Research on Target Recognition of Low Resolution Anti-ship Missile Based on Classifier Combination

摘要

In order to deal with complex naval battle field, it is of great significance to study how to improve the target recognition performance of anti-ship missile. The target recognition performance can be improved effectively by making full use of the RCS frequency characteristic of frequency agility anti-ship missile, Establish a RCS frequency characteristic database with six target classes from various aspect angles, and nine statistically based RCS frequency features are defined on the database, three kinds of classifiers are used as base classifiers: quadratic discriminant classifier, 7-nearest neighbor classifier and decision tree classifier, the proposed combination rule utilizes a LM-BP neural network combiner, and the performance of this combiner is compared to base classifiers as well as four other non-trainable combiners, which shows that classifier combination based LM-BP can improve the target recognition performance of anti-ship missile.