ICIP6-Underwater-Image-Synthesis

This project aims to generate pairs of clean RGB-D and underwater images for different water types which can be used to train models for underwater image restoration. It contains the datafiles of attenuation coefficients, spectral response of camera, illumination on the water surface and clean RGB-D images which are the parameters for the mathematical model. Two images are already provided as an example. Different RGB-D images can also be used to produce outputs. The main function is implemented in the Python (.py) file. Please check the jupyter notebook (.ipynb) to generate the required outputs.