The unique high/cold environment of the Qinghai–Tibet Plateau (QTP) limits the natural distribution of the population living there and threatens local residents’ health. Thus, exploring the quality of human settlements in this area is of great significance. In this study, 5 first-level indicators and 25 second-level indicators were initially selected, and the entropy TOPSIS method was used to determine the weight of each indicator and evaluate the quality of the human settlements in each county of the QTP. Then, the coefficient of variation and spatial autocorrelation were used to analyze the spatial differences in human settlement quality. Finally, the obstacle degree model was used to identify those obstacles that affect the quality of the human settlements in the QTP. This study has gathered important findings. (1) The human settlement quality in these counties can be divided into 18 high-level areas, 45 mid- and high-level areas, 44 mid-level areas, 79 mid- and low-level areas, and 28 low-level areas. (2) In terms of spatial patterns, the north is higher than the south, the east is slightly higher than the west, and the surrounding area is higher than the interior. (3) In the clustering model, the high–high clustering trend is mainly concentrated in the north of the QTP, whereas the south-central part of the QTP and the zone where Tibet, Qinghai, and Sichuan meet exhibit obvious low–low clustering. (4) The variability of human settlement quality occurs in the order of Sichuan < Yunnan < Gansu < Xinjiang Autonomous Region < Tibet Autonomous Region < Qinghai. (5) The main first-level obstacles affecting human settlement quality in the counties of the QTP are living conditions, construction level of public service facilities, and infrastructure. The main second-level obstacles are the number of living service facilities, the number of residential districts, and the density of the road networks.
The classification of architectural style for Chinese traditional settlements (CTSs) has become a crucial task for developing and preserving settlements. Traditionally, the classification of CTSs primarily relies on manual work, which is inefficient and time consuming. Inspired by the tremendous success of deep learning (DL), some recent studies attempted to apply DL networks such as convolution neural networks (CNNs) to achieve automated classification of the architecture styles. However, these studies suffer overfitting problems of the CNNs, leading to inferior classification performance. Moreover, most of the studies apply the CNNs as a black box providing limited interpretability. To address these limitations, a new DL classification framework is proposed in this study to overcome the overfitting problem by transfer learning and learning-based data augmentation technique (i.e., AutoAugment). Furthermore, we also employ class activation map (CAM) visualization technique to help understand how the CNN classifiers work to abstract patterns from the input. Specifically, due to a lack of architectural style datasets for the CTSs, a new annotated dataset is first established with six representative classes. Second, several representative CNNs are leveraged to benchmark the new dataset. Third, to address the overfitting problem of the CNNs, a new DL framework is proposed which combines transfer learning and AutoAugment to improve the classification performance. Extensive experiments are conducted on the new dataset to demonstrate the effectiveness of our framework. The proposed framework achieves much better performance than baselines, greatly mitigating the overfitting problem. Additionally, the CAM visualization technique is harnessed to explain what and how the CNN classifiers implicitly learn for recognizing a specified architectural style.
A watershed ecosystem is a compound ecosystem composed of land and rivers, and its health is closely related to the sustainable development of the region it is located in. The Yihe River Basin (YRB) in central China’s Henan province, which is located in the north–south transition zone and has a mountain–hill–plain landscape from the upstream to the downstream, is adopted as the research area in this study. A watershed ecosystem health assessment system is constructed based on an ecosystem vigor–organization–resilience–service supply and demand harmony (EVORSH) framework and utilized to assess the ecosystem health in the YRB by taking a 3 km × 3 km grid as the evaluation unit. Thirteen factors are selected from natural and human social factors, and from them, the factors that influence watershed ecosystem health through the generation of spatial heterogeneity are identified using the geographical detector model. The following findings are obtained. (1) The mean value of ecosystem health levels in the YRB is 0.65 and at the good level. The ecosystem health has considerable spatial heterogeneity. The areas with high–high concentration are distributed in the mountains in the upper reaches of the YRB, and the areas with low–low concentration are mainly distributed in the plain areas in the middle reaches of the YRB. (2) The geographical detector result shows that 9 of 13 factors have a considerable impact on the spatial distribution of the YRB’s ecosystem health. The interaction between two factors is enhanced synergically. The decisive power of population density, rainfall, and potential evapotranspiration are more than 0.5, so these three are the main factors that influence the distribution of ecosystem health in the YRB. (3) The EVORSH framework is suitable for the measurement of ecosystem health in the YRB. The evaluation result is consistent with the actual situation in the YRB. A 3 km × 3 km grid is used as the basic research unit, and it can more accurately and scientifically express the spatial heterogeneity of ecosystem health in the YRB compared with the macro evaluation unit. This study can provide a scientific basis for ecological protection and high-quality development planning in the YRB. By integrating multi-dimensional data and methods, the EVORSH framework proposed in this study can quickly and scientifically assess the status of ecosystem health, identify the influencing factors of spatial heterogeneity, and could be applied in other similar watersheds.
The unique high-frigid environment and poor natural conditions of Qinghai–Tibet Plateau (QTP) have limited sustainable economic and social development. The construction of the beautiful QTP is a concrete implementation of the United Nations Sustainable Development Goals. However, identifying the progress and system coupling relationships of beautiful QTP construction entails some barriers due to data and methodological issues. To evaluate beautiful QTP construction and achieve a coordinated development regime, this paper employs an analytic hierarchy process and coupling model to quantify the comprehensive index and the coupling relationships of five subsystems (i.e., ecological environment, cultural inheritance, social harmony, industrial development, and institutional perfection) based on point of interest (POI) data, which are highly accurate, containing quantity and location information. Meanwhile, spatial autocorrelation analysis is conducted on the comprehensive index and coupling coordination degree for identifying the spatial clustering characteristics of the two. Results show that the progress of the beautiful QTP construction in most counties are under a very low or low level. For the system coupling perspective, 86% of counties are under the coupling stage indicating a strong interaction among the subsystems. However, coordination is out of harmony in most counties. For the spatial clustering characteristics, the comprehensive index and the system coupling relationships of beautiful QTP construction show a positive spatial correlation, indicating an aggregation effect. The aggregation is mostly “low–low” and “high–high” aggregation indicating the spatial differences and regional imbalances. The government should adopt measures to make the five subsystems of beautiful QTP construction more synergistic to achieve the sustainable development of the QTP. Our study formed a sample case of special areas where statistical data are scarce while constructing a technical framework of Beautiful China construction that is applicable to these areas. The findings of this study can serve as a reference for improving the beautiful QTP or other similar areas of construction.
This paper investigates state-of-the-art deep learning techniques to achieve automatic architectural style classification of the Chinese traditional settlements. First, a new annotated dataset is built with six typical Chinese architectural styles, consisting of over 1000 web-crawled images and an original image collection of Chinese traditional settlements. Second, a state-of-the-art convolutional network named DenseNet is benchmarked on the new dataset to learn the effectiveness of the deep learning networks. Yet, the DenseNet network suffered server overfitting on the small-sized new dataset. Third, to overcome the common overfitting problem, a new deep learning framework named DenseNet-TL-Aug is developed by leveraging transfer learning (TL) and data augmentation (DA) techniques, e.g., AutoAugment. The experimental results demonstrate that the new developed framework achieves much better classification performance in classifying the Chinese traditional style images than the original DenseNet, significantly mitigating the overfitting problem. This study will contribute to automated landscape gene recognition as well as the design and development of traditional tourism.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.