ENG
✕

How to Evaluate Data Quality in the Relative and Absolute Orientation Stages？

Relative Orientation Stage

  1. Camera Spatial Pose Check

– For circular shooting, the spatial distribution of cameras should form a circle.

  1. Sparse Point Cloud Check:

– Sparse point clouds should show no obvious misalignment.

– For objects with fewer surface features, dense reconstruction can help assess the point cloud quality.

 

After confirming that the relative orientation results are accurate:

– Proceed to absolute orientation and align individual photos to refine the global alignment.

– After completing register, return to the absolute orientation stage to randomly check the camera alignment. If there are no issues, you can proceed to the texturing stage.

 

Absolute Orientation Stage

  1. Photo Alignment Check:

– If most photos are correctly aligned but a few are not, it may indicate issues with the relative orientation results.

  1. Overall Alignment Check:

– If all photos fail to align, it likely indicates errors in absolute orientation.

– For objects with fewer surface features, consider using point cloud and model matching to achieve absolute orientation.

 

By following these steps, you can systematically ensure the quality of your data.