![Sample our Geography journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/5937/TOC-Subject-Sample-2021-Geography.jpg"})
Abstract
This work deals with the classification of radar echoes and the removal of clutter caused by the Earth’s surface. Two incoherent radar sites are considered, which are the regions of Sétif (Algeria) and Bordeaux (France) where different climates and landforms prevail. To perform this task, we used a combination of textural and fuzzy approaches. For the textural technique, we applied grey-level co-occurrence matrices that are widely used in the analysis of texture images. We have shown that among nine parameters, only energy and local homogeneity are considered to be effective in discriminating between precipitation echoes and clutter. Then, these parameters are used as inputs for the fuzzy system, while the two radar echo types are its output classes. Image processing done by using this approach has reduced ground echoes by more than 93.5% for Sétif and 92.3% for Bordeaux sites, while more than 97.6% of precipitation echoes are stored at both sites. In addition, over 96% of the anomalous propagations observed only in Bordeaux site are removed. The proposed approach gives a filtering average rate that is 94.5% higher than that obtained for the textural technique alone, which is 91.5%.
Acknowledgements
The authors would like to thank all those who contributed to the radar data set used, notably the National Meteorology Office of Algeria for the Sétif data and Météo-France for the Bordeaux images. We would also like to thank the reviewers for their valuable comments and suggestions.