Drones and machine learning for habitat monitoring

The increasing decline in the status of habitats, mainly due to anthropogenic stressors, has spurred the development and implementation of many conservation-related legislation. This legislation involves mapping, a critical task to gather important information about the habitats, including their locations, spatial extents and changes over time. Advancements in mapping technologies, such as drones and machine learning, can be integrated with conventional field surveys to improve the efficiency and effectiveness of mapping habitats. The study assessed the effectiveness of drone-acquired data and machine learning as tools for accurate and detailed mapping of highly dynamic and fine-scale mosaics of habitats in a coastal dune environment. Drone imagery and field data were acquired in a dune site in Kerry, Ireland, during the growing season in 2020: May (early), July (mid) and October (late). Topographic data representing the terrain of the site were also generated during the photogrammetric processing of the images. These datasets were then processed and analysed using the Random Forest machine learning technique to classify dune habitats at this site. The results showed that using multiple drone datasets acquired throughout the vegetation growing season achieved higher classification accuracy compared to using just a single dataset (92.37% vs. 84.09%, respectively). Also, including topographic data consistently improved the accuracy, regardless of the number of datasets. Comparing the three drone-acquired datasets, the analysis suggested that the dataset acquired in the middle period of the growing season, i.e., the flowering period, was better than those acquired in the early or late periods for dune habitat mapping.

A critical aspect of habitat conservation is tracking the location and expansion of invasive species, which is considered a major threat to and pressures on habitats. This study also explored the potential of utilising drone imagery and deep learning (DL) techniques for mapping invasive species, including those in the early stages of invasion, i.e., occurring in small patches. However, creating a robust DL model is challenging due to the requirement for large and diverse labelled training data. The study implemented a DL semantic segmentation on drone imagery and investigated the potential of applying data augmentation and pseudo-labelling to increase the amount and diversity of labelled data. Results showed that DL-based segmentation achieved high accuracy (mean Intersection-Over-Union [mIOU] score=0.832). The model trained on the augmented and pseudo-labelled data achieved an mIOU score of 0.712 on an independent dataset, while there was a decrease of 0.158 in model performance when only the original labelled data were used. This result suggests the potential of using data augmentation and pseudo-labelling techniques in creating more robust models.

Overall, the combined use of high-resolution drone data and machine learning techniques offers massive potential for repeatable and systematic approaches to fine-scale habitat characterisation. The resulting detailed maps from these approaches can provide critical information to guide and inform habitat conservation efforts. Moreover, these maps can support and contribute to the evidence-based implementation of SDG 15, which focuses on “protecting, restoring and promoting sustainable use of terrestrial ecosystems and preventing biodiversity loss”.