The conventional method of surveying utilizing manned vessels requires a large investment of labor-intensive and time-consuming efforts. With the phenomenal progress of unmanned surface vessels (USVs), they have become a useful tool for surveyors and engineers who have been seeking a more productive and low-cost method as an alternative. This paper depicts a novel design of USVs for autonomous detection and recognition of buried submarine pipeline. The design adopted a parametric subbottom profiling system with embedded algorithms for path planning, autonomous obstacle avoidance, and autonomous pipeline recognition and navigation. The pipeline detection is based on the analysis of quadratic functions generated by the subbottom data set. Compared to the conventional method, the use of USVs equipped with subbottom profiling system turns out to be more useful and efficient for accurate detections of submarine pipeline.