This preview shows page 1. Sign up to view the full content.
Unformatted text preview: ake advantage of the fact that when we read a
handwritten document, we have a good idea of its context (what the topic is). Hence we
can often read a handwritten document even if we can decipher only a few letters in each
word. But teaching a computer to perform the same task is really difficult.
Optical Mark Reader (OMR)
These scanners are capable of recognizing a pre-specified type of mark made by pencil or
pen. For example, many students might have appeared in objective type tests where they
had to mark their answers to questions on a special pre-printed test scoring sheet by
darkening a small square, circular or oval shaped box by a pencil to indicate their correct
choice out of the various given options (see Figure 9.12). These answer sheets are
directly fed to a computer for grading with the use of an optical mark reader.
The actual technique used by an OMR device for recognition of marks involves focusing
a light on the page being scanned and detecting the reflected light pattern from the marks.
Pencil marks made with a soft lead pencil (high graphite content) reflect the light
enabling the reader to determine which responses are marked.
Note that OMR devices require that the pre-printed forms use good quality paper and the
boxes to be marked on them are accurately aligned for correct interpretation of input data.
Furthermore, the pre-printed forms should not be folded so that the marks can be read
The use of OMR is not limited only to the grading of objective type tests. In fact, any
input data that is of a choice or selection nature can be recorded for OMR input. Thus
OMR is very useful for grading tests with objective type questions in which a large
number of students appear and for researchers who need to tabulate responses to large
surveys where responses can be restricted to one or more out of a few possibilities.
Data coded in the form of small lines (known as bars) are known as bar codes. Bar codes
represent alphanumeric data by a combination of adjacent vertical lines (bars) by varying their width and the spacing between them. They are particula...
View Full Document
This document was uploaded on 04/07/2014.
- Spring '14