Development of image processing Algorithm to detect corrosion in Underwater infrastructures

Abstract

With the recent development, the technique has been highly newlineadvanced for the identification of objects as human and objects on land. newlineThough, due to unavoidable circumstances, it is relatively uncommon in the newlinemarine field. Multiple variables such as watercolor, lighting uniformity, newlineunderwater video acquisition which are somewhat tricky affect the reason for newlineanalysis primarily two points as under localization and classification. Thus, newlineuseful object classification, as well as recognition, has a more extraordinary newlinesignificant aimed at marine equipment intelligence. The underwater object is newlineto confine and identify objects in underwater sceneries. This research is one newlineof the attracted topic due its complete range application for oceanography newlinefield. Thus, it is a demanding task because of promising setting and lighting newlineconditions. newlineObject detection system which is based on deep learning newlinemethodology identifies the better outcome but still unsatisfied. Based on newlinepattern analysis to detect object using underwater video processing has been newlineproposed in first methodology. Firstly, the input image is involved the preprocessing newlinetask to remove the noise by using Laplacian Bell pattern method. newlineThis image enhancement method enhances the quality of image. Then, the newlineenhanced image involves the pattern extraction of image by using the L.G.P. newlineThis method extracts the local features from image based on the parameters. newlineFinally, the classification task performs for tracking the object and the target newlineis identified by blobs extraction newline