Application of Deep Learning in Globeland30-2010 Product Refinement

Abstract

GlobeLand30, as one of the best Global Land Cover (GLC) product at 30-m resolution, has been widely used in many research fields. Due to the significant spectral confusion among different land cover types and limited textual information of Landsat data, the overall accuracy of GlobeLand30 is about 80 %. Although such accuracy is much higher than most other global land cover products, it cannot satisfy various applications. There is still a great need of an effective method to improve the quality of GlobeLand30. The explosive high-resolution satellite images and remarkable performance of Deep Learning on image classification provide a new opportunity to refine GlobeLand30. However, the performance of deep leaning depends on quality and quantity of training samples as well as model training strategy. Therefore, this paper 1) proposed an automatic training sample generation method via Google earth to build a large training sample set; and 2) explore the best training strategy for land cover classification using GoogleNet (Inception V3), one of the most widely used deep learning network. The result shows that the fine-tuning from first layer of Inception V3 using rough large sample set is the best strategy. The retrained network was then applied in one selected area from Xi’an city as a case study of GlobeLand30 refinement. The experiment results indicate that the proposed approach with Deep Learning and google earth imagery is a promising solution for further improving accuracy of GlobeLand30.

Improving Dryland Urban Land Cover Classification Accuracy Using a Classical Convolution Neural Network

“…To date, numerous CNN models have been designed to tackle land cover classification problems; recent studies have proven that the GoogLeNet inceptionV3 model is effective and efficient for land cover type identification based on high-resolution imagery [43][44][45]. Compared with typical CNN models, inceptionV3 introduced the new concept of separable convolutional layers, which can reduce the number of computing parameters and significantly improve the feature learning speed [46][47][48].…”

“…To achieve this goal, typical dryland samples were extracted from a public benchmark dataset derived from Google Earth high-resolution imagery for model training. This dataset covered abundant labeled land types in China (i.e., cultivated land, forest, grassland, shrubland, water bodies, Land 2023, 12, 1616 6 of 20 artificial surfaces, bare land, and permanent snow and ice) [43]. First, 10,000 labeled images of each land type in typical dryland regions were randomly selected as training samples based on their geographical location (Figure 3).…”

Multi-Scenario Simulations of Land Use and Habitat Quality Based on a PLUS-InVEST Model: A Case Study of Baoding, China

“…The main data sources used in this study include the following: (1) Baoding grain production data and acreage data from the Baoding Statistical Yearbook and grain price data from the National Compilation of Cost and Benefit Information on Agricultural Products [30]; (2) land-use data from the Globeland30 global surface use database (http://www. globallandcover.com, accessed on 10 September 2022), in which the images for the land cover classification of development and update of GlobeLand30 are mainly 30 m multispectral images, including Landsat TM5 ETM+ [31]; (3) elevation data from the geospatial data cloud platform of the Computer Network Information Center of the Chinese Academy of Sciences (http://www.gscloud.cn, accessed on 10 September 2022), with a spatial resolution of 30 × 30 m; (4) temperature and precipitation data from the National Science and Technology Infrastructure Platform at the National Earth System Science Data Centre-Loess Plateau Sub-Center (http://loess.geodata.cn, accessed on 12 September 2022) [32]; (5) GDP and population data from the Resource and Environment Science Data Registration and Publication System; (6) roads, water, and railway data are cited from the National Geographic Information Resources Catalogue Service (www.webmap.cn). To unify the spatial accuracy of the data, the above data were processed by using the cropping and resampling tools of ArcGis 10.5 to convert them into 30 m × 30 m raster files.…”

Improving the Accuracy of Remote Sensing Land Cover Classification by GEO-ECO Zoning Coupled with Geostatistical Simulation

“…Classical supervised and unsupervised classification technologies are commonly used, but the accuracy is not high, which is about 60% to 70% [13,14]. A deep learning algorithm has not been applied to the classification of land cover products on a large scale [15]. The second-level land cover product classification is more difficult to obtain reliable results via automatic classification.…”