No abstract

To address the problem of subjectivity in determining the poverty-returning risk among registered poor households, a method of monitoring and analyzing the poverty-returning risk among households based on BP neural network and natural breaks method was constructed. In the case of Yunyang District, Hubei Province, based on the data of the poverty alleviation and development system, we constructed a monitoring system for the poverty-returning risk for the registered poor households. The spatial distribution pattern of households under the poverty-returning risk was analyzed from two scales of district and township, respectively, by combining Geographic Information Science, and the influence degree of indicators on the poverty-returning risk using mean impact value (MIV). The results show that: (1) The spatial distribution of the poverty-returning risk among the registered poor households in the study area basically coincides with the local natural poverty-causing factors and the degree of social and economic development. (2) The Poverty-Returning Risk Index for each township represents a globally strong spatial dependence with a Moran’s I coefficient of 0.352. (3) The past poverty identification status of registered poor households is the main factor to reduce the poverty-returning risk, and the past policy should remain unchanged for a period of time. (4) Improving the quality of education within households and focusing on helping households with older average age can further reduce the poverty-returning risk.

No abstract

No abstract

Efficient processing of ultra-high-resolution images is increasingly sought after with the continuous advancement of photography and sensor technology. However, the semantic segmentation of remote sensing images lacks a satisfactory solution to optimize GPU memory utilization and the feature extraction speed. To tackle this challenge, Chen et al. introduced GLNet, a network designed to strike a better balance between GPU memory usage and segmentation accuracy when processing high-resolution images. Building upon GLNet and PFNet, our proposed method, Fast-GLNet, further enhances the feature fusion and segmentation processes. It incorporates the double feature pyramid aggregation (DFPA) module and IFS module for local and global branches, respectively, resulting in superior feature maps and optimized segmentation speed. Extensive experimentation demonstrates that Fast-GLNet achieves faster semantic segmentation while maintaining segmentation quality. Additionally, it effectively optimizes GPU memory utilization. For example, compared to GLNet, Fast-GLNet’s mIoU on the Deepglobe dataset increased from 71.6% to 72.1%, and GPU memory usage decreased from 1865 MB to 1639 MB. Notably, Fast-GLNet surpasses existing general-purpose methods, offering a superior trade-off between speed and accuracy in semantic segmentation.