论文部分内容阅读
在基于高分辨距离像的雷达目标识别中,幂变换是一种常用的非线性预处理方法,对提高识别率有一定的作用。但幂变换对高分辨距离像目标识别的性能提升受信噪比情况影响较大,且幂变换指数的选取对识别结果影响明显,甚至有不利影响。文中针对幂变换存在的问题,提出用对数变换对高分辨距离像进行预处理。比较分析了对数变换与幂变换的作用机理的异同点,并重点分析了对数变换对高分辨距离像类内和类间相关系数的影响。研究表明:对数变换相较于幂变换更适于对高分辨距离像进行预处理,且实测数据的实验结果表明对数变换比幂变换能更有效地提高识别率。
In radar target recognition based on high resolution range images, power transformation is a commonly used nonlinear preprocessing method, which has a certain effect on improving the recognition rate. However, the power conversion has a great effect on the performance improvement of high resolution distance target recognition due to the signal-to-noise ratio, and the selection of power conversion index has a significant or even adverse effect on the recognition result. In this paper, aiming at the existent problem of power conversion, the paper proposes to use logarithmic transform to preprocess high resolution range images. The similarities and differences between logarithmic transformation and power transformation are compared and analyzed, and the influence of logarithmic transformation on the correlation coefficient between the inner and the outer classes of high resolution distance image is emphatically analyzed. The results show that compared with power conversion, logarithmic transformation is more suitable for preprocessing high resolution range images. Experimental results show that logarithmic transformation can improve recognition rate more effectively than power transformation.