| Title | Leveraging AI to estimate caribou lichen in UAV orthomosaics from ground photo datasets |
| |
| Author | Richardson, G ;
Leblanc, S G; Lovitt, J; Rajaratnam, K; Chen, W |
| Source | Drones vol. 5, issue 3, 2021 p. 1-16, https://doi.org/10.3390/drones5030099  Open Access |
| Image |  |
| Year | 2021 |
| Alt Series | Natural Resources Canada, Contribution Series 20210220 |
| Publisher | MDPI |
| Document | serial |
| Lang. | English |
| Media | on-line; digital |
| File format | pdf; html |
| Subjects | geophysics; Science and Technology; Nature and Environment; remote sensing; photogrammetric surveys; photography; mapping techniques; spectral analyses; models; Lichen; Cladonia rangiferina; caribou
lichen; Methodology; drones; Artificial intelligence; Classification |
| Illustrations | digital images; aerial photographs; flow diagrams; models; tables; plots |
| Program | Canada Centre for Remote Sensing Remote Sensing Science Program - Optical methods and applications |
| Released | 2021 09 17 |
| Abstract | Relating ground photographs to UAV orthomosaics is a key linkage required for accurate multi-scaled lichen mapping. Conventional methods of multi-scaled lichen mapping, such as random forest models and
convolutional neural networks, heavily rely on pixel DN values for classification. However, the limited spectral range of ground photos requires additional characteristics to differentiate lichen from spectrally similar objects, such as bright logs.
By applying a neural network to tiles of a UAV orthomosaics, additional characteristics, such as surface texture and spatial patterns, can be used for inferences. Our methodology used a neural network (UAV LiCNN) trained on ground photo mosaics to
predict lichen in UAV orthomosaic tiles. The UAV LiCNN achieved mean user and producer accuracies of 85.84% and 92.93%, respectively, in the high lichen class across eight different orthomosaics. We compared the known lichen percentages found in 77
vegetation microplots with the predicted lichen percentage calculated from the UAV LiCNN, resulting in a R2 relationship of 0.6910. This research shows that AI models trained on ground photographs effectively classify lichen in UAV orthomosaics.
Limiting factors include the misclassification of spectrally similar objects to lichen in the RGB bands and dark shadows cast by vegetation. |
| Summary | (Plain Language Summary, not published)
Caribou population in Canada have been generally declining since the mid 1980's. Several aspects can contribute to the decline. One of these aspects may
be food availability. The main source of winter food for caribou is lichen, but lichen availability is not well know in all of Canada. Satellite images have a potential to be used to map lichen, but so far lichen has been found to be difficult to map
in some areas of Canada. To improve the mapping techniques, this study adapts a method using artificial intelligence developed for field photography to images taken with drones in order to map lichen locally over areas of about 15-20 ha (37-50
acres). The maps made by drones will later be used to improve other maps covering larger areas. |
| GEOSCAN ID | 328790 |
|
|