基于小波变换的侧扫声纳图像镶嵌研究

Study on side-scan sonar image mosaic based on wavelet transformation

传统的基于简单加权的侧扫声纳图像镶嵌方法,很容易造成图像信噪比的降低以及出现拼接痕迹.本文提出了一种基于小波变换的镶嵌方法,通过实测数据处理,表明该方法优于传统的镶嵌方法,不仅能够很好地将源图像细节融合在一起,而且能够很好地显示源图像各自的信息.

Traditional side-scan sonar image mosaic is based on the simple weighted method, which often results in the signal-to-noise ratio loss and untrue mosaic imprint. This paper introduces a new mosaic method based on wavelet transformation. The processed results of the actual data indicate that the method shows obvious advantages compared to the traditional methods. It can not only merge the source imagest details very well, but also display the information of source images clearly.