A novel CNN-attention-LSTM method for airport remote sensing image caption
Boqi Shan, Wei Yang, X. Wei, Hongcheng Zeng, Yue Wang, J Y Chen · IET conference proceedings. · 2026
At present, remote sensing images have become an important source of geospatial information acquisition, but the spatial differences and size changes of targets in remote sensing images make it difficult for the naked eye to efficiently analyse image information. In this study, we use the image caption algorithm based on CNN-attention-LSTM to apply and evaluate the airport optical remote sensing images. We analyse the influence of different dataset combinations on the image caption effect, discuss the specific application effect of the Attention mechanism in image caption, and further improve the dataset, so that the model can generate accurate captions of the image scene including fine-grained elements such as aircraft colour, quantity, and status for airport optical remote sensing images, improve the automation level of remote sensing image interpretation, and provide more comprehensive and accurate information support for related fields.