Semi-supervised segmentation for coastal monitoring seagrass using RPA imagery

Abstract

Intertidal seagrass plays a vital role in estimating the overall health and dynamics of coastal environments due to its interaction with tidal changes. However, most seagrass habitats around the globe have been in steady decline due to human impacts, disturbing the already delicate balance in environmental conditions that sustain seagrass. Miniaturization of multi-spectral sensors has facilitated very high resolution mapping of seagrass meadows, which significantly improve the potential for ecologists to monitor changes. In this study, two analytical approaches used for classifying intertidal seagrass habitats are compared: Object-based Image Analysis (OBIA) and Fully Convolutional Neural Networks (FCNNs). Both methods produce pixel-wise classifications in order to create segmented maps, however FCNNs are an emerging set of algorithms within Deep Learning with sparse application towards seagrass mapping. Conversely, OBIA has been a prominent solution within this field, with many studies leveraging in-situ data and multiscale segmentation to create habitat maps. This work demonstrates the utility of FCNNs in a semi-supervised setting to map seagrass and other coastal features from an optical drone survey conducted at Budle Bay, Northumberland, England. Semi-supervision is also an emerging field within Deep Learning that has practical benefits of achieving state of the art results using only subsets of labelled data. This is especially beneficial for remote sensing applications where in-situ data is an expensive commodity. For our results, we show that FCNNs have comparable performance with standard OBIA method used by ecologists, while also noting an increase in performance for mapping ecological features that are sparsely labelled across the study site.

Pseudo-label based and transformation-consistent self-ensembling model for underwater life images

“…However, traditional convolutional neural networks require a large amount of labeled data to obtain relatively excellent performance. For traditional CNN, a large amount of labeled data often means time-consuming and labor-intensive [11,12,13]. So subsequent researchers began to study the semisupervised learning framework.…”

“…His method exploits unreliable pseudo-labels from more layers and achieves the best performance [16]. Ma Zhiyuan et al proposed a different underwater semi-supervised semantic segmentation network (US-Net), which utilizes both pseudo-label and semi-supervised boundary detection sub-networks, using the boundary detection self-network to improve the segmentation effect , and the pseudo-label generator can use unstandardized data for learning, which achieved better performance [17]. K Sohn et al proposed to use a simple interpolation method to introduce disturbances, and conduct unsupervised training through consistent regularization and pseudo-label technology, which greatly improved the prediction performance of the model [18].…”

Biological traits approaches in benthic marine ecology: Dead ends and new paths

“…New data can be collected via innovative technology and methods development; for example, benthic sampling with vehicles such as AUVs/ROVs could record in situ morphological traits, size, position, body form, etc. ; passive and active acoustic monitoring and sensors could be used to record movement rates; remotely piloted aircraft (i.e., drones) have potential for obtaining high‐resolution images of intertidal benthos (e.g., Chand et al, 2020; Hobley et al, 2021). Some of these technologies are still expensive, but low‐cost technologies are emerging and gaining traction.…”

Evaluation of Semi-Supervised Learning for CNN-Based Change Detection

“…A very recent application of the mean teacher training scheme in a remote sensing setting was reported by (Hobley et al, 2021), where the training scheme is used to train a Fully Convolutional Network for seagrass monitoring from Remotely Piloted Aircraft (RPA) Very High Resolution (VHR) imagery. The method was compared to a fully supervised training setup as well as to an Object-based Image Analysis (OBIA) approach, resulting in improved results compared to the fully supervised setting but still not as good as the results achieved using OBIA.…”