Fuzzy Gravitational Search Algorithm Based Image Zooming Interpolation Scheme

Abstract

This thesis aims to apply fuzzy gravitation search algorithms to decrease the image zooming inconsistent condition The image interpolation method distinguishes between the two categories: single frame and multi-frame The latter is often used in visually due to the continuous access to live images so a real-time zooming effect can be achieved The former is mostly used in repair reconstruction and local pictures to enlarge the view In the image scaling process the hardest part is to increase or maintain the sharpness and smoothness of the image and to reduce the blurring The proposed method is to modify the traditional linear interpolation method and make use of the fuzzy gravitational search algorithm in order to achieve optimal compensation rate of pixel Even if in high-magnification scaling we still have a clear image Simulation results demonstrate that the proposed scheme gives a higher peak-signal-to-noise ratio (PSNR) and shows a better images results in comparison with traditional method

Date of Award	2015 Jul 30
Original language	English
Supervisor	Tzuu-Hseng S. Li (Supervisor)