ct_assist, or CameraTransform-Assistance, is a library for finding head-feet pairs on images. This allows for further automation of the CameraTransform. It extracts head-feet pairs from a binary mask generated using instance segmentation. This requires the user to input a pytorch tensor.

This project was made for EU-Project ASSISTANCE. See below for more.

This library will be applied to images of fluid spills in order to estimate their area and release rate. See fluid_estimator, request access from the the author.

This repo extends CameraTransform's functionality by trying to find the reference objects with Computer Vision.

We provide a Anaconda environment file [env.yaml] for installing this repo's dependencies.

1. git clone https://github.com/Casper-Smet/ct_assist.git
2. cd .\ct_assist\
3. conda env create -f env.yml
4. conda activate ct_assist
5. pip install .

Table: RMSE per camera property, ten runs

Plot with results:

Adapted Situation AwareneSS tools and taIlored training scenarios for increaSing capabiliTies and enhANcing the proteCtion of First RespondErs – ASSISTANCE is an international research project funded by the European Commission under the Horizon 2020 programme in Secure Societies Challenge addressing the SU-DRS02-2018-2019-2020 (Technologies for first responders) topic.

The main purpose of ASSISTANCE project is twofold: to help and protect different kind of first responders’ (FR) organizations that work together during the mitigation of large disasters (natural or man-made) and to enhance their capabilities and skills for facing complex situations related to different types of incidents.

For more, see: https://assistance-project.eu/