TY - JOUR AU - Jeelani, Azra AU - M B, Veena PY - 2017/05/10 Y2 - 2024/03/29 TI - Denoising the Underwater Images by using Adaptive Filters JF - European Journal of Applied Sciences JA - EJAS VL - 5 IS - 2 SE - Articles DO - 10.14738/aivp.52.2939 UR - https://journals.scholarpublishing.org/index.php/AIVP/article/view/2939 SP - 01 AB - The Sound Navigation And Ranging and synthetic aperture radar images are perturbed by the multiplicative noise called speckle noise . The presence of speckle noise leads to incorrect analysis and has to be handled carefully. Images have a strong variation from one pixel to another which reduces the efficiency of the algorithms for detection and classification. In this paper, the most well-known filters are analyzed by using underwater images. It is shown that they are based on a test related to the local coefficient of variation of the observed image, which describes the scene heterogeneity. Linear noise removing models can remove noise but are not able to preserve edges of the images in an efficient manner . On the other hand, Non-linear models can handle edges in a much better way than linear models. It is found that the linear filters   and nonlinear filters can remove noise from small area objects and homogeneous areas but not in heterogeneous areas. Adaptive filters are used to remove noise not only from homogeneous area but also from heterogeneous areas. This paper presents an optimum filter by using locally estimated parameter values to provide minimum mean square error in order to smoothen the sonar images. It is also shown that the optimum filter is computationally efficient. The performance of this adaptive filter is compared (qualitatively) with other filters and found that the complexity of the frost filter is reduced. ER -