| Summary: | 碩士 === 國立成功大學 === 資訊工程學系碩博士班 === 95 === Scaling is a very important issue in image processing. It has been used in many applications such as HDTV, LCD-TVs, digital video camcorders, copy-print machines, medical imaging and so on. In this thesis, a low-complexity scaling algorithm for digital image is proposed. We define a derived adaptive area-pixel model with only fixed-point coefficients and use it with the local characteristics in the source image to estimate the luminosity of each target pixel. Experiment results show that our method can reserve edge characteristics efficiently, and performs better than other previous techniques in terms of both quantitative evaluation and visual quality.
Since our method is simple and regular, it is very suitable for hardware implementation. A high performance nine-stage pipelined architecture for the proposed method is also presented in this thesis. In the simulation, our design can operate at 200 MHz properly with the TSMC 0.18μm technology.
|
|---|
