SATELLITE IMAGE ENHANCEMENT TECHNIQUE BASED ON LANCZOS INTERPOLATION AND NLFMT FILTERING

Satellite images are being used now a days in various fields such as weather forecasting, remote sensing, cartography, oceanography
etc. Obtaining high quality images are of prime importance in such cases. But it may not be always possible due to the sensor
limitations, atmospheric conditions and due to transmission. Thus the images suffer from the drawback of losing high frequency
contents (which results in blurring). A wavelet domain approach based on method noise thresholding and non local means (NLM) is
proposed for the enhancement of the images. An input image is passed through a NLM filter, then its method noise is thresholded.
The NLM filtered image is summed with the thresholded method noise to cater for the image details removed by the NLM filter.
The denoised image is interpolated using the Lanczos interpolator for resolution enhancement. Objective and subjective analyses
reveal superiority of the proposed technique over the conventional and state-of-the-art image enhancement techniques.


!PDF