Feature extraction of underwater images by combining fuzzy c-means color clustering and LBP texture analysis algorithm with empirical mode decomposition

Abstract

Submerged objects discovery has been broadly completed by utilizing an acoustic device like side-scan sonar (SSS) which captured the pictures of seabed silt and structures. Such pictures are known as SSS image and it is of low contrast due to pixels intensity exists wider in a restricted range of the histogram. Therefore, the items in this sort of pictures are not clear and distinct. This paper presents fuzzy c-means (FCM) with local binary pattern. (LBP) and empirical mode decomposition (EMD) combined for enhancement of the SSS images. In this, EMD is used for image enhancement and FCM utilized to segment the image in order to extract the feature of the SSS image and Local Binary Pattern (LBP) algorithm is used to find texture of the enhanced image. The EMD is a versatile algorithm helpful for breaking down nonlinear and non-stationary signals. Thereby, intrinsic mode functions (IMF) of the three shading channels (Red, Green, and Blue) is calculated independently. Then, all the three channels IMFs are combined with ideal weights. It induces that the pixels of enhanced SSS images are uniformly distributed in the histogram range and also improved the color contrast problem. Therefore, the proposed approach has better density upgrade and the ocean bed structure will be fortified concentrate and dregs easily.