using the OCTA-500 dataset and provides a comparison between OCT and OCTA input modalities.Yayın Enhancing real estate listings through image classification and enhancement: a comparative study(Multidisciplinary Digital Publishing Institute (MDPI), 2025-05-22) Küp, Eyüp Tolunay; Sözdinler, Melih; Işık, Ali Hakan; Doksanbir, Yalçın; Akpınar, GökhanWe extended real estate property listings on the online prop-tech platform. On the platform, the images were classified into the specified classes according to quality criteria. The necessary interventions were made by measuring the platform’s appropriateness level and increasing the advertisements’ visual appeal. A dataset of 3000 labeled images was utilized to compare different image classification models, including convolutional neural networks (CNNs), VGG16, residual networks (ResNets), and the LLaVA large language model (LLM). Each model’s performance and benchmark results were measured to identify the most effective method. In addition, the classification pipeline was expanded using image enhancement with contrastive unsupervised representation learning (CURL). This method assessed the impact of improved image quality on classification accuracy and the overall attractiveness of property listings. For each classification model, the performance was evaluated in binary conditions, with and without the application of CURL. The results showed that applying image enhancement with CURL enhances image quality and improves classification performance, particularly in models such as CNN and ResNet. The study results enable a better visual representation of real estate properties, resulting in higher-quality and engaging user listings. They also underscore the importance of combining advanced image processing techniques with classification models to optimize image presentation and categorization in the real estate industry. The extended platform offers information on the role of machine learning models and image enhancement methods in technology for the real estate industry. Also, an alternative solution that can be integrated into intelligent listing systems is proposed in this study to improve user experience and information accuracy. The platform proves that artificial intelligence and machine learning can be integrated for cloud-distributed services, paving the way for future innovations in the real estate sector and intelligent marketplace platforms.
