Terahertz data compression using wavelets

📄 Project Abstract

Terahertz pulsed imaging (TPI) is a newly developed imaging technique with a wide range of possible applications. This project set out to investigate compression methods on the data produced by TPI, with a focus on using wavelets because they have recently emerged as powerful signal compression tools. Software was produced using Matlab to investigate different lossy compression methods using both the wavelet transform and the wavelet packet transform. It also compared quantization of the transform coefficients to a simple truncation approach in order to decide on an optimum compression level where the original signal's optical properties were not significantly altered. To complete the compression, lossless Run-length encoding was compared to a value and position entropy method, after which the final compression ratios of every method combination were compared. The first half of this paper begins by introducing Fourier analysis, the main branch of mathematics that led to the discovery o...