基于深度学习语义分割模型的高分辨率遥感图像水体提取

沈骏翱, 马梦婷, 宋致远, 柳汀洲, 张微

Water information extraction from high-resolution remote sensing images using the deep-learning based semantic segmentation model

SHEN Jun’ao, MA Mengting, SONG Zhiyuan, LIU Tingzhou, ZHANG Wei

表1 S&CMNet模型语义精度

Tab.1 S&CMNet model semantic accuracy

模型名称 PA/% mIoU/% F1- Score/% TIME/ (s·100-1) NDWI 67.52 41.48 51.30 — U-Net 90.43 88.33 93.50 252.25 SegNet 93.04 87.85 93.22 419.03 S&CMNet V2 92.66 89.82 94.43 291.31 PSPNet 88.64 86.07 92.16 223.62 DeeplabV3+ 92.83 88.09 93.38 239.88 S&CMNet V1 91.37 85.09 91.51 190.66