This repository contains implementation of our GRSL paper titled as Two-Pass Bilateral Smooth Filtering for Remote Sensing Imagery. In this paper, we propose a two-pass bilateral filter, TP-based BF, and an adaptive control scheme of range kernels for noise-invariant edge-preserving image smoothing. Experimental results on four aerial-imagery benchmark datasets show that our TP-based BF outperforms the existing bilateral filters in terms of both feature-aware and gradient-aware measures.
Authors: Bo-Hao Chen, Hsiang-Yin Cheng, Yi-Syuan Tseng, and Jia-Li Yin
Paper: PDF
- MATLAB R2019a
- MATLAB R2017b
- Windows 10
- Ubuntu 16.04
Might work under others, but didn't get to test any other OSs just yet.
- To build noise dataset, you'll also need following datasets, and put the data in
./data/img_ori/
.
- Run the following script to generate noise image, and results will be saved in:
./data/img_noise/
.
$ git clone https://github.com/bigmms/chen_grsl21_tpbf.git
$ cd chen_grsl21_tpbf
$ matlab
>> demo_noise
- Run the following script to generate ground truth image, and results will be saved in:
./data/img_gt/
.
>> demo_BF
- Structure of the generated data should be:
├── data
├──img_gt #folder for storing ground truth images
│ ├── 0001.png
│ ├── 0002.png
│ └── ...
├──img_noise #folder for storing noise images
│ ├── 0001.png
│ ├── 0002.png
│ └── ...
└──img_ori #folder for storing original images
├── 0001.png
├── 0002.png
└── ...
>> demo_TPBF
The test results will be saved in: ./img_output/
The TPBF code is licensed under CC BY-NC-SA 4.0. Commercial usage is not permitted. If you use this code in a scientific publication, please cite the following paper:
@ARTICLE{ChenGRSL2021,
author={Chen, Bo-Hao and Cheng, Hsiang-Yin and Tseng, Yi-Syuan and Yin, Jia-Li},
journal={IEEE Geoscience and Remote Sensing Letters},
title={Two-Pass Bilateral Smooth Filtering for Remote Sensing Imagery},
year={2022},
volume={19},
number={},
pages={1-5},
doi={10.1109/LGRS.2020.3048488}}