Akademik Veri Yönetim Sistemi
Tezin Türü: Yüksek Lisans
Tezin Yürütüldüğü Kurum: Orta Doğu Teknik Üniversitesi, Mühendislik Fakültesi, Elektrik ve Elektronik Mühendisliği Bölümü, Türkiye
Tezin Onay Tarihi: 2019
Tezin Dili: İngilizce
Öğrenci: KAMİ ÇEVİK
Danışman: Tayfun Akın
Özet:Infrared cameras generally require pre-processing on raw image data in order to improve the output image quality. Raw output of the infrared cameras occupies a narrow part of the available dynamic range. Contrast enhancement is one of the pre-processing operations and used for improving dynamic range of the low contrast images. This thesis proposes an adaptive histogram equalization method for enhancing the contrast of long wavelength infrared camera output with 640×480 resolution and provides a FPGA implementation architecture of this method that operates in real-time. Proposed method aims to improve the contrast of the background and foreground details while avoiding from over-enhancement effects and preserving thermal characteristics of the frame. Design of the proposed contrast enhancement algorithm based on integration of Adaptive Double Plateaus Histogram Equalization and Weighting Mean-Separated Sub-Histogram Equalization algorithms which are two widely used contrast enhancement methods for infrared and grayscale images. Threshold calculation is rearranged in order to prevent the excessive enhancement effects. Since the proposed method is a modified version of these two methods, it named as Weighting Mean-Separated Double Plateaus Histogram Equalization. Various contrast enhancement algorithms are implemented and compared quantitatively in MATLAB for evaluating the performance of the proposed method. FPGA implementation output is also compared with MATLAB output for testing the implementation process. Real-time image processing pipeline is designed as compatible for integration of the implemented design with targeted infrared imaging system. Test results show that the developed method improves background and foreground gray level intensity values with reducing excessive enhancement effects.