MCDD: A Novel Macao Change Detection Dataset for Sea Reclamation

Sea reclamation is essential for expanding land resources and supporting sustainable urban development in coastal areas. With the advancement of Remote Sensing (RS) technology and Convolutional Neural Network (CNN), sea reclamation land cover Change Detection (CD) has become feasible through high-resolution imagery analysis. However, there is a lack of publicly available large-scale sea reclamation datasets. Therefore, Macao Change Detection Dataset (MCDD) is developed, consisting of <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$10,000\ 256 \times 256$</tex> resolution image pairs containing 15 years of land cover changes in Macao. This dataset provides valuable information on sea reclamation and lays the foundation for training and evaluating deep learning models for RS tasks. The performance of multiple state-of-the-art models on MCDD is evaluated in terms of IoU, Precision, Recall, F1, Kappa, Overall Accuracy (OA), parameters and Floating-point Operations (FLOPs). Experimental results demonstrate that MCDD is robust and effective for sea reclamation land cover CD.