Scholarlink

<p>Urban sprawl analysis using remote sensing is essential for understanding land use and land cover (LULC) dynamics and supporting sustainable urban planning. This study proposes a hybrid quantum–classical convolutional neural network (HQCNN) for multitemporal urban sprawl analysis using remote sensing data. Multispectral linear imaging self-scanning sensor-III (LISS-III) satellite imagery of Mysuru City, India, from 2004, 2014, and 2024 was used to evaluate the proposed approach. The hybrid framework integrates quantum feature extraction with classical optimization to improve classification accuracy and computational efficiency. The model achieved overall accuracies of 95.32%, 95.62%, and 94.94%, outperforming conventional convolutional neural networks (CNNs) and existing quantum convolutional neural network (QCNN) models. Multitemporal change detection revealed urban expansion of 16.07% between 2004 and 2014 and 9.38% between 2014 and 2024, primarily replacing vegetation and non-urban land. A key limitation is the use of medium-resolution imagery, which restricts fine-scale urban feature representation. Overall, the results demonstrate the effectiveness of hybrid quantum–classical models for regional-scale urban sprawl analysis.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Hybrid quantum–classical convolutional neural network for multitemporal urban sprawl analysis using remote sensing data

  • 期刊论文

  • 发表日期：

    2026年4月18日

查看全文
  • H. N. Mahendra,
  • V. Pushpalatha,
  • S. Mallikarjunaswamy,
  • Tanvir Habib Sardar,
  • S. Rama Subramoniam

摘要

Urban sprawl analysis using remote sensing is essential for understanding land use and land cover (LULC) dynamics and supporting sustainable urban planning. This study proposes a hybrid quantum–classical convolutional neural network (HQCNN) for multitemporal urban sprawl analysis using remote sensing data. Multispectral linear imaging self-scanning sensor-III (LISS-III) satellite imagery of Mysuru City, India, from 2004, 2014, and 2024 was used to evaluate the proposed approach. The hybrid framework integrates quantum feature extraction with classical optimization to improve classification accuracy and computational efficiency. The model achieved overall accuracies of 95.32%, 95.62%, and 94.94%, outperforming conventional convolutional neural networks (CNNs) and existing quantum convolutional neural network (QCNN) models. Multitemporal change detection revealed urban expansion of 16.07% between 2004 and 2014 and 9.38% between 2014 and 2024, primarily replacing vegetation and non-urban land. A key limitation is the use of medium-resolution imagery, which restricts fine-scale urban feature representation. Overall, the results demonstrate the effectiveness of hybrid quantum–classical models for regional-scale urban sprawl analysis.