Classifying Clear Air Echoes via Static and Motion Streams Network

Yuxun Qu,Chenyang Zhang,Xuebing Yang,Yajing Wu,Wensheng Zhang,Guoping Zhang
DOI: https://doi.org/10.1109/lgrs.2021.3097098
IF: 5.343
2022-01-01
IEEE Geoscience and Remote Sensing Letters
Abstract:Classification of nonprecipitation echoes of radar is an inevitable step in radar-based precipitation estimation. Among nonprecipitation echoes, clear air echoes are specifically difficult to distinguish for their similarity to precipitation echoes. This letter aims to conduct a pixelwise classification of clear air echoes for image sequences of the radar reflectivity. We propose the Static and Motion streams Network (SMNet) to simultaneously utilize the static and motion features. SMNet realizes capturing the spatiotemporal characteristics while maintaining the details of the current frame via a fusion structure and a novel training method. For feature fusion, the static and motion streams are concatenated. Then, for model training, we adopt a dynamic weight assignment strategy to further extract rich information. Finally, we validate our method on an S-band single-polarization radar in Beijing, China, from May to September 2018. The results demonstrate that the overall performance of SMNet is superior to other competitors.
imaging science & photographic technology,remote sensing,engineering, electrical & electronic,geochemistry & geophysics
What problem does this paper attempt to address?