Adaptive Threshold Model in Google Earth Engine: A Case Study of Ulva prolifera Extraction in the South Yellow Sea, China

Abstract

An outbreak of Ulva prolifera poses a massive threat to coastal ecology in the Southern Yellow Sea, China (SYS). It is a necessity to extract its area and monitor its development accurately. At present, Ulva prolifera monitoring by remote sensing imagery is mostly based on a fixed threshold or artificial visual interpretation for threshold selection, which has large errors. In this paper, an adaptive threshold model based on Google Earth Engine (GEE) is proposed and applied to extract U. prolifera in the SYS. The model first applies the Floating Algae Index (FAI) or Normalized Difference Vegetation Index (NDVI) algorithm on the preprocessed remote sensing images and then uses the Canny Edge Filter and Otsu threshold segmentation algorithm to extract the threshold automatically. The model is applied to Landsat8/OLI and Sentinel-2/MSI images, and the confusion matrix and cross-sensor comparison are used to evaluate the accuracy and applicability of the model. The verification results show that the model extraction of U. prolifera based on the FAI algorithm has higher accuracy (R2 = 0.99, RMSE = 5.64) and better robustness. However, when the average cloud cover is more than 70% in the image (based on the statistical results of multi-year cloud cover information), the model based on the NDVI algorithm has better applicability and can extract the algae distributed at the edge of the cloud. When the model uses the FAI algorithm, it is named FAI-COM (model based on FAI, the Canny Edge Filter, and Otsu thresholding). And when the model uses the NDVI algorithm, it is named NDVI-COM (model based on NDVI, the Canny Edge Filter, and Otsu thresholding). Therefore, the final extraction results are generated by supplementing NDVI-COM results on the basis of FAI-COM extraction results in this paper. The F1-score of U. prolifera extracted results is above 0.85. The spatiotemporal distribution of U. prolifera in the South Yellow Sea from 2016 to 2020 is obtained through the model calculation. Overall, the coverage area of U. prolifera shows a decreasing trend over the five years. It is found that the delay in recovery time of Porphyra yezoensis culture facilities in the Northern Jiangsu Shoal and the manual salvage and cleaning-up of U. prolifera in May are among the reasons for the smaller interannual scale of algae in 2017 and 2018.

Urbanization shifts long‐term phenology and severity of phytoplankton blooms in an urban lake through different pathways

“…It is difficult to choose a reasonable threshold for the index model to extract bloom regions, and any choice of a fixed threshold is a compromise (Shi, Zhang, Qin, et al, 2019). Because long‐term satellite images may be affected by different sensors, environmental changes, and different atmospheric conditions, the fixed threshold may not be applicable to images at all time nodes (Zhang, Wu, et al, 2021). For example, the fixed threshold may overestimate the area of PBs, which is inconsistent with reality (Zhang, Wu, et al, 2021).…”

“…Field validation for the area of PBs has proven to be extremely difficult (Hu et al, 2010; Wang & Hu, 2021), especially for the long‐term and large‐scale area data of PBs (Fang et al, 2022; Hou et al, 2022). Since PBs in the visible and near‐infrared bands have spectral characteristics similar to terrestrial vegetation, the results of the visual interpretation of objectively present PBs in satellite images can be used as ground truth (Qi et al, 2016; Zhang, Wu, et al, 2021), and it has been widely used to verify the area of PBs (Hu et al, 2010; Mu et al, 2019; Wang & Hu, 2021). In this study, different satellite images from each year were selected for the extraction of the PB regions of interest to verify the accuracy of the OTSU threshold automatic segmentation algorithm, and the root mean square error (RMSE) and relative error (RE) were used to evaluate the results (Figure S4).…”

“…Because long-term satellite images may be affected by different sensors, environmental changes, and different atmospheric conditions, the fixed threshold may not be applicable to images at all time nodes (Zhang, Wu, et al, 2021). For example, the fixed threshold may overestimate the area of PBs, which is inconsistent with reality (Zhang, Wu, et al, 2021). Thus, we need to manually set an optimal threshold for each image to remove this effect and ensure that bloom regions are extracted with high accuracy for each image in the long-term series.…”

Extraction of Aquaculture Pond Region in Coastal Waters of Southeast China Based on Spectral Features and Spatial Convolution

“…Otsu's method is an automatic threshold segmentation algorithm based on the firstorder statistical characteristics of the gray histogram. This method is not affected by image brightness and contrast, has fast operation speed and high storage efficiency, and is widely used in the research of image threshold segmentation [45,46]. We used Otsu's method to obtain the optimal threshold of water and non-water segmentation in the test area.…”

Automatic Detection of Floating Macroalgae via Adaptive Thresholding Using Sentinel-2 Satellite Data with 10 m Spatial Resolution

“…The combination of images from high-spatial-resolution satellite data and Otsu's method excels in comparison with satellite data with various spatial resolutions. Finally, Zhang et al [24] used Otsu's method to detect floating macroalgae in the South Yellow Sea area using Landsat-8 and Sentinel-2 satellite data. FAI images of Sentinel-2 satellite data contain only a small number of environmental factors, such as thin clouds and sun glint, so this index exhibits high accuracy and stable results.…”