Underwater man-made object prediction using line detection technique

Underwater imaging is primarily focused on search and rescue, underwater mine detection, underwater cable and pipeline overhauling and underwater geological survey. Main challenge in underwater imaging is blurriness. In underwater environment blurriness is caused by many factors which includes microscopic organism, impurities and density of water which effects refractive index of water, and bokeh which is blurred effect on those region of image that are out of focus in range. Picture of a moving object also have a blur effect, reason is motion blur. To detect object in underwater image, integration of different image processing technique has been made. It includes preprocessing to reduce blurriness and noise in image and Euclidean shape prediction by detecting lines in the image. Computationally feasible technique is also discuss in this paper which is not only independent of image data bank but also less time consuming to process.