Unmanned Aerial Vehicle for infrastructure inspection with image processing for quantification of measurement and formation of facade map

摘要

The increasing age of infrastructure such as bridges or tall buildings requires a significant effort to be performed with respect to inspection such that damage being considered critical can be recognized early in advance and a respective infrastructure management can be performed. Many countries take the advantage of advance in Unmanned Aerial Vehicle technologies and applied for mostly aerial mapping and some infrastructure inspection to lower down the cost and increase the safety of inspection personnel. Most of the UAV system in the market produce just either photo or HD video, which does not contain any quantifiable reading for engineer to identify the damage level. Second, the image or video captured are not always perpendicular to the target surface, therefore distortion of the image could cause the inaccurate of the actual measurement on the photo/video. Third, current inspection method only done by few pictures or video, but do not contain “useful map” of the structure. This paper present a in house design autopilot system integrated with different sensors and optical camera together to perform a quantification of the image, also same time data fusion allow image correction to increase the accuracy of the measurement, and most importantly, the corrected images are used to form “facade map” of the building for engineers to inspect the building easily.