The LiDARMill Spatial Fuser processing report is intended for data processors to glean insight into details pertaining to specific processing pipeline configurations. This report focuses primarily on sensor calibration and data optimization details contributing to the the quality and accuracy of deliverable data products.
This section outlines key processing parameters
This section displays information pertaining to the Camera sensor used in the project. The # represents the number of sensors in the mission. The count begins at 0 and increases depending on the number of Camera sensors (e.g. Camera 0, Camera1, etc).
This section displays information pertaining to the LiDAR sensor used in the project. The # represents the number of sensors in the mission. The count begins at 0 and increases depending on the number of LiDAR sensors (e.g. LiDAR 0, LiDAR 1, etc).
Optimization/Calibration Options:
Trajectory optimization: This tool performs feature matching within overlapping swaths of LiDAR data to determine correction offsets that are applied to the mission's trajectory to improve point cloud relative accuracy. Enabled/disabled within Fuser pipeline configuration.
Sensor Calibration: This tool applies an angular correction to LiDAR sensor (pitch, yaw, roll correction) to resolve misalignments from IMU to sensor. Depending on the LiDAR model, an additional ranging scale correction, tilt angle offset, or encoder calibration correction may be calculated and applied. Enabled/disabled within Fuser pipeline configuration.
Relative accuracy, the measure of how well overlapping flightlines match each other, is determined for the mission(s). Surface models are developed for each flightline. Relative accuracy is calculated from these surfaces using two metrics, magnitude and dZ. Magnitude is the average of the absolute values of the vertical offsets between a single flightline surface and points from overlapping flightlines. dZ is the average value of the vertical offsets between a single flightline surface and the points from overlapping flightlines. An average magnitude for all flightlines represents the project's overall relative accuracy.
A histogram showing the distribution of vertical distance error residuals between correspondences in overlapping strips of LiDAR. Correspondences are utilized for LiDAR calibration and/or optimization. The combination of a high correspondence count, low correspondence distance errors, and well distributed surface normal orientations is a good indication of a high quality calibration and/ or optimization.
A plot showing the surface normal orientations that were sampled during the Lidar calibration and/or optimization procedure. This is a quality control plot that represents the robustness of observations within the dataset used to perform LiDAR calibration and/or optimization.
Plot of the laser return count per laser as a function of time.
Plot of the distribution of reflectance values per laser.
Plot of the distribution of laser return indexes (first, second, third, etc).
The first plot shows the difference between the Navlab post processed trajectory and the LiDARMill optimized trajectory in terms of north, east and up positional difference as a function of time. Optimized trajectories include vertical translations per flightline interval to improve relative accuracy. Optimized trajectories also account for survey CONTROL points, and a single vertical translation is applied to the post processed trajectory when CONTROL points are enabled, in order to best fit the LiDAR pointcloud to ground control.
The second plot shows the difference between the Navlab post processed trajectory and the LiDARMill optimized trajectory in terms of roll, pitch and heading as a function of time.
Pointcloud colored by corresponding RGB pixel values from mission imagery.
Pointcloud colored by corresponding laser intensity.
Pointcloud colored by corresponding ellipsoidal elevation values.
Pointcloud colored by distances above ground model derived from ground classified points
Pointcloud colored by strip number.
Pointcloud colored by number of overlapping strips.
Pointcloud colored by point density
Pointcloud colored by Classification (top-down view).
Pointcloud colored by classification (bottom-up view).
Digital surface model with overlaid GCPs colored by dZ value.
