基于图像扩散去噪的距离特征量提取研究

距离特征量反映了目标运动过程中的距离变化率，该观测量可由目标辐射噪声LOFAR谱图的干涉条纹中提取得到。低信噪比情况下，由于噪声的影响，干涉条纹特征会变模糊，提取的距离特征量值精度会降低。为提高提取精度，提出了一种边缘定向增强型扩散方程去噪方法，对LOFAR谱图进行处理，使条纹特征更加明显。迭代次数对扩散方程去噪效果影响很大，根据相关性原理，提出了一种迭代次数选择方法。为了使条纹特征更加易于区分，提出了一种基于区域分割的图像二值化方法。数值模拟实验和海试数据处理结果表明，LOFAR谱图经过处理后，提取出的距离特征量精度明显提高。

Research on extraction of distance characteristic feature based on image diffusion denoising method

The distance change ratio of target in movement can be described by distance characteristic feature. It can be extracted from LOFAR (Low Frequency Analysis Recording) spectrum striations of target radiated noise. Low precision of distance characteristic feature has been obtained because of the low signal to noise ratio and striation pollution. In order to improve the precision, a new edge-directed enhancement diffusion equation model has been put forward to denoise the LOFAR spectrum image. Because denoising performance has been greatly affected by the iteration times of diffusion equation, a method of choosing the times is presented. In order to let the striations characteristic more clear, a method of binary image based on image dividing is proposed. The results of numerical simulation and ocean experiments show that the precision of distance characteristic feature is one times higher after LOFAR spectrum image processing.