Master Degree / Yüksek Lisans Tezleri
Permanent URI for this collectionhttps://hdl.handle.net/11147/3008
Browse
2 results
Search Results
Master Thesis Design of an Offline Ottoman Character Recognition System for Translating Printed Documents To Modern Turkish(Izmir Institute of Technology, 2019) Küçükşahin, Naz; Ünlü, Mehmet Zübeyir; Ünlü, Mehmet Zübeyir; 03.05. Department of Electrical and Electronics Engineering; 03. Faculty of Engineering; 01. Izmir Institute of TechnologyOptical character recognition (OCR) is one of the most studied topics for many years. As a result of these studies, systems developed especially for the Latin alphabet have become more accurate even for handwritten texts. However, there are very limited studies on Ottoman OCR systems in the literature and it is still a subject of interest due to the complexity of the language in grammar, writing and spelling. In this thesis, it is aimed to design an offline OCR system that recognizes Ottoman characters using deep convolutional neural networks. The proposed work consists of several steps such as image processing, image digitization and character segmentation, adaptation of inputs to the network, training of the network, recognition and evaluation of results. Firstly, a character dataset was created by segmenting text images of different lengths that was selected among scanned samples of various Ottoman literature from the digital database of Turkish National Library. Two convolutional neural networks of different complexity were trained with the created character dataset and the relationship between recognition rates and network complexity was evaluated. Secondly, using the Histogram of Oriented Gradients and Principal Component Analysis, the features of the created dataset were extracted and the Ottoman characters were classified with k-Nearest Neighbor Algorithm and Support Vector Machines which are widely used classification methods in the literature. The performed analyzes have shown that both networks provide acceptable recognition rates compared to the conventional classifiers, however complex deep neural network showed better accuracy and lower loss.Master Thesis Shape Based Detection and Classification of Vehicles Using Omnidirectional Videos(Izmir Institute of Technology, 2015) Karaimer, Hakkı Can; Baştanlar, Yalın; Baştanlar, Yalın; 03.04. Department of Computer Engineering; 03. Faculty of Engineering; 01. Izmir Institute of TechnologyTo detect and classify vehicles in omnidirectional videos, an approach based on the shape (silhouette) of the moving object obtained by background subtraction is proposed. Different from other shape based classification techniques, the information available in multiple frames of the video is exploited. Two different approaches were investigated for this purpose. One is combining silhouettes extracted from a sequence of frames to create an average silhouette, the other is making individual decisions for all frames and use consensus of these decisions. Using multiple frames eliminates most of the wrong decisions which are caused by a poorly extracted silhouette from a single video frame. The vehicle types which are classified are motorcycle, car (sedan) and van (minibus). The features extracted from the silhouettes are convexity, elongation, rectangularity, and Hu moments. Three separate methods of classification is applied. The first one is a flowchart based (i.e. rule based) method, the second one is K nearest neighbor classification, and the third one is using a Deep Neural Network. 60% of the samples in the dataset are used for training. To ensure randomization, the procedure is repeated three times with the whole dataset split each time differently into training and testing samples (i.e. three-fold cross validation). The results indicate that using silhouettes in multiple frames performs better than using single frame silhouettes.