Abstract: Ramming settlement is a crucial indicator for evaluating the quality of dynamic compaction foundation reinforcement. At present, it mainly relies on manual monitoring, which seriously affects the efficiency of dynamic compaction construction, and safety cannot be guaranteed. To address these issues, this study proposes a non-contact measurement method using monocular photogrammetry to accurately measure ramming settlement. The method begins by analyzing and scientifically defining the suitable definition of ramming settlement for monocular visual monitoring. Next, based on the fundamental principles of visual imaging, a numerical equation is developed to measure the three-dimensional attitude of the rammer. This is achieved by leveraging the rotation invariance of the rammer to identify geometric feature points and solve for their three-dimensional coordinates. Finally, the proposed method is then tested on the construction site, and the algorithm is optimized using the conditional adjustment method. The experimental and engineering application results demonstrate that this method offers high precision, fast processing speed, and robustness. It can effectively automate the calculation of ramming settlement, a key parameter for monitoring the quality of dynamic compaction. This research provides a new direction and means for advancing the current automatic monitoring technology of dynamic compaction.

Key words: ramming settlement monitoring, monocular photogrammetry, rammer pose, rotation invariance