Abstract
dc:description.abstract<p>Machine vision principles are increasingly applied in the areas of manufacturing and automated inspection. The objective of this thesis work is to investigate different methods that could be implemented for matching printed dot matrix characters with the actual characters to be printed.</p><p>In support to this investigation, a survey of various methods proposed by researchers in the field of Optical Character Recognition (OCR) is presented. Survey includes a detailed description of the state-of-the art methods. It is shown that due to restrictions of hardware cost and the time available to inspect each character, the standard approaches are not practical. Hence, three heuristic approaches have been investigated.</p><p>Two of the three methods require the use of a linear array video scanner. The video data obtained is analyzed for printing errors. First method utilizes Fourier Transformation of linear array output in order to identify errors in the frequency domain. In the second method, the data output of the entire string printed is passed through the AND or OR logical operator. The third method requires an area video scanner. This algorithm attempts to recognize the characters printed.</p><p>Another method that exploits the basic nature of the printer mechanics was also investigated for this project. However, due to reason of propriety ownership it has not been discussed in this report. For further information about this approach, the reader is referred to a detailed report by Dr. Dave.</p>
Degree
thesis:*- Name thesis:degree_name
- Master of Science in Mechanical Engineering - (M.S.)
- Discipline thesis:degree_discipline
- Mechanical and Industrial Engineering
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Premnath, Thatta V.
- Contributors dc:contributor
-
- Rajesh N. Dave
- Avraham Harnoy
- Harry Herman
Subjects
dc:subject × 4Identifiers
dc:identifier.*- Repository record dc:identifier
- https://digitalcommons.njit.edu/theses/2871