Doppler weather radar clutter suppression based on texture feature

Abstract

Clutter suppression becomes a key point of quality control to Doppler weather radar reflectivity image, because the clutter from non-weather targets can interfere with weather targets. A novel method of suppressing clutter is proposed in this paper, which is mainly based on the Gray-level Co-occurrence Matrix(GLCM) feature. The feature possesses rotation invariability and strong ability to distinguish weather targets between clutters, which is obtained after the reflectivity image is processed by the morphological method and is segmented with the region growing method. According to statistics, there are two thresholds in different areas and Doppler velocity values in the clutters are always between -1 and 1 m/s, but precipitation clouds are always of outside this range. So the proposed feature, regional area and the Doppler velocity are combined to form a 3-layered identification method. Test result shows that the clutter clearance rate of the paper is improved from 95.7% under 2.8% loss rate to 96.5% under 1.0% loss rate.