摘要: 在逆合成孔径雷达(ISAR)图像飞机目标的识别中,传统的特征是ISAR图像的低频傅里叶系数、几何不变矩、形状特征和能量特征等,但上述特征都没有考虑ISAR的纹理。为此,引入一种新特征用于ISAR图像中飞机目标的识别,即提取ISAR的LBP特征来描述图像的纹理,用k近邻分类器进行分类。实验结果表明,该方法达到98.7%的识别率,表明该特征的有效性。
关键词:
ISAR图像,
LBP特征,
飞机目标识别
Abstract: In the Inverse Synthetic Aperture Radar(ISAR)image plane recognition, the traditional features are lowfrequency Fourier coefficients, geometric moment invariants, shape features and energy features of ISAR images. But these features do not take into account the texture of ISAR. This paper introduces a new feature for the ISAR image plane target recognition, the LBP feature is extracted to describe the ISAR image texture, and k neighbor classifier is used to classify. Experimental result shows an average recognition rate of 98.7% is reached, the LBP feature is effective.
Key words:
ISAR image,
LBP feature,
plane target recognition
中图分类号:
季战领, 冯晓毅, 阎昆. 一种ISAR图像中飞机目标识别的新特征[J]. 计算机工程, 2010, 36(23): 192-193,196.
JI Zhan-Ling, FENG Xiao-Yi, YAN Hun. New Feature for Plane Target Recognition in ISAR Image[J]. Computer Engineering, 2010, 36(23): 192-193,196.