It is of importance to oil and gas industry to inspect the pipelines laid down underwater for maintainance. Human efforts are not only expensive and time consuming but also there exists a chance of hazard and danger to human divers working on such tasks. Autonomous Underwater Vehicles and Remotely Operated Vehicles have shown major involvment in improving the quality of inspection over last couple of decades and reduction in human efforts. However, there are still challenges to be addressed. One such challenge that needs to be addressed is to track pipelines in and out of burial in an efficient manner. This project will investigate statistical methods to keep track of target(pipelines) using multiple sensors. Data from these sensors will be incorporated in statistical multi-sensor fusion framework using Probability Hypothesis Density (PHD) filter for having better estimates of state of the target(s).