A Combined Convolutional Neural Network for Urban Land-Use Classification with GIS Data

Analysis of the integrated role of the Yangtze River Delta based on the industrial economic resilience of cities during COVID-19

The enhancement of regional comprehensive development ability is significantly impacted by the study on the implementation effect of regional integration strategies. The integration strategy's impact on urban development during COVID-19 in the Yangtze River Delta(YRD) is unclear. According to prior industrial transfer theory, Hefei, Anhui's capital, is difficult to transfer industries, and other YRD cities push industry integration in Anhui. This study employs the theory of economic and land resource use to examine the resilience of the industrial economy during an epidemic by using industrial land as a representation of industrial economic development. The three cities in Anhui-Wuhu, Maanshan, and Chuzhou (Wu-ma-Chu) were selected as the research area. The study employed the UNet deep learning method to detect the land use types in Wu-ma-Chu. The land transfer matrix and the standard deviation ellipse were utilised to research the alterations in industrial land use and the spatial distribution of industrial output value, respectively. The results showed that the industrial land in Machu continued to grow during the outbreak, highlighting the resilience of the region's industrial economy. During 2019-2022, the elliptical ring of industrial output value is distributed in Nanjing, revealing the radiating role of Nanjing in integrating into the integration of the YRD. This confirms China's YRD integration strategy, strengthens regional economic resilience, and encourages coordinated regional economic development.

Groundwater potential delineation using geodetector based convolutional neural network in the Gunabay watershed of Ethiopia

Groundwater potential delineation is essential for efficient water resource utilization and long-term development. The scarcity of potable and irrigation water has become a critical issue due to natural and anthropogenic activities in meeting the demands of human survival and productivity. With these constraints, groundwater resource is now being used extensively in Ethiopia. Therefore, an innovative convolutional neural network (CNN) is successfully applied in the Gunabay watershed to delineate groundwater potential based on the selected major influencing factors. Groundwater recharge, lithology, drainage density, lineament density, transmissivity, and geomorphology were selected as major influencing factors during the groundwater potential of the study area. For dataset training, 70% of samples were selected and 30% were used for serving out of the total 128 samples. The spatial distribution of groundwater potential has been classified into five groups: very low (10.72%), low (25.67%), moderate (31.62%), high (19.93%), and very high (12.06%). The area obtains high rainfall but has a very low amount of recharge due to lack of proper soil and water conservation structures. The major outcome of the study showed that moderate and low potential is dominant. Geodetoctor results revealed that the magnitude influences on groundwater potential have been ranked as transmissivity (0.48), recharge (0.26), lineament density (0.26), lithology (0.13), drainage density (0.12), and geomorphology (0.06). The model results showed that using a convolutional neural network (CNN), groundwater potentiality can be delineated with higher predictive capability and accuracy. CNN based AUC validation platform showed that, 81.58% and 86.84% were accrued from the accuracy of training and testing values, respectively. Based on the findings, the local government can receive technical assistance for groundwater exploration, and sustainable water resource development in the Gunabay watershed. Finally, the use of a detector-based deep learning algorithm can provide a new platform for industrial sectors, groundwater experts, scholars, and decision-makers.

Combining Deep Learning and Multi-Source GIS Methods to Analyze Urban and Greening Changes

Although many authors have observed a degradation in greening cover alongside an increase in the built-up areas, resulting in a deterioration of the essential environmental services for the well-being of ecosystems and society, few studies have measured how greening developed in its full spatiotemporal configuration with urban development using innovative remote sensing (RS) technologies. Focusing on this issue, the authors propose an innovative methodology for the analysis of the urban and greening changes over time by integrating deep learning (DL) technologies to classify and segment the built-up area and the vegetation cover from satellite and aerial images and geographic information system (GIS) techniques. The core of the methodology is a trained and validated U-Net model, which was tested on an urban area in the municipality of Matera (Italy), analyzing the urban and greening changes from 2000 to 2020. The results demonstrate a very good level of accuracy of the U-Net model, a remarkable increment in the built-up area density (8.28%) and a decline in the vegetation cover density (5.13%). The obtained results demonstrate how the proposed method can be used to rapidly and accurately identify useful information about urban and greening spatiotemporal development using innovative RS technologies supporting sustainable development processes.

Unified DeepLabV3+ for Semi-Dark Image Semantic Segmentation

Semantic segmentation for accurate visual perception is a critical task in computer vision. In principle, the automatic classification of dynamic visual scenes using predefined object classes remains unresolved. The challenging problems of learning deep convolution neural networks, specifically ResNet-based DeepLabV3+ (the most recent version), are threefold. The problems arise due to (1) biased centric exploitations of filter masks, (2) lower representational power of residual networks due to identity shortcuts, and (3) a loss of spatial relationship by using per-pixel primitives. To solve these problems, we present a proficient approach based on DeepLabV3+, along with an added evaluation metric, namely, Unified DeepLabV3+ and S3core, respectively. The presented unified version reduced the effect of biased exploitations via additional dilated convolution layers with customized dilation rates. We further tackled the problem of representational power by introducing non-linear group normalization shortcuts to solve the focused problem of semi-dark images. Meanwhile, to keep track of the spatial relationships in terms of the global and local contexts, geometrically bunched pixel cues were used. We accumulated all the proposed variants of DeepLabV3+ to propose Unified DeepLabV3+ for accurate visual decisions. Finally, the proposed S3core evaluation metric was based on the weighted combination of three different accuracy measures, i.e., the pixel accuracy, IoU (intersection over union), and Mean BFScore, as robust identification criteria. Extensive experimental analysis performed over a CamVid dataset confirmed the applicability of the proposed solution for autonomous vehicles and robotics for outdoor settings. The experimental analysis showed that the proposed Unified DeepLabV3+ outperformed DeepLabV3+ by a margin of 3% in terms of the class-wise pixel accuracy, along with a higher S3core, depicting the effectiveness of the proposed approach.