tel: 208 426 2214
fax: 208 426 2470
E-mail:
EBarneySmith@boisestate.edu
Area of Expertise
Document imaging, optical scanners
Biography
Elisa Barney-Smith is an associate professor and associate chair of the Electrical & Computer Engineering department at Boise State University. She received a B.S. in Computer Science and the M.S. and Ph.D. degrees in Electrical, Computer and Systems Engineering all from Rensselaer Polytechnic Institute, Troy, NY. Elisa's main research focus is in document imaging. Current projects are centered around developing models of the degradations produced by optical page scanners, printers, FAX machines and photocopiers. These models will be used to improve the recognition and processing of the images.
Lecture Title(s)
Effects of Scanning Degradations on Document Images
Optics in camera systems have been characterized by many methods for many purposes. Most often the effect of the optics is considered for grey scale or color images. Bilevel images such as found in text documents and line drawings are also affected by the optics, but because of the thresholding the way the optics and optical system in desk top scanners affect (and degrade) images produces different artifacts. These artifacts affect both human readability and machine readability. What is legible and comfortable for humans to read is not the same as what is legible for a computer to read and convert documents into recognized documents. The types of effects that are seen, a discussion about human versus machine legibility and measurement techniques appropriate for the bilevel case will be discussed.
Document Image Degradations and Analysis of Statistical Image Differences
Document image quality is degraded through processes such as scanning, printing, and photocopying. The resulting bilevel image degradations can be categorized based either on observable degradation features or on degradation model parameters. The image degradation features can be related mathematically to model parameters. In this talk the degradation model will be introduced. Then pairs of populations of degraded character images created with different model parameters will be statistically compared. The probability that the character populations were degraded by the same model parameters correlates with the relationship between observable degradation features and the model parameters. Two metrics of character difference are used: Hamming distance and moment feature distance. Knowledge about the conditions under which characters will be similar and when they will be different can influence the choice of parameters for future experiments.
