Abstract:In order to correct the interference echoes of X-band weather radar, this article collected and selected abnormal data of X-band weather radar.Analysis found that electromagnetic interference echoes mainly appear in two forms: point shaped spiral bands, strip shaped pockmarks, and straight lines.The traditional electromagnetic interference filtering method uses a single window threshold method, which eliminates interference echoes while filtering out the true meteorological echo edges due to misjudgment.This article constructs feature parameters and determines the radial threshold in each direction, assigns flag values, and implements corresponding quality control methods for different flag values radially.Finally, a multi-directional rectangular window is used to judge isolated point echoes, and the relative relationship between the single distance library and surrounding echoes is checked from multiple angles, effectively reducing misjudgment of meteorological echo edges.After batch processing non meteorological echo data, it is found that the recognition algorithm with added feature parameters has a better recognition effect on radial interference echoes.