A study on multi-dimension error compensation and application to the foundation pit monitoring

摘要

Error compensation is the key issue during camera calibration with displacement measurements. The overall error can be reduced significantly after compensation and different data intervals can be selected to reduce the calibration time while maintaining a high level of accuracy. In this study, we propose a multidimensional error compensation method that considers the error due to multiple causes. First, a multi-objective optimization method is used to optimize center recognition based on the k-means clustering algorithm. The algorithm has three steps comprising noise removal for the best co-ordinates, multi-target area identification, and target center calculation. Local and global features are then employed in a support vector regression model to estimate the error compensation. We applied the method to improve the detection of unexpected deformation and movements in a foundation pit using a real dataset obtained from the foundation pit monitoring system for the Zhoushan industry pack project. The results demonstrated that the centers determined after compensation were closer to the actual target than the original centers.