| Summary: | 碩士 === 國立交通大學 === 電子工程系 === 89 === In this thesis, we propose several approaches to do pixel interpolation, image scaling, edge enhancement, and brightness enhancement in digital image acquisition systems. About pixel interpolation, we propose a pattern-based method and an interpolation method based on polynomial fitting. These two approaches reduce the appearance of false color and lattice effect. About image scaling, we discuss both cubic spline interpolation and B-spline interpolation. About edge enhancement, we adjust the enhancement factor at each pixel according to the image content near that pixel. This approach can provide sharp edges, fewer ringing artifacts, and lower noise enhancement. About brightness enhancement, we select the pixels around edges and build a hitogram for hem. Then, by modifying the histogram in a proper way, this approach provides enhanced brightness without affecting the perception of the original image.
|
|---|
