Development of concrete compaction management method using AI-based vibrator detector

Abstract

Manual concrete compaction using a vibrator depends on the experience and skills of skilled workers, and currently cannot be managed quantitatively. Quantitative compaction management in real time has become an urgent issue. Therefore, in order to manage the compaction position and time, it is necessary to recognize and detect the target vibrator.The purpose of this study is to develop an essential elemental technology to manage compaction position and time from images. In this study, the location and time will be determined by identifying the vibrating part of the vibrator and images of colored tape attached to the vibrator cord. Using three-dimensional image restoration technology, we calculated the spatial coordinates of feature points (vibrating parts and colored tape). Additionally, in order to recognize and detect features in real time, we have attempted to speed up the processing of huge amounts of information.