Abstract: To enhance spatial resolution and measurement efficiency, we present a novel system for measuring surface velocity fields of flows on large-scale hydraulic laboratory scale models. The basic idea is to match characteristic image features. To this end, flow is seeded with tracer and a camera is recording the scenery in plumb vertical top view in a height of 5 m. A matching algorithm correlates characteristic image feature points, and instantaneous and time-averaged velocity fields are provided automatically. Applications of the system to both weir flow and jet flow experiments yield reasonable velocity fields with high spatial resolution. This velocimetry system shows to be effective, efficient, and flexible. The output velocity field has high temporal-spatial resolution and keeps considerable accuracy with PIV, which can truly reflect characteristics of indoor surface flow.

Key words: image processing, velocity measurement, feature matching, river model experiment, surface velocity field