Cilt:62 Sayı:02 (2020)

Browse

Recent Submissions

Now showing 1 - 6 of 6
  • Item
    Towards Real Tıme Image Dehazıng On Androıd Operatıng System
    (Ankara Üniversitesi Fen Fakültesi, 2020-07-01) Cimtay, Yücel; Other; Other
    Sis, görüntü ve videonun kalitesini düşüren en önemli etkilerden biridir. Bu, kontrastı azaltır ve görsel verimliliği azaltır. Sis giderme için genellikle Atmosferik ışık saçılım modeli (AISM) kullanılır. Bu modelde ölçülecek iki bilinmeyen vardır: atmosferik ışık ve iletim. Bu tür hesaplamalar kolay değildir ve atmosferik ışığın hesaplanması çok zaman alır. Bu durum, gerçek zamanlı olarak sisin çözülmesini zorlaştırır. Sis giderme uygulamaları uzun süredir yaygın olarak uygulanmasına rağmen, bu çalışma android işletim sistemi üzerinde gerçek zamanlı sis gidermnin ilk denemelerinden biridir. Bu, gerçek zamanlı sis gidermeyi bir mobil uygulama veya araca dönüştürmek açısından çok önemlidir.
  • Item
    A Concatenated Up And Down Tapered Fıber For Sımultaneous Measurement Of Straın And Temperature
    (Ankara Üniversitesi Fen Fakültesi, 2020-07-01) Bilsel, Mustafa; Navruz, İsa; Elektrik-Elektronik Mühendisliği; Mühendislik Fakültesi
    A novel fiber optical sensor based on in-line fiber Mach-Zehnder interferometer for simultaneous measurement of strain and temperature is proposed and demonstrated experimentally. The interferometer is simple, extremely robust and highly sensitive and consists of two concatenated parts; one is a down-tapered fiber (DTF) and the other is an up-tapered fiber (UTF). UTF and DTF sections of the sensor are fabricated by using a commercial fiber splicer and a non-commercial setup based on heating and stretching a portion of a standard single-mode fiber, respectively. While UTF section behaves as a beam splitter to decompose the injected light into core and cladding modes, DTF section provides evanescent field to access the surrounding environment. Experimental results indicate that the resolutions of 0.83 °C and 45.80 micro-epsilon were achieved in temperature and strain, respectively, for simultaneous measurement with a 10 pm of wavelength resolution.
  • Item
    Modern Learnıng Technıques And Plant Image Classıfıcatıon
    (Ankara Üniversitesi Fen Fakültesi, 2020-07-01) Ünal, Metehan; Bostancı, Erhan; Güzel, Mehmet Serdar; Aydın, Ayhan; Bilgisayar Mühendisliği; Mühendislik Fakültesi
    The intelligent machines concept is born in sci-fi scenarios. Today it seems to be we are much closer to realizing this idea than ever before. By imitating the human nervous system, machines can learn many things. This paper explains modern learning techniques like artificial neural networks, transfer learning. Later purposes an experiment to classify plant seedling images to test the transfer learning with two different CNN architectures. Although the architects were not actually created for this task, result were quite accurate for a different classification task.
  • Item
    Lpı Radar Waveform Classıfıcatıon Usıng Bınary Svm And Multı-Class Svm Based On Prıncıpal Components Of Tfı
    (Ankara Üniversitesi Fen Fakültesi, 2020-12-31) Bektaş, Almıla; Ergezer, Halit; Other; Other
    Since cognition has become an important topic in Electronic Warfare (EW) systems, Electronic Support Measures (ESM) are used to monitor, intercept and analyse radar signals. Low Probability of Intercept (LPI) radars is preferred to be able to detect targets without being detected by ES systems. Because of their properties as low power, variable frequency, wide bandwidth, LPI Radar waveforms are difficult to intercept with ESM systems. In addition to intercepting, the determination of the waveform types used by the LPI Radars is also very important for applying counter-measures against these radars. In this study, a solution for the LPI Radar waveform recognition is proposed. The solution is based on the training of Support Vector Machine (SVM) after applying Principal Component Analysis (PCA) to the data obtained by Time-Frequency Images (TFI). TFIs are generated using Choi-Williams Distribution. High energy regions on these images are cropped automatically and then resized to obtain uniform data set. To obtain the best result in SVM, the SVM Hyper-Parameters are also optimized. Results are obtained by using one-against-all and one-against-one methods. Better classification performance than those given in the literature have been obtained especially for lower Signal to Noise Ratio (SNR) values. The cross-validated results obtained are compared with the best results in the literature.
  • Item
    Analyzıng The Performance Of Pure Lateratıon In Indoor Envıronments Wıth Varıous Performance Metrıcs
    (Ankara Üniversitesi Fen Fakültesi, 2020-12-31) Cengiz, Korhan; Other; Other
    Nowadays, determining the location of the users and devices in indoor buildings is promising research topic. Accurate position determination of the users for indoor environments is used for numerous applications such as public safety, supermarkets, health care applications, travelling, social networks and tourism. However, global positioning systems created for outdoor localizations cannot be used for indoor positioning systems (IPS) because detecting the exact position of a target is an issue for IPS. For indoor environments, there are several positioning algorithms such as lateration, fingerprinting, dead reckoning etc. Lateration is low cost and easy to deploy when compared to other existing algorithms. Therefore, in this study, received signal strength based pure lateration that uses synthetic data generated from MATLAB is proposed. The performance of pure lateration is investigated in terms of several performance metrics such as effect of varying number of the access points (AP), varying dimensions of the measurement area, varying Gaussian Noise power and varying number of test points in the field. The simulation of the pure lateration algorithm is conducted in MATLAB. The effect of the performance metrics are investigated and discussed in details. According to the results, accuracy performance of lateration is increased when the number of APs increase in the area, however this will bring some hardware costs. In addition, when the number of test points increases in the field, in other words the step size between two test points decreases in the field the error performance of lateration is also enhanced however, this will also cause to computational costs. Finally, enlarging the measurement area causes to decrease the accuracy performance of lateration as expected. The main purpose of this study is to obtain the optimum conditions for lateration to provide a solution for real time applications. For future work, the real time implementations of this study are performed and to improve the accuracy performance, it is aimed to use a curve fitting idea to the measured values.
  • Item
    A Statıstıcal Overvıew On Sleep Scorıng
    (Ankara Üniversitesi Fen Fakültesi, 2020-12-31) Özbek, Levent; Sütçigil, Levent; Aydın, Hamdullah; Yetkin, Sinan; Özgen, Fuat; İstatistik; Fen Fakültesi
    In this study, sleep electroencephalography (EEG) which is frequently used in statistical modelling has been modelled with the autoregressive (AR) time-series model and what kind of a structure the variance of the term white noise included in the model represented in different sleep stages has been observed. Taking all of the stages scored in accordance with Rechtschaffen and Kales criterion into account separately, epoches in each stage have been modelled with the AR and the variance of the term white noise in this model has been monitored. The study has evaluated the sleep EEG variances of a subject. In accordance with the results, the heterogeneity at Stage 2 was thought to be the reason why the objective differences appeared in scoring. It is thought that this data pointed out a necessity that the period in Rechtschaffen and Kales scoring which is called Stage 2 must be revised.