Acquiring traffic flow parameters of inland waterway is one key issue in developing intelligent inland waterway transport system. The paper proposes a method of vision measurement to acquire waterway traffic flow parameters. It uses vision sensors to capture panoramic view image of channel

extracts interested objects (ships)

and establishes visual perception measurement model for acquiring geometric parameters of the ship under multi-view. A model which used CCD vision sensors to capture image information and broad knowledge-based sensors to fuse information is constructed and is used to compute water transport parameters

including the tonnage of ships

ship types

ship speed

occupancy rate of waterway

traffic density

the security state of ships

and other basic parameters of important waterway transport infrastructure

which serves as basis for traffic management

planning

and decision-making for waterway transport management sector. The results show that the system can obtain the traffic flow parameters of water access reliably

which meets the engineering requirements.