2016/12/11 by Jie Wang, Luyan Ji, Wang, Jie +9
Earth and Planetary Sciences · Engineering · Environmental Science · #Applications (stat.AP) #Computer Vision and Pattern Recognition (cs.CV) #FOS: Computer and information sciences #Remote Sensing and Land Use #Remote Sensing in Agriculture #Remote-Sensing Image Classification
paper · pdf · doi:10.48550/arxiv.1612.03373
openalex publication_date 2016/12/11 · openalex created_date 2022/10/04 · openalex updated_date 2026/07/28
There is a trend to acquire high accuracy land-cover maps using multi-source\nclassification methods, most of which are based on data fusion, especially\npixel- or feature-level fusions. A probabilistic graphical model (PGM) approach\nis proposed in this research for 30 m resolution land-cover mapping with\nmulti-temporal Landsat and MODerate Resolution Imaging Spectroradiometer\n(MODIS) data. Independent classifiers were applied to two single-date Landsat 8\nscenes and the MODIS time-series data, respectively, for probability\nestimation. A PGM was created for each pixel in Landsat 8 data. Conditional\nprobability distributions were computed based on data quality and reliability\nby using information selectively. Using the administrative territory of Beijing\nCity (Area-1) and a coastal region of Shandong province, China (Area-2) as\nstudy areas, multiple land-cover maps were generated for comparison.\nQuantitative results show the effectiveness of the proposed method. Overall\naccuracies promoted from 74.0% (maps acquired from single-temporal Landsat\nimages) to 81.8% (output of the PGM) for Area-1. Improvements can also be seen\nwhen using MODIS data and only a single-temporal Landsat image as input\n(overall accuracy: 78.4% versus 74.0% for Area-1, and 86.8% versus 83.0% for\nArea-2). Information from MODIS data did not help much when the PGM was applied\nto cloud free regions of. One of the advantages of the proposed method is that\nit can be applied where multi-temporal data cannot be simply stacked as a\nmulti-layered image.\n