| Popis: |
Gaussian filter smoothens digital images by reevaluating every pixel during preserving edges. In computing the new pixel value, a window is centered on the pixel of interest with local neighborhood pixels. In this paper, an attempt is made to apply Gaussian filtering technique over Landsat -7 ETM+ satellite data product to analyze the impact of bandwidth. The selection of window is found on the statistical analysis which best enhances the image while preserving the edges. For satellite image with spatial resolution around 30m, window 9X9 for SD = 3, 7X7 for SD = 1.5, 5x5 for SD = 0.75 and 5x5 for SD = 0.375 are recommended. Resulting in blurred image, the largest window 9x9 was recommended to obtain better results. |
