A semantic template matching framework for remote sensing image registration

image-20210408200648767

TensorFlow implementation of semantic template matching.

Requirements

Please use Python 3.7, install NumPy, OpenCV (3.4.2) and TensorFlow (2.0.0).

train sample

data
google earth image https://drive.google.com/drive/folders/1LV8n80daRKCySmCCQ1nZP6lB4aRsN3CM?usp=sharing
landsat-8
GF-2
SAR-optical https://drive.google.com/drive/folders/12x2m2temb5IdUjUfhWuEzCK1sXXT2ZME?usp=sharing

Example of a training data

Framework image-20210408211154158 image-20210408211354390
reference image----------------------,---------------------template image,---------------------------,label

Get started

First download the training data, place it under the project, and then generate the .tfrecord file using the code in generate data.

Training scripts

Training semantic template matching models using model files

ISPRS paper

Li L, Han L, Ding M, et al. A deep learning semantic template matching framework for remote sensing image registration[J]. ISPRS Journal of Photogrammetry and Remote Sensing, 2021, 181: 205-217.