Bridges may get deteriorated over time, and the design live load has been increased due to the growth of related industries. Therefore, the structural integrity assessment of bridge structures has drawn lots of interests for engineers. Load carrying capacity is one of the widely used structural integrity index for the bridge rating and maintenance. In Korean bridge design specification, load carrying capacity is calculated based on the truck loading tests and the corresponding measured displacement data. However, for measuring displacement, conventional sensors need stable reference points plummeted to earth from the measured points. So, the installation of conventional sensors for displacement measurement is technically not easy and costly, if not impossible, especially when bridges cross over sea channels, large rivers, or a road with passing traffic. To resolve this issue, a vision-based real-time displacement measurement system using digital image processing techniques was employed. The effectiveness of the present system was verified through a laboratory test, and it was utilized for the evaluation of load carrying capacity of a steel-plate girder bridge. The results obtained from the present system were found to show good agreement with those obtained from the conventional displacement measurement system.