Development of Adaptive Image Estimate Based on Minimum Description Length Criteria for Simultaneous Noise Reduction and Compression of Image

Document Type



Master of Science (MS)


Computer Engineering

First Advisor's Name

Jean Andrian

First Advisor's Committee Title

Committee Chair

Second Advisor's Name

Kang Yen

Third Advisor's Name

Armahdo Barreto

Date of Defense



Traditional image processing approaches have separated the problem of noise reduction and data compression, in a sense that the image is processed at various stages, each proposed to address specific data redundancies. Compression techniques based on Wavelets have addressed the combination of noise reduction and compression more effectively due to the application of thresholding. With that, however, the disadvantage of manual thresholding and selection of best estimating Wavelet basis is an associated limitation. This thesis will present a development of two-dimensional Adaptive Minimum Description Length (AMDL) algorithm and further expand it to an Adaptive Image Estimate by incorporating the Mean Square Error for selection of the ‘best’ threshold in reference to the ‘best’ available Wavelet basis resulting a simultaneous noise reduction and compression of images. This approach may have applications in analyzing any noisy image such as images captured by satellites on a cloudy day, medical or industrial microscopic images containing interference, and the list goes on ....



This document is currently not available here.



Rights Statement

Rights Statement

In Copyright. URI:
This Item is protected by copyright and/or related rights. You are free to use this Item in any way that is permitted by the copyright and related rights legislation that applies to your use. For other uses you need to obtain permission from the rights-holder(s).