With the development of high-resolution optical sensors, the classification of ground objects combined with multivariate optical sensors is a hot topic at present. Deep learning methods, such as convolutional neural networks, are applied to feature extraction and classification. In this work, a novel deep belief network (DBN) hyperspectral image classification method based on multivariate optical sensors and stacked by restricted Boltzmann machines is proposed. We introduced the DBN framework to classify spatial hyperspectral sensor data on the basis of DBN. Then, the improved method (combination of spectral and spatial information) was verified. After unsupervised pretraining and supervised fine-tuning, the DBN model could successfully learn features. Additionally, we added a logistic regression layer that could classify the hyperspectral images. Moreover, the proposed training method, which fuses spectral and spatial information, was tested over the Indian Pines and Pavia University datasets. The advantages of this method over traditional methods are as follows: (1) the network has deep structure and the ability of feature extraction is stronger than traditional classifiers; (2) experimental results indicate that our method outperforms traditional classification and other deep learning approaches.

Green spaces play an important role in improving the health and quality of life of urban residents, and environmental justice has become one of the most debated topics for scholars and governments in China. The accessibility of green space has been recognized as an effective method for the assessment of environmental justice. Population and green space are the two basic elements for evaluating green space accessibility. However, high-precision population data are difficult to obtain. To address this issue, we propose a straightforward method to assess green space accessibility using urban artificial building data and investigate how green space distribution is correlated with the population. Our analysis uses urban artificial building data and green space data from 2018 in Nanjing, China. The results show that the overall amount of green space, is not only very low, but also unevenly distributed. From the urban center to the city outskirts, the green space changed representing an M-shaped pattern, while the population sharply declined. Given the serious mismatch between the population and green space, only about 26.4% and 51.8% of Nanjing residents could access green space within 400 and 800 m, respectively, which was far below that in cities in developed countries. For residents’ wellbeing, governments must consider increasing green space provision and promoting equal access.

Scientific estimations and the dynamic monitoring of the development trend of carbon emissions from energy consumption with a long time series can provide the scientific basis for formulating and implementing regional carbon-reduction strategies. Based on DMSP-OLS and NPP-VIIRS night-time light data, a pixel-scale estimation model of energy-consumption carbon emissions in Jiangsu Province from 2000 to 2019 was constructed. The spatiotemporal evolution characteristics and influencing factors were analyzed using the GIS method and a GTWR (geographically and temporally weighted regression) model. The results showed that: (1) The goodness of fit of the image-fusion correction of the two night-time light data sources from 2012 to 2013 was 0.894; the goodness of fit of the carbon-emission estimation model by stages was above 0.99; and the average relative error was 7.71%, which met the requirement for the estimation accuracy. (2) During the study period, the total carbon emissions from energy consumption in Jiangsu Province continued to increase, rising from 94.7618 million tons to 313.3576 million tons, with an annual growth rate of 6.50%; and the growth rate presented an upward trend of “slow-accelerate-decelerate”. Spatially, it showed an unbalanced distribution pattern of “low north and high south”. (3) Per-capita GDP and energy intensity were the core driving factors affecting carbon emissions in Jiangsu Province over the past 20 years. Energy intensity had the greatest driving effect on carbon emissions in southern Jiangsu, while per-capita GDP had the greatest influence in central and northern Jiangsu. Coordinating the relationship between central, north, and south Jiangsu is of great significance for the realization of the sustainable economic and social development of the double carbon goal.

Information entropy and interclass separability are adopted as the evaluation criteria of dimension reduction for hyperspectral remote sensor data. However, it is rather single-faceted to simply use either information entropy or interclass separability as evaluation criteria, and will lead to a single-target problem. In this case, the chosen optimal band combination may be unfavorable for the improvement of follow-up classification accuracy. Thus, in this work, inter-band correlation is considered as the premise, and information entropy and interclass separability are synthesized as the evaluation criterion of dimension reduction. The multi-objective particle swarm optimization algorithm is easy to implement and characterized by rapid convergence. It is adopted to search for the optimal band combination. In addition, game theory is also introduced to dimension reduction to coordinate potential conflicts when both information entropy and interclass separability are used to search for the optimal band combination. Experimental results reveal that compared with the dimensionality reduction method, which only uses information entropy or Bhattacharyya distance as the evaluation criterion, and the method combining multiple criterions into one by weighting, the proposed method achieves global optimum more easily, and then obtains a better band combination and possess higher classification accuracy.