Engineering Commons^™

Eeg Data Compression, Yun-Chu Wu

Theses

This paper presents two different ways to compress EEC data-direct data com pression and a data transformation technique. The Adaptive Delta modulation and Huffman coding are used in the former method to predict or interpolate the data. Linear orthognal transformation algorithms are used in the latter method to detect and reduce the redundancies of the data by analyzing the spectral and energy distribution. Each method is implemented by programming the computer. The experimental results of their efficiencies and errors with different requirements and under different situations are compared and discussed. By comparing the EEC data compression degree and normalized square …

Multiresolution Techniques : Laplacian Pyramid Coding And Its Comparison With Subband Coding, Danmei Chang

Theses

The multiresolution pyramid and related structures have been developed as intermediate representations between the purely frequency domain and purely spatial domain. In this thesis. we discuss two important pyramid structures which are often used in multiresolution representation: Laplacian pyramid and subband coding.

Laplacian pyramid. a typical pyramid structure, is studied in detail. By means of software implementation. we obtain a comprehensive understanding of its features and applications in image representation and data compression. Several modifications of pyramid structure are used to compress the data dynamic range, reduce the computational complexity and improve performance. Performance of the conventional Laplacian pyramid and …

Comparative Study Of Prediction Gain Based On Neural Network Architecture, Prashant M. Shah

Theses

This thesis describes the Neural Network approach to design predictor using Delta and Generalized Delta Rule. The predictor is designed by supervised training based on the typical sequence of pixel values. Neural Network is used to find the coefficients of the predictor. Both 1-D and 2-D scheme of the pixels as well as linear and non-linear correlations are used to find the coefficients by training. Different combinations of pixels are used to find the "best" combination among the order of the predictor.