Yazar "Firat, Huseyin" seçeneğine göre listele
Listeleniyor 1 - 10 / 10
Sayfa Başına Sonuç
Sıralama seçenekleri
Öğe 3D residual spatial-spectral convolution network for hyperspectral remote sensing image classification(Springer London Ltd, 2023) Firat, Huseyin; Asker, Mehmet Emin; Bayindir, Mehmet Ilyas; Hanbay, DavutHyperspectral remote sensing images (HRSI) are 3D image cubes that contain hundreds of spectral bands and have two spatial dimensions and one spectral dimension. HRSI analysis are commonly used in a wide variety of applications such as object detection, precision agriculture and mining. HRSI classification purposes to assign each pixel in HRSI to a unique class. Deep learning is seen as an effective method to improve HRSI classification. In particular, convolutional neural networks (CNNs) are increasingly used in remote sensing field. In this study, a hybrid 3D residual spatial-spectral convolution network (3D-RSSCN) is proposed to extract deep spatiospectral features using 3D CNN and ResNet18 architecture. Simultaneously spatiospectral features extraction is provided using 3D CNN. In deeper CNNs, ResNet architecture is used to achieve higher classification performance as the number of layers increases. In addition, thanks to the ResNet architecture, problems such as degradation and vanishing gradient that may occur in deep networks are overcome. The high dimensionality of the HRSIs increases the computational complexity. Thus, most of studies apply dimension reduction as preprocessing. In the proposed study, principal component analysis (PCA) is used as the preprocessing step for optimum spectral band extraction. The proposed 3D-RSSCN method is tested with Indian pines, Pavia University and Salinas datasets and compared against various deep learning-based methods (SAE, RPNet, 2D CNN, 3D CNN, M3D CNN, HybridSN, FC3D CNN, SSRN, FuSENet, S3EResBoF). As a result of the applications, the best classification accuracy among these methods compared in all datasets is obtained with the proposed 3D-RSSCN. The proposed 3D-RSSCN method has the best accuracy and time performance in classifying.Öğe A Comprehensive CNN Model for Age-Related Macular Degeneration Classification Using OCT: Integrating Inception Modules, SE Blocks, and ConvMixer(Mdpi, 2024) Yusufoglu, Elif; Firat, Huseyin; Uzen, Huseyin; Ozcelik, Salih Taha Alperen; Cicek, Ipek Balikci; Sengur, Abdulkadir; Atila, OrhanBackground/Objectives: Age-related macular degeneration (AMD) is a significant cause of vision loss in older adults, often progressing without early noticeable symptoms. Deep learning (DL) models, particularly convolutional neural networks (CNNs), demonstrate potential in accurately diagnosing and classifying AMD using medical imaging technologies like optical coherence to-mography (OCT) scans. This study introduces a novel CNN-based DL method for AMD diagnosis, aiming to enhance computational efficiency and classification accuracy. Methods: The proposed method (PM) combines modified Inception modules, Depthwise Squeeze-and-Excitation Blocks, and ConvMixer architecture. Its effectiveness was evaluated on two datasets: a private dataset with 2316 images and the public Noor dataset. Key performance metrics, including accuracy, precision, recall, and F1 score, were calculated to assess the method's diagnostic performance. Results: On the private dataset, the PM achieved outstanding performance: 97.98% accuracy, 97.95% precision, 97.77% recall, and 97.86% F1 score. When tested on the public Noor dataset, the method reached 100% across all evaluation metrics, outperforming existing DL approaches. Conclusions: These results highlight the promising role of AI-based systems in AMD diagnosis, of-fering advanced feature extraction capabilities that can potentially enable early detection and in-tervention, ultimately improving patient care and outcomes. While the proposed model demon-strates promising performance on the datasets tested, the study is limited by the size and diversity of the datasets. Future work will focus on external clinical validation to address these limita-tions.Öğe Automatic Thresholding Method Developed With Entropy For Fabric Defect Detection(Ieee, 2019) Uzen, Huseyin; Firat, Huseyin; Karci, Ali; Hanbay, DavutFabric defect detection is one of the most important areas for quality control of products in the textile industry. Many different studies have developed methods for this problem. In this study, an automatic thresholding method developed with entropy has been proposed. Due to the low cost of calculation, the proposed automatic thresholding method will be very suitable for real-time applications. In this study, automatic thresholding method which is supported by 4 different entropy method was compared with otsu method which is one of automatic thresholding methods. Various tests have been made on different fabric types for comparisons. As a result of experimental studies, successful results of automatic thresholding methods supported with entropy were obtained for fabric defect detection. Renyi entropy method was the most successful result among the proposed methods.Öğe Central serous retinopathy classification with deep learning-based multilevel feature extraction from optical coherence tomography images(Elsevier Sci Ltd, 2025) Uzen, Huseyin; Firat, Huseyin; Ozcelik, Salih Taha Alperen; Yusufoglu, Elif; Cicek, Ipek Balikci; Sengur, AbdulkadirCentral Serous Chorioretinopathy (CSCR) is an ocular disease characterized by fluid accumulation under the retina, which can lead to permanent visual impairment if not diagnosed early. This study presents a deep learning-based Convolutional Neural Network (CNN) model designed to automatically diagnose acute and chronic CSCR from Optical Coherence Tomography (OCT) images through multi-level feature extraction. The proposed CNN architecture consists of consecutive layers like a traditional CNN. However, it also extracts various features by creating feature maps at four different levels (F1, F2, F3, F4) for the final feature map. The model processes information using group-wise convolution and Pointwise Convolution Block (PCB) at each level. In this way, each feature group is further processed to obtain more representative features, enabling more independent learning. After the PCB outputs, the 4 feature maps are vectorized and combined, thus creating the final feature map. Finally, classification prediction scores are obtained by applying a fully connected layer and softmax function to this feature map. The experimental study utilized two datasets obtained from Elazig Ophthalmology Polyclinic. The dataset includes 3860 OCT images from 488 individuals, with images categorized into acute CSCR, chronic CSCR, wet AMD, dry AMD, and healthy controls. Our proposed method achieves an increase in accuracy of 0.77%, attaining 96.40% compared to the highest previous accuracy of 95.73% by ResNet101. Precision is enhanced by 0.95%, reaching 95.16% over ResNet101 ' s 94.21%. The sensitivity (recall) is improved by 0.90%, achieving 95.65% versus ResNet101 ' s 94.75%. Additionally, the F1 score is increased by 0.93%, attaining 95.38% compared to ResNet101 ' s 94.45%. These results illustrate the effectiveness of our method, offering more precise and reliable diagnostic capabilities in OCT image classification. In conclusion, this study demonstrates the potential of artificial intelligence-supported diagnostic tools in the analysis of OCT images and contributes significantly to the development of early diagnosis and treatment strategies.Öğe Classification of hyperspectral remote sensing images using different dimension reduction methods with 3D/2D CNN(Elsevier, 2022) Firat, Huseyin; Asker, Mehmet Emin; Hanbay, DavutThe high dimensionality of hyperspectral remote sensing images (HRSI) affects the classification performance. Therefore, most HRSI classification methods use dimension reduction methods as a solution for high dimensionality. It is aimed to extract useful features with dimension reduction methods. At the end of this process, the data dimension is reduced and the transaction cost is decreased. In this study, LDA, PCA, IPCA, ICA, SPCA, RPCA and SVD dimension reduction methods were applied as a preprocessing step to improve HRSI classification performance. Since HRSI is volumetric data and has a spectral dimension, 2D CNN cannot extract good distinguishing features from spectral dimensions. Because 2D CNN only considers spatial information. With 3D CNN, spectral-spatial features are extracted simultaneously. However, 3D CNN increases the computational cost. Therefore, in this study, Hybrid 3D/2D CNN method is used together with dimension reduction methods. Hybrid CNN method consists of a combination of 3D CNN, 2D CNN and depthwise separable convolution. While 3D CNN extracts common spectral-spatial features, more spatial features are learned with 2D CNN used after 3D CNN. With depthwise separable convolution, it reduces the number of parameters and overfitting is prevented. The applications performed on the frequently used HRSI benchmark datasets show that the classification performance of the proposed method is better than the compared methods. In addition, Indian pines, HyRANK-Loukia, Botswana and Pavia of University datasets are used to examine the effect of dimension reduction methods used together with the hybrid 3D/2D CNN method on classification performance. As a result of the applications, the best classification accuracies were obtained in PCA, LDA and IPCA with Indian pines, PCA with Pavia of university, PCA and IPCA with Salinas, PCA, RPCA and LDA dimension reduction methods with HyRANK-Loukia.Öğe Comparison of 3D CNN based deep learning architectures using hyperspectral images(Gazi Univ, Fac Engineering Architecture, 2023) Firat, Huseyin; Hanbay, DavutHyperspectral images (HSI) are 3-dimensional (3D) image cubes with two spatial and one spectral dimensions. The development of deep learning methods has had a significant impact on HSI classification. Especially convolutional neural network (CNN) based methods are getting more attention in this field. In this study, we make use of the deep learning architectures LeNet5, AlexNet, VGG16, GoogleNet and ResNet50, which are among the successful examples of CNN for the HSI classification problem. We use a 3D CNN-based hybrid approach when using these architectures. Because, using 3D CNN, spectral-spatial features are extracted simultaneously. In this case, the classification accuracy of HSIs is increased with the spectral-spatial-based deep learning architecture. However, in the proposed model, principal component analysis (PCA) is used as a preprocessing technique for optimal band extraction from HSIs. After applying PCA, 3D cubes are obtained by neighborhood extraction and given to the input of deep learning architectures. Indian pines, Salinas, Botswana and HyRANK-Loukia datasets were used to compare the classification performances of 3D CNN-based deep learning architectures. As a result of the applications, the best classification accuracy was obtained with VGG16 architectures in Indian pines dataset, ResNet50 in Botswana dataset, VGG16 in HyRANK-Loukia dataset, LeNet5 and VGG16 architectures in Salinas dataset.Öğe Deep Learning-Assisted Detection and Classification of Thymoma Tumors in CT Scans(Mdpi, 2025) Kilic, Murat; Biyikli, Merve; Ozcelik, Salih Taha Alperen; Uzen, Huseyin; Firat, HuseyinBackground/Objectives: Thymoma is a rare epithelial neoplasm originating from the thymus gland, and its accurate detection and classification using computed tomography (CT) images remain diagnostically challenging due to subtle morphological similarities with other mediastinal pathologies. This study presents a deep learning (DL)-based model designed to improve diagnostic accuracy for both thymoma detection and subtype classification (benign vs. malignant). Methods: The proposed approach integrates a pre-trained VGG16 network for efficient feature extraction-capitalizing on its capacity to capture hierarchical spatial features-and an MLP-Mixer-based feature enhancement module, which effectively models both local and global feature dependencies without relying on conventional convolutional mechanisms. Additionally, customized preprocessing and post-processing methods are employed to enhance image quality and suppress redundant data. The model's performance was evaluated on two classification tasks: distinguishing thymoma from healthy cases and discriminating between benign and malignant thymoma. Comparative analysis was conducted against state-of-the-art DL models including ResNet50, ResNet34, SEResNeXt50, InceptionResNetV2, MobileNetV2, VGG16, InceptionV3, and DenseNet121 using metrics such as F1 score, accuracy, recall, and precision. Results: The model proposed in this study obtained its best performance in thymoma vs. healthy classification, with an accuracy of 97.15% and F1 score of 80.99%. In the benign vs. malignant task, it attained an accuracy of 79.20% and an F1 score of 78.51%, outperforming all baseline methods. Conclusions: The integration of VGG16's robust spatial feature extraction and the MLP-Mixer's effective feature mixing demonstrates superior and balanced performance, highlighting the model's potential for clinical decision support in thymoma diagnosis.Öğe Grad-CAM Enhanced Explainable Deep Learning for Multi-Class Lung Cancer Classification Using DE-SAMNet Model(Mdpi, 2026) Kilic, Murat; Biyikli, Merve; Yelman, Abdulkadir; Firat, Huseyin; Uzen, Huseyin; Cicek, Ipek Balikci; Sengur, AbdulkadirBackground/Objectives: Lung cancer (LC) is the leading cause of cancer-related mortality worldwide, making early and accurate diagnosis crucial for improving patient outcomes. Although chest computed tomography (CT) enables detailed assessment of lung abnormalities, manual interpretation is time-consuming, requires expert expertise, and is prone to diagnostic variability. To address these challenges, this study proposes DE-SAMNet, a hybrid deep learning framework for automated multi-class LC classification from CT scans. Methods: The model integrates two pre-trained convolutional neural networks-DenseNet121 and EfficientNetB0-operating in parallel to extract complementary multi-scale features. A Spatial Attention Module (SAM) is applied to each feature stream to emphasize clinically important regions. Final classification is performed through a compact fusion mechanism involving global average pooling, batch normalization, and a fully connected layer. DE-SAMNet was evaluated on two datasets: a public dataset (IQ-OTH/NCCD) with benign, malignant, and normal cases, and a private clinical dataset including benign, malignant, cystic, and healthy cases. Results: On the public dataset, the model achieved a 99.00% F1-score, 98.41% recall, 99.64% precision, and 99.54% accuracy. On the private dataset, it obtained 95.96% accuracy, 95.99% precision, 96.04% F1-score, and 96.21% recall, outperforming existing approaches. To enhance reliability, explainable AI (XAI) techniques such as Grad-CAM were used to visualize the model's decision rationale. The resulting heatmaps effectively highlight lesion-specific regions, offering transparency and supporting clinical interpretability. Conclusions: This explainability strengthens trust in automated predictions and demonstrates the clinical potential of the proposed system. Overall, DE-SAMNet delivers a highly accurate and interpretable solution for early LC detection.Öğe Hybrid 3D/2D Complete Inception Module and Convolutional Neural Network for Hyperspectral Remote Sensing Image Classification(Springer, 2023) Firat, Huseyin; Asker, Mehmet Emin; Bayindir, Mehmet Ilyas; Hanbay, DavutClassification in hyperspectral remote sensing images (HRSIs) is a challenging process in image analysis and one of the most popular topics. In recent years, many methods have been proposed to solve the HRSIs classification problem. Compared to traditional machine learning methods, deep learning, especially convolutional neural networks (CNNs), is commonly used in the classification of HRSIs. Deep learning-based methods based on CNNs show remarkable performance in HRSIs classification and greatly support the development of classification technology. In this study, a method in which the Hybrid 3D/2D Complete Inception module and the Hybrid 3D/2D CNN method are used together has been proposed to solve the HRSIs classification problem. In the proposed method, multi-level feature extraction is performed by using multiple convolution layers with the Inception module. This improves the performance of the network. Conventional CNN-based methods use 2D CNN for feature extraction. However, only spatial features are extracted with 2D CNN. 3D CNN is used to extract spatial-spectral features. However, 3D CNN is computationally complex. Therefore, in the proposed method, a hybrid approach is used by first using 3D CNN and then 2D CNN. This reduces computational complexity and extracts more spatial features. In addition, PCA is used as a preprocessing step for optimum spectral band extraction in the proposed method. The proposed method has been tested using Indian pines, Salinas, University of Pavia, HyRANK-Loukia and Houston datasets, which are frequently used in studies for HRSIs classification. The overall accuracy of the proposed method in these five datasets are 99.83%, 100%, 100%, 90.47% and 98.93%, respectively. These results reveal that the proposed method provides higher classification performance compared to state-of-the-art methods.Öğe A Hybrid Methodology Using Heuristic Methods for Two-Dimensional Cutting and Packing Problem with Rectangular Pieces(Gazi Univ, 2019) Firat, Huseyin; Alpaslan, Nuh; Hanbay, DavutThe cutting and packing problem is the process of cutting small pieces of certain sizes and proportions from materials used for different purposes in industries. Because this problem cannot be expressed by mathematical models, combinational optimization in multidimensional space is utilized for the solution. The aim of this problem is to increase the usability of the material used for the placement process and to minimize the trim loss. In this study, a solution is presented to two-dimensional regular cutting and packing problem by a combined method consisting of improved bottom-left, bottom-left fill placement algorithms, no-fit polygon and first fit decreasing heuristic algorithms. The improved bottom-left placement algorithm for the placement of parts starting from the bottom-left part according to a certain permutation order, bottom-left fill algorithm for the placement of suitable pieces to the available free spaces in placement model, no-fit polygon method for preventing the geometric overlap between the parts and the first fit decreasing heuristic algorithm is used as the selection algorithm after ordering from large to small according to the parts areas. Placement process and performance evaluation was performed for 11 different test data. As a result of the studies carried out with combined heuristic methods, it is seen that there is a placement without any waste in P2 and P10 placement models. This shows that the optimal solution is obtained. In other placement models, a trim loss was obtained between 4.54% and 16.7%. The experimental results show the effectiveness of the proposed heuristic methods for the solution of the cutting and packing problem.
