基于直方图参数估计CFAR的SAS图像目标检测算法的研究
首发时间:2010-01-06
摘要:本文给出了一种合成孔径声纳图像目标检测的新算法。该算法利用这类图像目标相对于杂波所占的像素数目比很小的特点,以图像的直方图为基础,与典型背景杂波分布的概率密度函数进行快速拟合,并采用最小均方误差准则来确定最终的杂波分布及参数。根据得到的参数进行恒虚警率检测,从而将目标从图像中提取出来。实践表明,本算法计算迅速,检测概率高,能够为后续的识别工作提供较好的检测结果。
For information in English, please click here
The Research of Target Detection Algorithm of SAS Images Using CFAR Based on Histogram Parameter Estimation
Abstract:In this paper, a kind of new algorithm for target detection of SAS images is presented. This algorithm takes advantage of the feature that the pixels on behalf of the target are far more than that of the background in SAS iamges. So the histogram of the image is calculated for the quick fitting of the pdf of standard noise distribution, which is used to determine the best distribution model and parameters according to the least mean square (LMS) criterion. The parameters are then provided for CFAR detection, and the target will be extracted from the entire image. This algorithm works fast in practice, and could offer good result for subsequent identification with high efficiency.
Keywords: SAS target detection histogram CFAR
基金:
论文图表:
引用
No.3852051017712627****
同行评议
共计0人参与
勘误表
基于直方图参数估计CFAR的SAS图像目标检测算法的研究
评论
全部评论0/1000