Adaptive Higher-Order Spectral Analysis for Image Recovery Under Distortion of Moving Water Surface using Dragonfly-Colliding Bodies Optimization

AbstractReconstruction of an underwater object from a sequence of images distorted by moving water waves is a challenging task. Most of the environmental research has been employing image data in recent days. The precision of this research is often dependent on the superiority of image data. In the existing approaches, the problem of analyzing video sequences when the water surface is disturbed by waves. The water waves will affect the appearance of the individual video frames such that no single frame is completely free of geometric distortion. Thus, the image acquisition from the environmental condition is more complex and crucial, but it must be focused on getting the high spectral and spatial quality. The primary intent of this paper is to plan for the intelligent higher-order spectral analysis for recovering the images from the moving water surface. The three main phases of the proposed image recovery model are (a) image pre-processing, (b) lucky region selection, and (c) image recovery. Once the pre-processing of the image is carried out, the lucky region selection is performed by computing the dice coefficient method. As a modification to the existing methods, the proposed model adopts optimized bispectra to enhance the quality of the recovered image. A hybrid algorithm with Dragonfly-Colliding Body Optimization (D-CBO) is used for enhancing the bispectra method. The proposed model has been tested on distorted underwater images. From the experimental analysis, in terms...
Source: Sensing and Imaging - Category: Biomedical Engineering Source Type: research