In this paper, we present the autonomous pipeline navigation method of an in-pipe robot, based on the landmark recognition system and 3D orientation sensor. The fittings of pipelines such as elbows and branches can be used as the landmarks. In this paper, we introduce a noise tolerant landmark detection method using a line laser beam projected on the internal surface of the pipeline. Once the line laser is projected on the surface of landmarks, a unique line pattern is generated. By using this line pattern, the robot can easily detect and distinguish these landmarks while recognizing the direction of the pipeline path. Simultaneously, the robot's pose information from the orientatiom sensor and the driving motor's encoder is continuously recorded and used to reconstruct a map of the pipelines. By using this map and a specific navigation strategy, the robot can navigate in the 3D pipeline autonomously. The effectiveness of the proposed method is validated by experiments using the in-pipe robot called MRINSPECT V.