Speed compensation scheme for reading MICR data

• US 4,776,021 A
• Filed: 11/16/1987
• Issued: 10/04/1988
• Est. Priority Date: 11/16/1987
• Status: Expired due to Term

• Current Assignee: NCR Corporation (NCR Voyix Corporation)
• Original Assignee: NCR Corporation (NCR Voyix Corporation)
• Inventors: Ho, Benedict C. M.
• Primary Examiner(s): Boudreau, Leo H.
• Assistant Examiner(s): Daley, Donald J.
• Application Number:
  US07/121,372
• Time in Patent Office: 323 Days
• Field of Search: 382/7, 382/11, 382/61, 235/436, 235/474
• US Class Current: 382/139
• CPC Class Codes: G06V 30/2253 Recognition of characters printed with magnetic ink G06V30/2247 takes precedence

Abstract

A method and apparatus for providing a speed compensation scheme for reading MICR data on a document. The magnetic waveform for a character is sampled at periodic rates and converted to first bytes of data which are placed in a first memory. The velocity of the document is sampled at second periodic rates and the instantaneous values of the speed of the document are placed in a second memory as second bytes of data. A processor is used to remove the second bytes of data from memory and to calculate an average speed of the document for an associated examining window consisting of a predetermined number of the first bytes of data. The width of the examining window is adjusted in accordance with the average speed compared to an anticipated normal speed.

Claims

1. An apparatus for reading magnetic characters on a document, comprising:

reading means for reading magnetic characters on a document at a reading station;

moving means for moving said document in reading relationship with said reading means to produce magnetic waveforms corresponding to the characters being read;

an analog to digital converter for converting the magnetic waveform associated with a character into binary data;

sampling means for sampling said binary data at periodic times to produce first bytes of data representing said magnetic waveform;

a first memory means for storing said first bytes of data;

velocity sensing means for sensing the velocity of said document at said reading means at periodic times to produce second bytes of data representing instantaneous velocities of said document at said reading means;

a second memory means for storing said second bytes of data;

a predetermined number of said first bytes of data representing a window of a plurality of windows used in examining said magnetic waveforms, with said predetermined number occurring when the speed of a document at said reading station is an anticipated normal speed;

a predetermined number of said second bytes of data being associated with said predetermined number of first bytes of data; and

processor means for withdrawing a predetermined number of said second bytes of data from said second memory and for calculating an average speed therefrom, whereby said average speed is compared with said anticipated normal speed to obtain a variation in speed, if any;

said variation in speed from said anticipated normal speed being used to adjust the predetermined number of said first bytes of data associated with a window to arrive at an adjusted window; and

said processor means also having means for examining said first bytes of data in said adjusted window for peaks for use in template matching.

2. The apparatus as claimed in claim 1 in which said processor means also includes means for scaling the width of a said adjusted window to make it compatible with known template matching.

3. The apparatus as claimed in claim 2 in which said periodic times of said sampling means and said velocity sensing means are in a ratio of four to one, with the periodic times of said sampling means occurring more frequently.

4. A method of reading magnetic characters on a document comprising the steps:

(a) moving a document in reading relationship with a reading means to generate a magnetic waveform corresponding to a character to be read on the document;

(b) converting the magnetic waveform into a plurality of bytes of data corresponding to said magnetic waveform;

(c) sampling said bytes of data at periodic times to produce first bytes of data corresponding to said magnetic waveform, with a predetermined number of said first bytes of data corresponding to an examining window of a plurality of examining windows when the document is moved at an anticipated velocity in reading relationship with said reading means;

(d) storing said first bytes of data in a first memory;

(e) periodically sensing the velocity of said document at said reading means to produce second bytes of data representing instantaneous velocities of said document at said reading means so that a predetermined number of second bytes of data correspond to a said examining window of said plurality of windows when the document is moved at said anticipated velocity;

(f) storing said second bytes of data in a second memory;

(g) calculating the average velocity of said document at said reading means from said predetermined number of second bytes of data; and

(h) adjusting the width of a said examining window compared to its associated said predetermined number of first bytes by changing the number of first bytes included in the examining window in accordance with a comparison of said average velocity with said anticipated velocity; and

(i) examining the first bytes of data within the adjusted examining window of step h for peaks for use in template matching.

5. The method as claimed in claim 4 in which said sampling step is effected using a rate which is a predetermined multiple of the rate of sensing the velocity of said document.

6. The method as claimed in claim 5 in which said examining step includes the step of scaling the width of the adjusted examining window so as to make it compatible with existing template matching techniques.