AI Summary & Key Takeaways
AI InsightThe results shown that self-calibration of loop-close images enables the close results to the pre-calibration.
📌 3 Key Takeaways
Abstract
The camera calibration is an important issue that must be overcome to getting metric scene measurement. The imaging parameters are estimated by calibration of the camera. Basically, the camera calibration is performed individually from the photogrammetric evaluation. Today, 3-D point cloud generation and the camera calibration are usually attained simultaneously by using SfM approach photogrammetric evaluation. Stereo images that do not have camera intrinsic parameters can also be evaluated by SfM based photogrammetry. In this study, camera calibration models were investigated for point cloud generation of close-range photogrammetry. The results shown that self-calibration of loop-close images enables the close results to the pre-calibration. Otherwise, the images should be convergent as far as possible or projection-to-sparse point cloud ratio must be raised. The results show that the projection-to-sparse point cloud ratio of 13.22 created high accuracy to self-calibration. Consequently, the pre-calibration requires extra computation and time. However the self-calibration can be implemented for high accuracy measurement subject to convergence imaging or sufficient number of projection.