With the incessant advancement in optics, electronics and computer technologies during the
last three decades, commercial digital video cameras have experienced a remarkable evolution, and can now
be employed to measure complex motions of objects with sufficient accuracy, which render great assistance
to structural displacement measurement in civil engineering. This paper proposes a novel image-based approach
for dynamic measurement of structures. One digital camera is used to capture image sequences of planar
targets mounted on vibrating structures. The mathematical relationship between image plane and real
space is established based on computer vision theory. Then, the structural dynamic displacement at the target
locations can be quantified using point reconstruction rules. Compared with other tradition displacement measurement
methods using sensors, such as accelerometers, linear-variable-differential-transducers (LVDTs)
and global position system (GPS), the proposed approach gives the main advantages of great flexibility, a
non-contact working mode and ease of increasing measurement points. To validate, four tests of sinusoidal
motion of a point, free vibration of a cantilever beam, wind tunnel test of a cross-section bridge model, and
field test of bridge displacement measurement, are performed. Results show that the proposed approach can
attain excellent accuracy compared with the analytical ones or the measurements using conventional transducers,
and proves to deliver an innovative and low cost solution to structural displacement measurement