Search

Timely and accurate survey of construction and demolition waste (C&DW) distribution is of significance for managing C&DW and enhancing the quality of the urban environment, especially in rapidly urbanizing country of China. Automatic C&DW recognition using high-resolution remote sensing images is an extremely important surveying method. Existing C&DW recognition methods can only obtain a rough spatial distribution of C&DW. Mapping precise C&DW distribution on large-scale remote sensing images is still challenging. In this paper, to solve the problem of precise C&DW recognition, a novel deep learning algorithm based on high-resolution remote sensing images is proposed, which we refer to as the cross-channel multi-scale gated fusion network (CCMGNet). CCMGNet extracts depth features of C&DW from RGB and NIR bands by using two independent encoding streams. A gated fusion layer and a multi-scale attention module are designed to effectively fuse the features from two encoding streams. We manually labelled the sample set used for training and testing on 26 remote sensing images captured by the Gaofen-2 (GF-2) satellite. These images were taken in nine typical cities in China over the time period from 2019 to 2021. The results showed that the proposed method is effective in recognizing C&DW, realizing 89.42% and 84.52% in precision and IoU, and is superior to other state-of-the-art deep learning algorithms and existing C&DW recognition methods. The effectiveness of all significant components in CCMGNet was confirmed by ablation experiments. The proposed method was applied to extract C&DW regions in large-scale remote sensing images, with an example of a GF-2 image of Beijing. The high extraction efficiency and satisfactory visual effect demonstrate the potential of the proposed method for surveying the spatial distribution of C&DW. The novel method will play a crucial role in automatically surveying the spatial distribution and size of C&DW across the country. [ABSTRACT FROM AUTHOR]

The reflectance anisotropy of land surface serves as an important bridge between surface biophysical parameters and remote sensing observations. It can characterize by the linear kernel-driven bidirectional reflectance distribution function (BRDF), which is the combination of several kernel functions and kernel weights. These kernel weights can be estimated by remote sensing; however, the stability of current kernel weights products is still challenging, especially in urban areas with complex aerosol properties and heterogeneous surfaces. In this paper, we propose a method for robust estimation of kernel weights from the Moderate Resolution Imaging Spectroradiometer (MODIS) surface spectral reflectance products (MxD09GA) data based on the constrained least-squares method (CLSM) and a priori knowledge. The kernel weights data were obtained by the CLSM from 2014 to 2017 in Beijing region of China. Validations were carried out using the MxD09GA and BRDF/Albedo products (MCD43A1). The results show that the time series of kernel weights by the CLSM show small variability over different land cover types. The kernel weights estimated by the CLSM can clearly show the phenological signal and fitting ability of surface spectral reflectance is better than that of the MCD43A1 products in Beijing urban area. Experimental results demonstrate that the CLSM has the potential for the robust estimation of kernel weights in urban areas. [ABSTRACT FROM AUTHOR]

To investigate geological mining hazards using digital techniques such as high-resolution remote sensing, a semi-automatically geological mining hazards extraction method is proposed based on the case of the Shijiaying coal mine, located in Fangshan District, Beijing, China. In the method, the vegetation is first removed using the normalized difference vegetation index (NDVI) on the GeoEye-1 data. Then, geological mining hazards interpretation features are determined after color enhancement using principal component analysis (PCA) transformation. Bitmaps mainly covered by geological mining hazards are isolated by masking operation in the environment for visualizing images software. Next, each bitmap is classified into a two-valued imagery using support vector machine algorithm. In the two-valued imagery, 1 denotes the geological mining hazards, while 0 denotes none. Afterwards, the two-valued imagery is converted into a vector graph by corresponding functions in the ArcGIS software and no geological mining hazards regions in the vector graph are deleted manually. Finally, the correlation between factors (such as mining activity, lithology, geological structure, and slope) and geological mining hazards is analyzed using a logistic regression and a hazardous-area forecasting model is built. The results of field verification show that the accuracy of the geological mining hazards extraction method is 98.1% and the results of the hazardous-area forecasting indicate that the logistic regression is an effective model in assessing geological hazard risks and that mining activity is the main contributing factor to the hazards, while geological structure, slope, lithology, roughness of the surface, and aspect are the secondary. [ABSTRACT FROM AUTHOR]