This paper deals with the study of the detection and quantification of cracks in bridge structures using image data obtained by using unmanned aerial vehicle (UAV). In general, an UAV is influenced by fly environment such as a gust of wind and self-excitation during bridge inspection. Therefore, a feasibility study is required in order to verify the performance in dynamical environment. In this study, a 2D motion jig was fabricated to simulate the dynamical environment. And a steel and two concrete crack specimens were prepared. To detect crack of the specimens, improved image processing techniques (IPTs) are proposed. It is aimed to correct the image distortion due to the vibration that can occur in the appearance of the bridge using the drone. Performances of the technique were evaluated through lab-test. Finally, improved IPTs are applied to the detected regions to quantify cracks at 100 micrometers. The non-contact detection technology of bridge damage proposed in this study can be applied to the actual bridge inspection field and it is expected that the economic and technical efficiency will be high.