A university image analysis project to correct the following panda image using OpenCV to perform spatial and frequency domain corrections.
The DFT of the panda is taken using OpenCV functionality.
The DFT of the panda is shifted to visualise low frequency data in the center and high frequency data in the corners.
By applying various masks to the DFT, shifting back and taking the inverse DFT, the image quality has (arguably!) been improved.
Though tedious to make, the final specific filter saw great results and despite a large amount of static still visible on the panda, the image has been improved greatly by removing the common frequencies across the image