cnr-isti-vclab / TagLab

A CNN based image segmentation tool oriented to marine data analysis
https://taglab.isti.cnr.it/
GNU General Public License v3.0
84 stars 33 forks source link
annotation-tool cnn coral-reefs corals image-segmentation labeling-tool labels marine-data-analysis python semantic semantic-segmentation underwater-monitoring

TagLab: an image segmentation tool oriented to marine data analysis

|   Software Requirements   |   Install   |   Update   |   Citation   |

TagLab was developed to assist with the annotation and extraction of statistical data from ortho-maps of benthic communities. The tool features various AI segmentation networks specifically trained for agnostic recognition (focused solely on contours) and semantic recognition (which also considers species) of corals. TagLab is an ongoing project of the Visual Computing Lab.

ScreenShot

Interaction

TagLab allows users to interactively segment regions using the following methods:

Fully Automated Segmentation

TagLab enables users to automatically segment regions by using a custom segmentation model, which are trained using the DeepLab V3+ backbone. Using a custom model in TagLab is accomplished through the following steps:

  1. Creating and managing a training dataset
  2. Training a model using the Train Your Network interface
  3. Making predictions with the Fully Automated Semantic Segmentation tool

TagLab has many resources for image data management and analysis, including the ability to track changes among time points. For additional information, please visit the official TagLab website and documentation page.

Software Requirements

TagLab runs on Linux, Windows, and MacOS. To run TagLab, the main requirement is just 64bit Python 3.11.x.

GPU accelerated computations are not supported on MacOS and on any machine that does not have an NVIDIA graphics card with CUDA enabled.

Installing TagLab

See the instructions on the wiki.

Updating TagLab

If you already installed TagLab and you need to update to a new version, you can just run the update.py script from the terminal (be sure to be into the TagLab main folder, see step 2):

python3 update.py

or, on Windows:

python.exe update.py

The script will automatically update TagLab to the newest version available in this repository.

NOTE: If some package is missing, after an update, re-launch install.py.

Updating from 0.2

If you are updating TagLab from 0.2 version, in order to download also the new networks, please run the update.py script twice:

python3 update.py
python3 update.py

Citation

If you use TagLab, please cite it.

@article{TagLab,
    author = {Pavoni, Gaia and Corsini, Massimiliano and Ponchio, Federico and Muntoni, Alessandro and Edwards, Clinton and Pedersen, Nicole and Sandin, Stuart and Cignoni, Paolo},
    title = {TagLab: AI-assisted annotation for the fast and accurate semantic segmentation of coral reef orthoimages},
    year = {2022},
    journal = {Journal of Field Robotics},
    volume = {39},
    number = {3},
    pages = {246 – 262},
    doi = {10.1002/rob.22049}
}