No abstract

Deep learning is an important research method in the remote sensing field. However, samples of remote sensing images are relatively few in real life, and those with markers are scarce. Many neural networks represented by Generative Adversarial Networks (GANs) can learn from real samples to generate pseudosamples, rather than traditional methods that often require more time and man-power to obtain samples. However, the generated pseudosamples often have poor realism and cannot be reliably used as the basis for various analyses and applications in the field of remote sensing. To address the abovementioned problems, a pseudolabeled sample generation method is proposed in this work and applied to scene classification of remote sensing images. The improved unconditional generative model that can be learned from a single natural image (Improved SinGAN) with an attention mechanism can effectively generate enough pseudolabeled samples from a single remote sensing scene image sample. Pseudosamples generated by the improved SinGAN model have stronger realism and relatively less training time, and the extracted features are easily recognized in the classification network. The improved SinGAN can better identify sub-jects from images with complex ground scenes compared with the original network. This mechanism solves the problem of geographic errors of generated pseudosamples. This study incorporated the generated pseudosamples into training data for the classification experiment. The result showed that the SinGAN model with the integration of the attention mechanism can better guarantee feature extraction of the training data. Thus, the quality of the generated samples is improved and the classification accuracy and stability of the classification network are also enhanced.

The Central Yunnan Urban Agglomeration (CYUA) is an important zone of western development in China. The clarification of the spatial structure and changing trends in CYUA could help promote the coordinated development of the CYUA and enhance the overall competitiveness of the region. Based on data from the Yunnan Statistical Yearbook and the nighttime light data, this paper extracts the urban built-up area of the CYUA and analyzes the urban expansion and urban spatial connection intensity of the CYUA from 2000 to 2018 by using the urban gravity center model and the gravity model. The results show the following: (1) From 2000 to 2018, the urban built-up area of the CYUA expanded rapidly, and the urban built-up area increased by 369.35%, with Kunming accounting for 45.41% of the increased area. Kunming was the main contributor to the increase in the urban built-up area in the CYUA. From 2000 to 2018, the urban built-up areas of the CYUA were scattered in various mountain basins. (2) Overall, the urban gravity center of the CYUA has moved to Kunming, and the distance of the urban gravity center has increased since 2005, indicating that urban expansion has accelerated since 2005. (3) The development of the CYUA is extremely unbalanced. The urban spatial connection intensity between Kunming city, Yuxi city, and Qujing city, and Yi Autonomous Prefecture of Chuxiong is relatively strong, while the urban spatial connection intensity among cities other than Kunming is weak. Overall, the CYUA is characterized by stellar radiation with Kunming city as the core and Yuxi city as the secondary core.