Discrete cosine transformation (DCT) is similar to discrete Fourier transform except DCT uses real cosine coefficients. Most of the information for DCT is contained in a handful of coefficients making it ideal for data compression, e.g. JPEG compression. DCT is a linear function with respect to spatial domain therefore, any image processing operation would yield similar results when performed in spatial or DCT domains.

This tutorial demonstrates image denoising by averaging images in real and DCT spaces, respectively.

The code from this video is available at: https://github.com/bnsreenu/python_for_microscopists

Subscribe: https://www.youtube.com/channel/UC34rW-HtPJulxr5wp2Xa04w

#python #deep-learning #machine-learning