The real time monitoring of radar products is an important part of real time running and monitoring services in the CMA Meteorological Observation Centre, and the automated identification of abnormal radar products will therefore have practical significance and research value. The identification system based on image processing technology to achieve the automatic recognition of abnormal radar products is discussed, and the system model and automatic recognition algorithm of radar data abnormity are introduced in detail, aiming at the feature extraction, iterative learning, and so on. The analysis of the detected radar product abnormalities shows that the automatic identification system based on image processing technology can detect abnormal radar products efficiently in real time to facilitate the quick location of abnormal radar and timely repair, and can also be used as a basis for radar repair or for radar upgrading to improve the efficiency of radar applications.