Оптическое распознавание заполненных форм

Capturing of human marked data from document forms

Optical mark recognition (OMR) collects data from people by identifying markings on a paper. OMR enables the hourly processing of hundreds or even thousands of documents. For instance, students may remember completing quizzes or surveys that required them to use a pencil to fill in bubbles on paper (seen to the right). A teacher or teacher's aide would fill out the form, then feed the cards into a system that grades or collects data from them.

An optical answer sheet or bubble sheet is a special type of form used in multiple choice question examinations. OMR is used to detect answers. The Scantron Corporation creates many optical answer sheets, although certain uses require their own customized system. Optical answer sheets usually have a set of blank ovals or boxes that correspond to each question, often on separate sheets of paper. Bar codes may mark the sheet for automatic processing, and each series of ovals filled will return a certain value when read. In this way students' answers can be digitally recorded, or identity given.

The first optical answer sheets were read by shining a light through the sheet and measuring how much of the light was blocked using phototubes on the opposite side. As some phototubes are mostly sensitive to the blue end of the visible spectrum, blue pens could not be used, as blue inks reflect and transmit blue light. Because of this, number two pencils had to be used to fill in the bubbles—graphite is a very opaque substance which absorbs or reflects most of the light which hits it. In contrast, OMR software users design their own mark sense forms with a word processor or built in form editor, print them locally on a printer, and can save thousands of dollars on large numbers of forms. Identifying optical marks within a form, such as for processing census forms, has been offered by many forms processing (Batch Transaction Capture) companies since the late 1980s. Mostly this is based on a bitonal image and pixel count with minimum and maximum pixel counts to eliminate extraneous marks, such as those erased with a dirty eraser that when converted into a black and white image (bitonal) can look like a legitimate mark. So this method can cause problems when a user changes their mind, and so some products started to use grayscale to better identify the intent of the marker—internally scantron and NCS scanners used grayscale. OMR software is also used for adding OMR marks to mail documents so they can be scanned by folder inserter equipment. An example of OMR software is Mail Markup from UK developer Funasset Limited. This software allows the user to configure and select an OMR sequence then apply the OMR marks to mail documents prior to printing.

Optical mark recognition (OMR) is the scanning of paper to detect the presence or absence of a mark in a predetermined position. With modern OMR, where the presence of a pencil filled in bubble is recognized, the recognition is done via an optical scanner. The first mark sense scanner was the IBM 805 Test Scoring Machine; this read marks by sensing the electrical conductivity of graphite pencil lead using pairs of wire brushes that scanned the page. In the 1930s, Richard Warren at IBM experimented with optical mark sense systems for test scoring, as documented in US Patents 2,150,256 (filed in 1932, granted in 1939) and 2,010,653 (filed in 1933, granted in 1935). The first successful optical mark sense scanner was developed by Everett Franklin Lindquist as documented in US Patent 3,050,248 (filed in 1955, granted in 1962). Lindquist had developed numerous standardized educational tests, and needed a better test scoring machine than the then standard IBM 805. The rights to Lindquist's patents were held by the Measurement Research Center until 1968, when the University of Iowa sold the operation to Westinghouse Corporation. During the same period, IBM also developed a successful optical mark sense test scoring machine, as documented in US Patent 2,944,734 (filed in 1957, granted in 1960). IBM commercialized this as the IBM 1230 Optical mark scoring reader in 1962. This and a variety of related machines allowed IBM to migrate a wide variety of applications developed for its mark sense machines to the new optical technology. These applications included a variety of inventory management and trouble reporting forms, most of which had the dimensions of a standard punched card. While the other players in the educational testing arena focused on selling scanning services, Scantron Corporation, founded in 1972, had a different model; it would distribute inexpensive scanners to schools and make profits from selling the test forms. As a result, many people came to think of all mark sense forms (whether optically sensed or not) as scantron forms. In 1983, Westinghouse Learning Corporation was acquired by National Computer Systems (NCS). In 2000, NCS was acquired by Pearson Education, where the OMR technology formed the core of Pearson's Data Management group. In February 2008, M&F Worldwide purchased the Data Management group from Pearson; the group is now part of the Scantron brand. OMR has been used in many situations as mentioned below. The use of OMR in inventory systems was a transition between punch cards and bar codes and is not used as much for this purpose.

In the past and presently, some OMR systems require special paper, special ink and a special input reader (Bergeron, 1998). This restricts the types of questions that can be asked and does not allow for much variability when the form is being input. Progress in OMR now allows users to create and print their own forms and use a scanner (preferably with a document feeder) to read the information. The user is able to arrange questions in a format that suits their needs while still being able to easily input the data. OMR systems approach one hundred percent accuracy and only take 5 milliseconds on average to recognize marks. There is also the possibility of missing data in the scanning process, and incorrectly or unnumbered pages can lead to their being scanned in the wrong order. Also, unless safeguards are in place, a page could be rescanned, providing duplicate data and skewing the data. As a result of the widespread adoption and ease of use of OMR, standardized examinations can consist primarily of multiple choice questions, changing the nature of what is being tested.