侧扫声纳图像非下采样轮廓波变换域分区增强方法

Partition Enhancement Method for NSCT Domain of Side-scan Sonar Image

成像机理的限制以及海洋中存在的丰富噪声源,导致侧扫声纳图像出现噪声污染严重、目标和背景区域灰度值对比度低以及边缘呈现强度较弱等情况。针对上述问题,提出一种侧扫声纳图像非下采样轮廓波变换(NSCT)域分区增强方法。对于声纳图像低频部分,使用非线性函数增强方法,提升低频图像对比度;对于声纳图像高频部分,通过分析声纳高频图像在NSCT域上同一尺度不同方向子带系数最大值与最小值差值的分布规律,进行噪声和纹理边缘的划分以及对应的处理。将所提方法与小波硬阈值增强方法、小波Shrinkage自适应阈值增强方法做实验对比,结果表明,该方法不仅可以较好地消除噪声,而且可以抑制琐碎纹理、提升弱边缘,侧扫声纳图像增强效果更加突出。

The serious noise pollution,low contrast of gray values of target and background areas,and weak edges in the side-scan sonar image are due to the limitations of side-scan sonar imaging mechanism and the abundant noise sources in the ocean.In response to the above problems,a partition enhancement method for the non-subsampled contourlet transform(NSCT)domain of side-scan sonar images is proposed.For the low-frequency part of sonar image,a nonlinear function enhancement method is used to improve the contrast of low-frequency image.For the high-frequency part of sonar image,the noise and texture edges are partitioned and the corresponding processing is made by analyzing the distribution rule of the difference between the maximum and minimum values of sub-band coefficients in different directions on the same scale.The proposed method was compared with the gamma enhancement method and the wavelet threshold enhancement method through experiment.The results show that the proposed method not only denoises noise well,but also can suppresses trivial textures and enhances weak edges.The enhancement effect for side-scan sonar image is more prominent.