FanInSAR is a Fancy Interferometric Synthetic Aperture Radar (InSAR) time series analysis library written in Python. It aims to provide a foundational library for the development of InSAR algorithms, facilitating efficient processing of InSAR time series data by offering a Pythonic, fast, and flexible approach.
Most existing community InSAR software adopts a workflow-oriented approach. While this lowers the entry barrier for new users, it often compromises flexibility and extensibility. For algorithm researchers, integrating new methods into these rigid workflows can be challenging, highlighting the need for a more adaptable framework for InSAR time series analysis.
FanInSAR is designed to bridge this gap. It serves as a foundational library for InSAR time series processing, offering a flexible and extensible framework tailored for algorithm researchers and developers. FanInSAR is not a complete end-to-end InSAR processing system; rather, it provides building blocks for creating custom workflows. Its high-level API abstracts the complexity of the processing pipeline and hides low-level implementation details, allowing users to focus on developing and testing new algorithms. For researchers aiming to rapidly prototype and deploy their own InSAR methods, FanInSAR offers a fast and efficient starting point.
- Pythonic: FanInSAR is written in Python and provides a user-friendly API. For example, a series of well-known InSAR datasets are provided in the form of Python classes; loading data from
HyP3orLiCSARproducts is as simple as providing the corresponding home directory. Sampling values from an interferometric dataset is as easy as calling thequery()method by passing the spatial (Points, BoundingBox, Polygons) and temporal (Pairs) queries. The warping process during sampling (such as reprojecting and resampling) is automatically handled by the library. - Fast: The core computation in FanInSAR is implemented using
PyTorch, a high-performance deep learning library. This allows for efficient processing on both CPU and GPU, enabling faster execution. - Flexible: FanInSAR is designed to be flexible, allowing for customization and extension. Users can easily inherit classes or customize the processing pipeline for their specific needs.
FanInSAR is a Python package, and requires Python >= 3.8. You can install the latest release of FanInSAR using pip from the PyPI:
pip install FanInSARor from GitHub:
pip install git+https://github.com/Fanchengyan/FanInSAR.gitThe detailed documentation is available at: https://faninsar.readthedocs.io/en/latest/
⚠️ NoteFanInSAR is under active development and is currently in the alpha stage. Its API may change in the future until it reaches a stable version.
Fan, C., & Liu, L. (2024). FanInSAR: A Fancy InSAR time series library, in a Pythonic, fast, and flexible way (0.0.1). Zenodo. https://doi.org/10.5281/zenodo.11398347
@software{fan_2024_11398347,
author = {Fan, Chengyan and
Liu, Lin},
title = {{FanInSAR: A Fancy InSAR time series library, in a
Pythonic, fast, and flexible way}},
month = may,
year = 2024,
publisher = {Zenodo},
version = {0.0.1},
doi = {10.5281/zenodo.11398347},
url = {https://doi.org/10.5281/zenodo.11398347}
}