Evaluation of Speckle Noise Reduction of TerraSAR Intensity Imagery

Abstract

In this study, the effect of speckle noise reduction was evaluated with high resolution satellite TerraSAR intensity data. Multi-temporal polarization imagery (HH & VV) was acquired on different growth stages of rice plant. Three kinds of filters, i.e. Lee, Frost, Gamma Map, were applied for speckle noise filtering with different window sizes. Six indices, i.e. Standard deviation, Equivalent number of looks, Speckle suppression index, Edge preservation index, Image sharpness, and Image detail preservation index, were used in order to compare and assess the effect of speckle noise reduction of different filters. Based on the evaluation results, the most appropriate window size was determined for speckle noise filtering of utilized data.