Skip to content
Evaluation Tool for the ICDAR 2019 Competition on Table Detection and Recognition
Branch: master
Clone or download
Pull request Compare This branch is 7 commits ahead, 1 commit behind goucejia:master.
cndplab-founder resolve conflict
resolve conflict
Latest commit 738456d May 24, 2019
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
annotations
result-files omit non-essential files Apr 4, 2019
.gitignore add fix for issue content (remove test-related content parts) Apr 15, 2019
data_structure.py update measure tool May 24, 2019
eval.py update measure tool May 24, 2019
evaluate.py
readme.txt omit non-essential files Apr 4, 2019
test_eval.py some more test for table recognition Mar 28, 2019

readme.txt

Evaluation Tool for the ICDAR 2019 Competition on Table Detection and Recognition
This tool relies on Numpy and Shapely package.

Usage:
The command line takes 4 parameters as input.
Specify your track with 3rd parameter as flag: -trackA, -trackB1 or -trackB2
Specify your result file (must be a .tar.gz file) with the 4th parameter.


Example:
1. python evaluate.py -trackA ./your-file-path/you-file-name.tar.gz

2. python evaluate.py -trackB1 ./your-file-path/you-file-name.tar.gz

3. python evaluate.py -trackB2 ./your-file-path/you-file-name.tar.gz


Additional Notes:
python evaluate.py -trackA ./result-files/reg.tar.gz
run the above command for a example test.

You can’t perform that action at this time.