SVM Target Identification Method Based on HRRP Sample Partition

On the basic of analyzing the quality and characteristic of high resolution range profile (HRRP), aim at its azimuth sensitivity and feature instability, one kind of sample partition method based on the HRRP´s inner correlation coefficient threshold value was proposed, and it was adopted to partition experiment targets´ HRRP into some subspace. The support vector machine (SVM) was trained by the sample that made up from some HRRP selected from the some sample subspace. The targets were identified by the trained SVM, and the correct identification rate is clearly better than the SVM trained by random sample. So, the method was a sort of valid and correct target identification method.