fusion-accuracy-assessment

It includes the code and test data for assessing the accuracy of fused images

https://github.com/xzhu-lab/fusion-accuracy-assessment

Repository

It includes the code and test data for assessing the accuracy of fused images

Readme.txt

1. In the zipped folder "IDL code and sample data for computing accuracy metrics", there are two files:
   (1) an IDL code "optimal_accuracy_metrics.pro" which can compute the four metrics, AD, RMSE, edge, LBP for any fused images. Input the data according to the pop-up window when running the code. The result will be saved as a csv file in the folder of fused image and named as fused image name+_accuracy.csv (e.g.,fused_fine_image_FSDAF.tif_accuracy.csv). The first n rows are accuracies of n bands, and the last row is the average accuracies of multiple bands.
   Please note: the code used some functions from the classic ENVI, so classic ENVI should be open when compiling the code! 
   (2) sample data: a 3-band reference image (green, red, nir) on 25 Nov 2001 in the CIA site; a fused image by FSDAF

2. In the zipped folder "python code and sample data for computing accuracy metrics", there are three files:
   (1) an instruction file "Instruction of optimal_accuracy_metrics": detailed introduction on how to implement the code, suggesting read this file before running the code.
   (2) a python code "optimal_accuracy_metrics.py" which can compute the four metrics, AD, RMSE, edge, LBP for any fused images. Input the data according to the pop-up window when running the code. The result will be saved in the folder of input images. The first n rows are accuracies of n bands, and the last row is the average accuracies of multiple bands.
   Please note: the result from python code is slightly different with that of IDL code because these two codes process floating-point numbers in different ways.
   (3) sample data: a 3-band reference image (green, red, nir) on 25 Nov 2001 in the CIA site; a fused image by FSDAF

3. In the zipped folder "R code and sample data for drawing Taylor like diagram", there are two files:
   (1) a R code "APA_diagram.R" which can draw the all-round performance assessment(APA) diagram, a Taylor-like polar diagram to show the accuracies of different fused images with "fair" and "good" ranges.
   Please set the parameters and input the data file at the beginning of the code based on your own data. The produced diagram will be saved to your working direction with the filename you define (e.g., APA diagram example.png).  
   (2) data_for_APA_diagram.csv: a sample data for testing the code, which includes accuracies of 6 fusion methods for fusing the image on 25 Nov 2001 in the CIA site.


To Cite these codes and dataset in Publications:
Zhu, X., Zhan, W., Zhou, J., Chen, X., Liang, Z., Xu. S., Chen, J.2022. A novel framework to assess all-round performances of spatiotemporal fusion models, Remote Sensing of Environment，274，113002，https://doi.org/10.1016/j.rse.2022.113002

Owner

Citation (CITATION.cff)

# This CITATION.cff file was generated with cffinit.
# Visit https://bit.ly/cffinit to generate yours today!

cff-version: 1.2.0
title: Fusion-accuracy-assessment
message: Please cite this software using these metadata.
type: software
authors:
  - given-names: Xiaolin
    family-names: Zhu
  - given-names: Wenfeng
    family-names: Zhan
  - given-names: Junxiong
    family-names: Zhou
  - given-names: Xuehong
    family-names: Chen
  - given-names: Zicong
    family-names: Liang
  - given-names: Shuai
    family-names: Xu
  - given-names: Jin
    family-names: Chen
identifiers:
  - type: doi
    value: 10.5281/zenodo.6387281
    description: >-
      When requested, you can cite this code through
      the doi published by Zenodo.
keywords:
  - Spatiotemporal fusion
  - Accuracy assessment
  - Inter-comparison
  - Taylor diagram
  - Model performance
license: CC-BY-4.0

GitHub Events