Object-based analysis of UAS imagery to map emergent and submerged invasive aquatic vegetation: a case study

Abstract

Small unmanned aircraft systems (UAS) combined with automated image analysis may provide an efficient alternative or complement to labour-intensive boat-based monitoring of invasive aquatic vegetation. A small mapping drone was assessed for collecting high-resolution (≤5 cm/pixel) true-colour and near-infrared imagery revealing the distribution of invasive water soldier (Stratiotes aloides) in the Trent–Severn Waterway, Ontario (Canada). We further evaluated the capacity of an object-based image analysis approach based on the Random Forests classification algorithm to map features in the imagery, chiefly emergent and submerged water soldier colonies. The imagery contained flaws and inconsistencies resulting from data collection in suboptimal weather conditions that likely negatively impacted classification performance. Nevertheless, our best-performing classification had a producer’s and user’s accuracy for water soldier of 81% and 74%, respectively, an overall accuracy of 78%, and a kappa value of 61%, indicating “substantial” accuracy. This trial provides an instructive case study on results achieved in a “real-world” application of a UAS for environmental monitoring, notably characterized by time constraints for data collection and analysis. Beyond avoiding data collection in unfavourable weather conditions, adaptations of the image segmentation process and use of a true discrete-band multispectral camera may help to improve classification accuracy, particularly of submerged vegetation.

Open-Source Processing and Analysis of Aerial Imagery Acquired with a Low-Cost Unmanned Aerial System to Support Invasive Plant Management

“…In general, this approach can allow environmental managers to monitor the distribution and spread of invasive plants in similar situations once the classification models are adjusted to the respective species. Environmental managers should be strongly encouraged to use low-cost UAS approaches to avoid high costs and timeconsuming field surveys, especially in sites of low accessibility (Wan et al, 2014;Chabot et al, 2016;Hill et al, 2016;Müllerová et al, 2017). Minimizing in situ mapping and control operations is also an added advantage for using UAS, because reduction in habitat disturbance can encourage the recovery of native plants that are threatened by biological invasion (Mack and D'Antonio, 1998;Huston, 2004).…”

“…Unmanned Aerial Systems (UAS), popularly known as drones, introduce a new remote sensing technique that may become an applicable and affordable alternative to conventional approaches, as they reduce costs and increase the spatial resolution of aerial images (Wan et al, 2014;Dvořák et al, 2015;Chabot et al, 2016;Hill et al, 2016;Müllerová et al, 2016Müllerová et al, , 2017. The technical development, component miniaturization, and increased sales in recent years resulted in the rapid growth of UAS as an environmental remote-sensing platform.…”

Drones detect illegal and derelict crab traps in a shallow water estuary

“…Drones (also known as RPAS or UAV) have recently become sophisticated and affordable to create large‐scale impacts to various commercial and scientific operations, such as fauna surveys (Ezat, Fritsch & Downs, ; Kelaher et al., ), agriculture (Rey‐Caramés, Diago, Martín, Lobo & Tardaguila, ), anti‐poaching (Mulero‐Pazmany, Stolper, van Essen, Negro & Sassen, ), search and rescue (Karaca et al., ), engineering inspections (Omar & Nehdi, ) and habitat mapping (Ventura, Bruno, Jona Lasinio, Belluscio & Ardizzone, ; Chabot, Dillon, Ahmed & Shemrock, ). Drones are also being incorporated into fisheries management, but their use is still largely experimental, rather than part of routine operating procedures (Kopaska, ; Brooke et al., ; Toonen & Bush, ).…”

Assessment of Aquatic Weed in Irrigation Channels Using UAV and Satellite Imagery

“…They used the Support Vector Machine (SVM) algorithm to classify images, with a weed expert providing the classification training set. An unmanned aerial vehicle (UAV) with red-green-blue (RGB) and near-infrared cameras was used to map the invasive water soldier weed in [13], using object-based image analysis and the random forest algorithm. They selected training samples throughout the area based on visual interpretation with four classes (water soldier emergent and submerged, native vegetation and other) and achieved a classification kappa of 61%.…”