The use of very-high-resolution images to extract urban, suburban and rural roads has important application value. However, it is still a problem to effectively extract the road area occluded by roadside tree canopy or high-rise buildings to maintain the integrity of the extracted road area, the smoothness of the sideline and the connectivity of the road network. This paper proposes an innovative Cascaded Attention DenseUNet (CADUNet) semantic segmentation model by embedding two attention modules, such as global attention and core attention modules, in the DenseUNet framework. First, a set of cascaded global attention modules are introduced to obtain the contextual information of the road; secondly, a set of cascaded core attention modules are embedded to ensure that the road information is transmitted to the greatest extent among the dense blocks in the network, and further assist the global attention module in acquiring multi-scale road information, thereby improving the connectivity of the road network while restoring the integrity of the road area shaded by the tree canopy and high-rise buildings. Based on binary cross entropy, an adaptive loss function is proposed for network parameter tuning. Experiments on the Massachusetts road dataset and the DeepGlobe-CVPR 2018 road dataset show that this semantic segmentation model can effectively extract the road area shaded by tree canopy and improve the connectivity of the road network.

Urban growth copes with problems in sustainable development. In developing countries, particularly, sustainable development of urban growth copes with severe challenges with respect to sluggish economic and social growth, population boom, environmental deterioration, unemployment, slums and so on. Time series of remote sensing data provide critical support on sustainability assessment. However, the urban spatial extend cannot be accurately extracted from land cover data. Targeting the urban growth and its sustainability in Islamabad, the capital of Pakistan, this study extracts urban area from four periods of Landsat images between 1990 and 2018 using an innovative object-based backdating change detection method and two criteria for extracting urban land from impervious surface. We prove that impervious surface cover and urban area increased 273.10% and 426.21%, respectively, over the last 3 decades. We identify five factors playing important role in urban growth: population, transportation systems, master planning, industrial and real estate development, and neighbor urban effect. In this study, we assess the socio-economic sustainability associated with slum growth and census data, and the environmental sustainability in relation to the variations of normalized difference vegetation index (NDVI) in forest areas. We found that slums increased with the corresponding growth of urban area and population, reflecting sluggish economic increase in Islamabad. We found that the area of woodland increased 9.29%, but its NDVI decreased from 0.668 to 0.551, implying a deteriorative trend of environmental condition.

Classification of grasslands is a convenient method to measure and manage the sustainability of Chinese grasslands. In this study, a timely and reliable procedure was examined using remote-sensing (RS) techniques. Linear regression analysis between field survey data and Moderate-Resolution Imaging Spectroradiometer (MODIS) data showed that among 17 vegetation indices (VIs) evaluated, the enhanced vegetation index (EVI) was the best VI to simulate forage dry biomass and cover in the Gannan region. The results of precision estimation of the models showed that power and logarithm regression satisfactorily simulated grassland dry biomass and grassland cover, respectively. The index of classification management of grasslands (ICGs) was used to subdivide grasslands into conservation grasslands and moderately productive grasslands in the Gannan region, where no grasslands fell into intensively productive grasslands. Conservation grasslands accounted for 2.04% of the available grasslands, whereas moderately productive grasslands were 97.96% of the available grasslands, and this is related to the history of the grasslands’ use and the per capita income in the Gannan region. This study proposes that the area of conservation grasslands and that of moderately productive grasslands are determined by increases in per capita income and changes in the human use of grasslands.