03/05/2026
LiDAR SLAM or Visual SLAM — which would you choose?
This is a question many of us in mapping and 3D reconstruction face. But recently, I had an experience that really reshaped my perspective.
I processed a dataset captured with Tersus’ latest CaptuRTK using only a smartphone and Nuwa data collection App. The dataset included 500+ images covering ~500 m² — all processed offline on the phone (Snapdragon Gen2, 12GB RAM).
The result?
A high-quality point cloud that clearly shows how far survey-grade Visual SLAM + RTK has come.
We’re reaching a new level.
👉 For scenarios like:
Measuring hard-to-reach points
Volume calculations
Small-area outdoor point cloud generation
CaptuRTK — paired with a global shutter camera — can already deliver survey-grade accuracy, with incredible portability and efficiency.
👉 But if your work involves:
Frequent point cloud acquisition
Indoor or complex environments
Maximum robustness and consistency
A LiDAR solution like S1 is still the more reliable choice.
What stands out to me is this shift:
We’re moving from device-driven decisions to scenario-driven choices.
It’s no longer LiDAR vs. Visual —
It’s about using the right tool for the right job.
If you haven’t tried generating point clouds from imagery yet, now is a great time. The technology has matured to the point where even a smartphone can handle high-quality processing.
