In view of the problem that support vector machine classifier learning new knowledge is poor in real time, a new algorithm of radar emitter identification based on hull vector and Parzen window density estimation is studied. The algorithm takes the Parzen window density estimation to eliminate the outliers, and reduces the training time by using the hull vector of the sample set. The identification of radar signal samples is finished by using the trained classifier. The simulation results show that the online learning algorithm based on hull vector and Parzen window density estimation has good timeliness and high recognition rate.