Curvelet Approach for Deep-sea Sonar Image Denoising, Contrast Enhancement and Fusion

Abstract

Side-scan sonar acquires high quality imagery of the seafloor with very high spatial resolution but poor locational accuracy. However, multi-beam sonar obtains high precision position and underwater depth in seafloor points. In order to fully utilize all information of these two types of sonars, it is necessary to fuse the two kinds of sonar data. This paper gives curvelet transform for enhancing the signals or details in different scales separately. It also proposes a new intensity sonar image fusion method, which is based on curvelet transform. Considering the sonar image forming principle, for the low frequency curvelet coefficients, we use the maximum local energy method to calculate the energy of two sonar images. For the high frequency curvelet coefficients, we take absolute maximum method as a measurement. The main attribute of this paper is: Firstly, the multi-resolution analysis method is well adapted the cured-singularities and point-singularities. It is useful for sonar intensity image enhancement. Secondly, maximum local energy is well performing the intensity sonar images, which can achieve perfect fusion result. The experimental results show that the method can be used in the flat seafloor or the isotropic seabed. Compared with wavelet transform method, this method can get better performance.