The following process was used to generate the dense point cloud. The images used were acquired from a DJI Phantom 3 Professional flown over Black Beach, Falmouth on 18 March 2016. Eighteen ground control points (GCPs) were incorporated in the photogrammetric processing. Details and locations of the images and GCPs are provided by additional datasets in the larger work citation. The processing was performed using Agisoft Photoscan Professional v. 1.2.6 build 2834 (64 bit) software. The computer was a Mac Pro running OS X 10.11.6 with 8-Core Intel Xeon E5 CPUs running at 3 GHz with 128 GB RAM.
Add photos
1) Using the “Add photos…” tool, all 250 images in bb_20160318_UAS_images.zip (larger work) were added to a single “chunk” (Agisoft terminology). Image quality for the photos was estimated using “Estimate image quality…”. The resulting image quality metrics (which are relative non-dimensional measures in which a value of 1 indicates that the image does not have obvious blurring) ranged from 0.69 to 1.26.
2) Using ”Convert”, the coordinate system of the images (called “cameras” in Photoscan) was converted from native GPS geographic units (latitude/longitude, assumed to be in the WGS84 coordinate system) to meters in NAD83/UTM zone 19N (EPSG:26919) using the toWGS84 transformation in the NAD83 well-known text (WKT). Camera location accuracy was left at the default 10 m (found in Reference Settings on the Reference Pane).
Ground control points
3) “Detect Markers” was used to automatically identify targets in the photos, with settings “Cross (non-coded)” and a tolerance of “50” (on a scale of zero to 100, with 100 being the least discriminating). All 18 of the 4-ft square black and white targets deployed were automatically detected. The automatically-generated marker labels were added to the GCP location file (see larger work) by viewing the target label in the images and with reference to a map of the labeled GCPs.
3) “Import markers” was used to load the GCP location file (see larger work), which assigned precise GCP coordinates (northing, easting, and elevation in UTM Zone 19 North meters in NAD83 and NAVD88 coordinate systems) to the detected markers, and placed new markers for the GCPs that had not been auto-detected.
4) The locations of all markers were established and verified in all of the images in which they appeared, except when the image of the target was so poor that the reference point on the target could not be precisely determined. This was a manual process aided by the ability of the software to identify images in which each marker appeared and to maintain a centered view at constant zoom level across all of those images. Each of those images was inspected to verify and adjust the precise marker placement. Manual placement was a painstaking and somewhat subjective process that introduced slight uncertainties into the GCP location in the images. However, our experience indicates that addition of GCPs and pinpointing targets in as many images as possible improves the final alignment of the point cloud. The tie-point accuracy was set to 1 pixel in “Reference Settings”.
Initial alignment
5) ”Align Photos” was selected to align all of the cameras using the following settings: Accuracy: “High” (which did not downsample the images); Pair selection: “Reference” (which used GPS information identify nearby images when searching for tie points); Key point limit: 5,000; Tie point limit; 0 (unlimited). Adaptive camera model fitting option was selected. 211 cameras (images with varying viewpoints) were initially aligned.
6) ”Optimize Cameras” was used to perform initial lens calibration and camera alignment. Lens-calibration parameters f, cx, cy, k1, k2, k3 were included; higher-order parameters k4, b1, b2, p1, p2, p3, and p4 were not. These parameters define focal length (f), pixel coordinates of the principal point (cx, cy), and certain radial distortion coefficients (k1, k2, k3, k4, p1, p2, p3, and p4). The software generates a metric for assessing model fit called the standard unit weight error (SUWE). Values close to 1.0 are optimal. The initial SUWE was 0.165 and the overall alignment error for the cameras was 35.26 m.
Refinement of the sparse point cloud
The sparse point cloud representing tie points among the images consisted of approximately 690,000 points. An iterative method developed by Tommy Noble and used by Sherwood (2017) was used to identify and remove lower-quality tie points with the “Gradual Selection” tool. Iterations were performed with the following criteria and target values.
* Reconstruction uncertainty – Quality based on the geometry of the reconstruction. A dimensionless ratio of the maximum/minimum axes of the three-dimensional ellipse describing reconstruction uncertainty based on ray triangulation (target was 10)
* Projection accuracy – Quality of pixel matching among images. A weighted ranking (1 is best, larger numbers worse) based on the size and sharpness of tie-points (target was 3)
* Reprojection error – Estimate of residual error in tie-point location. A measure (pixels) of the precision of calculated tie-point locations based on the geometry (target was 0.3 pixels).
“Gradual Selection” was used and the target value was set, but if more than about 20% of the points were flagged at that setting, the threshold was adjusted to select only about 10% of the points. (The total number of points and the number of flagged points was shown on screen as selections were made). Selected points were deleted, and camera settings were optimized before the next iteration. After each iteration, the improvement in accuracy was assessed by checking the marker error for ground control points. This procedure was repeated three times for each criterion listed above (in order; i.e. points were selected based on reconstruction uncertainty three times before next selecting by projection accuracy).
When complete, approximately 26 percent of the points had been removed, leaving 581,000 tie points, and the following values for the target metrics were obtained:
* Reconstruction uncertainty - 10 (no units)
* Projection accuracy - 6 (no units)
* Reprojection error - 0.45 (pixels)
The marker error for the ground control points was reduced to 0.083 m (0.43 pixels) from 0.094 m (0.58 pixels).
As a final alignment optimization/tie point refinement step, we manually deleted outlying tie points, such as those that were clearly above or below ground. We did so conservatively. These points were identified as those that did not follow the known structure of the landscape, and were usually caused by photos in which water was moving and/or inconsistently obstructed the view of submerged areas. With these changes, the updated (and final) optimization produced an estimated marker error of 6 cm for about 500,000 tie points remaining in the sparse point cloud. From beginning to end, optimization removed about 190,000 tie points, 27 percent of the original sparse cloud. This reduced the total error by 90 percent, from 62 to 6 cm.
Dense point cloud
“Build Dense Cloud” was invoked with “High” quality and “Mild” depth filtering to generate a dense point cloud. "Export points" was used to export the point cloud in .LAZ format with coordinates referenced to NAD 83/UTM Zone 19N. The resulting dense point cloud containing 4,345,647 points is the associated data product.
Estimate uncertainty of point cloud
Uncertainty in the location of points in the dense point cloud is, in general, the quadrature sum of a) uncertainty in the locations of the ground control points (GCPs) to which the point cloud is referenced; b) uncertainty in the geometric reconstruction represented by the sparse point cloud, which includes uncertainty in the location of tie points, camera locations, camera look angles, and camera lens calibrations, assuming the GCP locations are exact; and c) interpolation errors associated with placing the dense-cloud points in the geometric reconstruction, which arise when the locations of dense-cloud points between sparse-cloud points differ from the real-world locations. The horizontal and vertical precision of the surveyed GCP locations (a), was estimated as the average of the two uncertainties reported for every GCP, which were +/- 0.011 m horizontally and +/- 0.014 m vertically. Our replacements for (b) were the horizontal and vertical RMS errors associated with reconstruction of the GCP marker locations reported by Photoscan (+/- 0.06 m and +/- 0.024 m, respectively). In lieu of values for (c), we combined the reported unscaled reprojection error reported by Photoscan (0.45 pixels) with the resolution of the images (1 pixel equaled approximately 0.025 m in nadir views) to derive a reprojection error of 0.01 m. These combined uncertainty estimates (root sums of squares of the uncertainty terms) are +/- 0.057 m horizontal and +/- 0.03 m vertical.
