An Image Enhancement Approach to Achieve High Speed Using Adaptive Modified Bilateral Filter for Satellite Images Using FPGA

Abstract

For real time application scenarios of image processing, satellite imaginary has grown more interest by researches due to the informative nature of image. Satellite images are captured using high quality cameras. These images are captured from space using on-board cameras. Wrong ISO setting, camera vibrations or wrong sensory setting causes noise. The degraded image can cause less efficient results during visual perception which is a challenging issue for researchers. Another reason is that noise corrupts the image during acquisition, transmission, interference or dust particles on the scanner screen of image from satellite to the earth stations. If quality degraded images are used for further processing then it may result in wrong information extraction. In order to cater this issue, image filtering or denoising approach is required. Since remote sensing images are captured from space using on-board camera which requires high speed operating device which can provide better reconstruction quality by utilizing lesser power consumption. Recently various approaches have been proposed for image filtering. Key challenges with these approaches are reconstruction quality, operating speed, image quality by preserving information at edges on image. Proposed approach is named as modified bilateral filter. In this approach bilateral filter and kernel schemes are combined. In order to overcome the drawbacks, modified bilateral filtering by using FPGA to perform the parallelism process for denoising is implemented.