Time-Series Model-Adjusted Percentile Features: Improved Percentile Features for Land-Cover Classification Based on Landsat Data

Shuai Xie,Jiangning Yang,Liangyun Liu

doi:10.3390/rs12183091

Shuai Xie, Jiangning Yang + Show 1 more

Open Access

https://doi.org/10.3390/rs12183091

Copy DOI

Abstract

Percentile features derived from Landsat time-series data are widely adopted in land-cover classification. However, the temporal distribution of Landsat valid observations is highly uneven across different pixels due to the gaps resulting from clouds, cloud shadows, snow, and the scan line corrector (SLC)-off problem. In addition, when applying percentile features, land-cover change in time-series data is usually not considered. In this paper, an improved percentile called the time-series model (TSM)-adjusted percentile is proposed for land-cover classification based on Landsat data. The Landsat data were first modeled using three different time-series models, and the land-cover changes were continuously monitored using the continuous change detection (CCD) algorithm. The TSM-adjusted percentiles for stable pixels were then derived from the synthetic time-series data without gaps. Finally, the TSM-adjusted percentiles were used for generating supervised random forest classifications. The proposed methods were implemented on Landsat time-series data of three study areas. The classification results were compared with those obtained using the original percentiles derived from the original time-series data with gaps. The results show that the land-cover classifications obtained using the proposed TSM-adjusted percentiles have significantly higher overall accuracies than those obtained using the original percentiles. The proposed method was more effective for forest types with obvious phenological characteristics and with fewer valid observations. In addition, it was also robust to the training data sampling strategy. Overall, the methods proposed in this work can provide accurate characterization of land cover and improve the overall classification accuracy based on such metrics. The findings are promising for percentile-based land cover classification using Landsat time series data, especially in the areas with frequent cloud coverage.

Highlights

Earth observation data acquired by satellites are commonly utilized to map and monitor land covers [1], which is essential for research on biological diversities [2], wetland ecosystems management [3], and forest disturbances and recoveries [4,5]
We proposed the time-series model (TSM)-adjusted percentile features with the aim of characterizing land cover accurately and improving the classification accuracy substantially
Classification of Percentiles Derived from Multispectral Reflectance and Normalized Difference Vegetation Index (NDVI) Time Series

Summary

Introduction

Earth observation data acquired by satellites are commonly utilized to map and monitor land covers [1], which is essential for research on biological diversities [2], wetland ecosystems management [3], and forest disturbances and recoveries [4,5]. The statistical metrics that are extracted using Landsat time-series data over a single year or consecutive years provide a novel spectro-temporal feature space for Landsat-based land-cover classification. These spectro-temporal statistic measurements have been demonstrated as feasible tools for distinguishing land cover classes [8,9,10]. The 20th, 50th, and 80th percentiles composites derived from monthly global Web-enabled Landsat Data (WELD) were used for land-cover mapping of North America [14]. The 0th, 25th, 50th, 75th, and 100th percentiles derived from the WELD monthly composites were included in the spectral inputs that were used to generate the continuous field for land-cover classes of the United States [13]

Objectives

Methods

Results

Conclusion