Keywords : Noise
Iraqi National Journal of Earth Sciences,
2007, Volume 7, Issue 1, Pages 1-12
Remotely sensed data especially satellite images contain noise in some cases for both types systematic and non-systematic. These noise represent by forms working to create problems in interpretation and analysis of remote sensing data. The non-systematic noise considered one of these problems because it appears by an random points in all parts of images and with different intensity of it surrounding areas. Therefore, the Non-systematic noise must be use active processing software to remove it, in order to avoid the effect on the other pixels which are not contain noise in the image. Oppositely, the systematic noise is shown by systematic frequently lines patterns, therefore it can been recognize and processing.
The present study shows an active method for removing the non-systematic noise without change the other pixels. The suggested method is represented by a design of decision function depending on mixed filter, which is formed by merging the median filter and average filter depending on the thresholds value which used in comparism.
The suggested method provided a good accuracy comparing with the classical methods, by removing the non-systematic noise without any pixel distortions of the original image.