Advanced search options

Advanced Search Options 🞨

Browse by author name (“Author name starts with…”).

Find ETDs with:

in
/  
in
/  
in
/  
in

Written in Published in Earliest date Latest date

Sorted by

Results per page:

You searched for id:"handle:1802/32167". One record found.

Search Limiters

Last 2 Years | English Only

No search limiters apply to these results.

▼ Search Limiters


University of Rochester

1. Stossel, Bryan Joseph (1963 - ). Image processing, coding, and compression with multiple-point impulse response functions.

Degree: PhD, 2017, University of Rochester

Aspects of image processing, coding. and compression with multiple-point impulse response functions are investigated. Topics considered include characterization of the corresponding random-walk transfer function. image recovery for images degraded by the multiple-point impulse response, and the application of the blur function to image coding and compression. </br>It is found that although the zeros of the real and imaginary parts of the random-walk transfer function occur in continuous, closed contours, the zeros of the transfer function occur at isolated spatial frequencies. Theoretical calculations of the average number of zeros per area are in excellent agreement with experimental results obtained from computer counts of the zeros. The average number of zeros per area is proportional to the standard deviations of the real part of the transfer function as well as the first partial derivatives. </br>Statistical parameters of the transfer function are calculated including the mean, variance, and correlation functions for the real and imaginary parts of the transfer function and their corresponding first partial derivatives. These calculations verify the assumptions required in the derivation of the expression for the average number of zeros. Interesting results are found for the correlations of the real and imaginary parts of the transfer function and their first partial derivatives. </br>The isolated nature of the zeros in the transfer function and its characteristics at high spatial frequencies result in largely reduced reconstruction artifacts and excellent reconstructions are obtained for distributions of impulses consisting of 25 to 150 impulses. </br>The multiple-point impulse response obscures original scenes beyond recognition. This property is important for secure transmission of data on many communication systems. The multiple-point impulse response enables the decoding and restoration of the original scene with very little distortion. </br>Images prefiltered by the random-walk transfer function yield greater compression ratios than are obtained for the original scene. The multiple-point impulse response decreases the bit rate approximately 40-70% and affords near distortion-free reconstructions. Due to the lossy nature of transform-based compression algorithms, noise reduction measures must be incorporated to yield acceptable reconstructions after decompression.

Record DetailsSimilar RecordsGoogle PlusoneFacebookTwitterCiteULikeMendeleyreddit

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6th Edition):

Stossel, B. J. (. -. ). (2017). Image processing, coding, and compression with multiple-point impulse response functions. (Doctoral Dissertation). University of Rochester. Retrieved from http://hdl.handle.net/1802/32167

Chicago Manual of Style (16th Edition):

Stossel, Bryan Joseph (1963 - ). “Image processing, coding, and compression with multiple-point impulse response functions.” 2017. Doctoral Dissertation, University of Rochester. Accessed August 23, 2017. http://hdl.handle.net/1802/32167.

MLA Handbook (7th Edition):

Stossel, Bryan Joseph (1963 - ). “Image processing, coding, and compression with multiple-point impulse response functions.” 2017. Web. 23 Aug 2017.

Vancouver:

Stossel BJ(-). Image processing, coding, and compression with multiple-point impulse response functions. [Internet] [Doctoral dissertation]. University of Rochester; 2017. [cited 2017 Aug 23]. Available from: http://hdl.handle.net/1802/32167.

Council of Science Editors:

Stossel BJ(-). Image processing, coding, and compression with multiple-point impulse response functions. [Doctoral Dissertation]. University of Rochester; 2017. Available from: http://hdl.handle.net/1802/32167

.