Bamfield Region UAV Imagery and Surface Model Data
Bibliographic record
Abstract
Bamfield UAV Data Capture April 9 & 10th 2016. Hakai Institute: Iain McKechnie, Will McInnes, and Keith Holmes. The Hakai Institute in partnership with the Bamfield Marine Science Centre (BMSC) have conducted UAV (unmanned aerial vehicle) missions in five locations around Bamfield, British Columbia, Canada. The missions took place April 9th and 10th, 2016. The main goal was to collect high spatial resolution imagery and surface models of key research areas for BMSC researchers. All UAV flights and data processing has been completed by the Hakai Institute and the aim is to collaborate research efforts to enhance the spatial knowledge of these key research areas. The low tide windows between 8:30 and 11:30 am on April 9th and 10th provided imagery with tides of .5 meters or less and image resolutions of between 3cm and 10cm. Weather was mixed and we had to improvise flights to compensate for low level clouds and fog. In the end we were able to accomplish all our major target areas. Data products that can be generated are the following: High resolution imagery mosaics Digital surface models LAS point cloud data 3D mesh data Raw imagery and video footage Georeferencing points and 2 survey markers for Diana Island Methodology: 1) Georeferencing targets were distributed throughout the image capture area. Targets are checkered black and white and 60cm x 60cm. 2) Phantom 3 Pro UAVs were used for all imagery acquisition. Flights were conducted with a grid pattern with greater than 50% overlap in imagery. Small area flights required less images like Wizard with 237 images processed and larger areas like Grappler requiring 504 images. Flights take roughly 15 minutes to conduct and cover up to 1.5 x 1.5 km in a single flight. Full specs for image capture are listed below. 3) After the flights are conducted the field team inspects the survey coverage using the DJI GO applicaton map. Once they are satisfied a georeferencing mission starts. Georeferencing requires the use of a Topcon DGPS GR5 survey grade GPS system. Targets placed before the flight and opportunistic targets (discussed later) are referenced to within 10 cm horizontal / 15 cm vertical resolution. DGPS soaking is used to enhance accuracy which takes anywhere from 15 seconds to 3 minutes to achieve acceptable levels of variance. Many of the targets achieved well below these numbers. It is essential that the targets are well dispersed. If some areas have not been covered well with targets or there was limited time pre flight to dispense targets opportunistic targets are used. Opportunistic targets are large stationary objects that are close to the ground and can easily be identified in imagery. Often driftwood or geographic features like large rocks or fissures are ideal in the field if no man made structures like cement pads are available. 4) Image processing is conducted using Pix4D software. Images are first inspected for quality (remove high glare or poor contrast images) and then the full data process takes place which includes a flight report, ray cloud / image tie points, triangulated mesh, mosaic and surface model. 5) Image mosaics and digital surface models are then processed in ArcGIS Desktop. Survey DGPS points from the field are used to link up the targets and a sline correction method is used because it assumes we have highly accurate survey data to correct the imagery with. The georeferencing pins are saved and processed again with the accompanying surface model. The final datasets are then exported to our servers for archiving.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.002 | 0.001 |
| Meta-epidemiology (narrow) | 0.003 | 0.002 |
| Meta-epidemiology (broad) | 0.002 | 0.000 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.001 | 0.003 |
| Scholarly communication | 0.001 | 0.009 |
| Open science | 0.007 | 0.011 |
| Research integrity | 0.002 | 0.002 |
| Insufficient payload (model declined to judge) | 0.000 | 0.022 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; both teacher heads agree on what is shown here.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".