摘要
海底声纳图像是海底探测中目标回波的数据可视化结果,海底声纳图像的去噪结果对后期目标识别具有重要作用。但是采用当前方法进行海底声纳数据去噪时,存在海底声纳数据边缘细节损失严重的问题,为此提出一种基于最小二乘自适应的海底声纳数据的可视化去噪算法。上述算法首先利用最小二乘自适应算法对一维海底声纳图像信号进行去噪处理,通过多次迭代获得滤波器参数,构成滤波掩模,再对二维海底声纳图像进行滤波,对含噪声的海底声纳二维图像进行NSCT分解,获得具有不同方向信息的高低频海底声纳图像,采用中值算法对海底声纳图像高频部分进行滤波,并采用非局部均值滤波处理海底声纳图像低频部分,综合滤波后的高低频海底声纳图像的噪声情况选取阈值,对不同区域的阈值利用不同的因子进行调整完成对海底声纳数据的可视化去噪。仿真证明,所提算法能够有效提高海底声纳数据的去噪效果,且具有较好的可视化效果。
In this paper, we propose a visualization de-noising algorithm of data of seabed sonar based on least squares self-adaption. Firstly, algorithm of least squares self-adaption was used to carry out de-noising processing for signal of seabed sonar image with one dimension and parameter of filter was obtained via multiple iteration to form filtering mask. Then, filtering was carried out for seabed sonar image with two dimension and NSCT decomposition was carried out for the two-dimensional image containing noise. Seabed sonar image with high and low frequency hav- ing different direction information was obtained. Moreover, mid-value algorithm was used to carry out filtering for high-frequency part of the sonar image and nonlocal average f;hering was used to deal with low-frequency part of the image. Threshold value was selected after synthesizing noise situation of the image with high and low frequency after filtering and threshold value of different area was adjusted via different factors. Finally, the visualization de-noising was completed. Following conclusion can be drawn from experimental simulation. The algorithnl can improve de-noi- sing effect of the sonar data effectively. It has better visualization effect.
作者
张杭琦
ZHANG Hang-qi(School of Mathematics and Information Science, North China University of Water Resources and Electric Power, Zhengzhou Henan, 450046,Chin)
出处
《计算机仿真》
北大核心
2017年第11期176-179,共4页
Computer Simulation
关键词
海底声纳数据:可视化
去噪算法
Seabed sonar data
Visualization
De-noising algorithm
