<p>In this paper, an autonomous underwater pipeline tracking is accomplished by using a bio-inspired robotic fish, AquaNavigator, with an image-based visual servoing control. AquaNavigator has a novel design architecture in terms of the propulsion tail mechanism which is based on a double-slider crank mechanism. The presented design also contains an onboard camera for vision input, allowing the robotic fish to perform intelligent operations in the underwater environment, such as autonomous pipeline tracking and inspection of any possible malfunction of underwater pipelines which transport hydro-chemicals and other resources. In this paper, a complete nonlinear dynamic model of the robotic fish is presented and a modular simulation environment is developed to validate any applied visual servoing control technique for pipeline tracking applications. A comparison of Position-based Visual Servoing and Image-based Visual Servoing control is shown in the simulator for the specific application of pipeline tracking. A Hough transformation based image processing technique is developed for the pipeline detection. A proportional-integral-derivative (PID) control scheme with a single angle tracking is established to achieve the both the position and orientation control of the robotic fish with respect to the pipeline. The image-based visual servoing control is implemented on a robotic fish to track the pipeline, and the tracking performance is experimentally validated in an indoor swimming pool. The experimental results show that the robotic fish can converge to the pipeline in a time span of 10 seconds with a distance covered less than 1 m.</p>