Near Infrared Imaging and Image Pre-Processing to Improve the Automatic Detection of Canada Geese
Bibliographic record
Abstract
Migratory shorebirds populations are adversely affected by climate change and loss of habitat thus careful monitoring of their populations is important for early detection of population loss.Current counting methods generally rely on intrusive and time-consuming manual identification.This work is part of a larger project to develop automated classification and counting methods using a remotely piloted aircraft system (RPAS).In addition to the use of RPAS, this work will also investigate if near-infrared (NIR) imaging captured by the RPAS yields detection improvements.Healthy vegetation reflects NIR wavelengths of light which can potentially create a greater contrast between an object and the surrounding vegetation.Preprocessing NIR raw images to enhance the contrast between vegetation and Canada geese (Branta canadensis) to improve object detection using the convolutional neural network (CNN) YOLOv4-Tiny have been investigated in this study.Training was done on a small dataset (423 and 1,269 images), a large dataset (2,000 images), artificially generated dataset (2,000 images) and hybrid datasets consisting of real and artificially generated images of Canada geese (4,000 and 5,000 images).A RPAS was used to obtain greyscale and NIR test images of geese decoys using the RPAS onboard camera and a NIR specific camera at varying altitudes.The NIR preprocessed ground test images showed detection improvements in both number of detections and confidence score percentages when validated against the YOLOv4-Tiny detector that was trained on an augmented dataset and the hybrid datasets.However, the greyscale aerial test images generally outperformed the NIR pre-processed images.
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 machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.001 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.001 | 0.001 |
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; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
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".