Deploying DJI Matrice 300 RTK & LiDAR in Bangladesh's Deltaic Terrain

Vlog Overview & Transcript

Bangladesh's deltaic river channels and floodplains feature highly dynamic terrain under constant monsoonal influence. Capturing accurate elevation models is critical for flood mitigation, highway design, and bridge planning. However, dense foliage and aquatic plants pose severe barriers to standard optical camera drone photogrammetry.

Optical cameras cannot pierce thick canopies; they only capture the top leaf layers, leading to digital elevation models that skew vertical accuracy by several meters in forested tracts. To solve this, we deploy active light detection and ranging (LiDAR) sensors on a DJI Matrice 300 RTK platform.

LiDAR works by emitting millions of laser pulses per second, which pass through small gaps in leaves and bounce back from the underlying soil. By classifying these returns, we strip away vegetation layers mathematically to reveal the true bare-earth Digital Terrain Model (DTM).

We set up physical geodetic Ground Control Points (GCPs) across the survey area using dual-frequency Trimble R12 GNSS receivers, tied directly to the Survey of Bangladesh (SOB) Geodetic Control Network monument datum points. These control points allow us to calculate the absolute vertical and horizontal error limits of the LiDAR scans. The resulting vertical absolute accuracy matches ASPRS Class 1 mapping thresholds (under 5 centimeters), providing engineering-grade topo data for structural designs.