Optimizing rural waste management: Leveraging high-resolution remote sensing and GIS for efficient collection and routing,International Journal of Applied Earth Observation and Geoinformation

当前位置： X-MOL 学术 › Int. J. Appl. Earth Obs. Geoinf. › 论文详情

Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)

Optimizing rural waste management: Leveraging high-resolution remote sensing and GIS for efficient collection and routing
International Journal of Applied Earth Observation and Geoinformation ( IF 7.6 ) Pub Date : 2024-11-06 , DOI: 10.1016/j.jag.2024.104219
Xi Cheng, Jieyu Yang, Zhiyong Han, Guozhong Shi, Deng Pan, Likang Meng, Zhuojun Zeng, Zhanfeng Shen

Accurate assessment of distribution patterns and dynamic insights into rural populations is pivotal for comprehending domestic waste generation, recycling, and transportation in rural territories. Given that the dispersion of rural inhabitants exhibits minimal variation and maintains stability, this research endeavors to establish a pragmatic model for rural domestic waste collection and routing, leveraging the capabilities of very high-resolution remote sensing combined with geographic information system (GIS) techniques. Specifically, the Dilated LinkNet model was employed to discern features such as buildings, roads, water bodies, farmlands, and forests from the high-resolution remote sensing imagery. A novel multiple K-means clustering approach was devised for building segmentation. Within these clusters, an assortment of spatial regulations and evaluations facilitated the judicious selection of environmentally-conscious waste collection sites (WCSs). The Pointer Network, augmented with reinforcement learning, executed a traveling salesman analysis on these chosen WCSs, yielding the optimal collection trajectory. Validated in Huangtu Town, a quintessential rural region in China, our model manifested superior recognition precision, recording IoU accuracies of 0.902, 0.926, 0.933, 0.891, and 0.849 for buildings, roads, water bodies, farmlands, and forests respectively. Notably, when compared to our field survey data, the optimized daily collection route in a rural context decreased from 256.40 km before optimization to 140.44 km, reflecting a substantial reduction of 45.23% in total distance. This study furnishes an effective model that relies solely on information from remote-sensing images for efficient rural waste collection and extends invaluable insights to planners and administrators in the realm of rural and township waste management.

更新日期：2024-11-06

点击分享查看原文

点击收藏

阅读更多本刊新发论文本刊介绍/投稿指南