International Business Machines Corporation

Richard G Casey

Bernadette Oudot

Alex M Miller

Karl S Zilles

Hiroyasu Takahashi

Glen G Langdon, Jr

Stephen Kane Boyer

James C King

Robin Williams

Kazuharu Toyokawa

Kottappuram Mohamedali Mohiuddin

Yohji Nakamura

Patrick E Mantey

David R Ferguson

Steven K Boyer

Theodore D Friedman

An automated optical character recognition method is provided for use in conjunction with a programmable digital processing device. The method inputs a sequence of values representing one or more characters in an array of characters to be optically recognized. The values define one or more dimensional characteristics of the characters. From the input values, a standard dimensional value is determined from a frequency distribution of a selected one of the character dimensional characteristics. For each of the input characters, a set of normalized values is determined from the standard dimensional value. The normalized values correspond to the one or more character dimensional characteristics. Optical character recognition is thereafter performed using the normalized values.

Speed and recognition enhancement for OCR using normalized height/width position

Character recognition apparatus capable of automatically modifying its recognition tree is disclosed. When a character is erroneously recognized, its image is used to select a leaf node of the character recognition tree which has a high probability of recognizing the character correctly. The tree is split at that node, creating one new leaf node which is assigned to the character formerly assigned to the split node, and a second new leaf node assigned to the erroneously recognized character. The split node is then assigned to test a pel address in a character image which has a high probability of discriminating between the two characters of its new leaf nodes.

Image recognition apparatus

An apparatus and method for optical recognition of chemical graphics allows documents containing chemical structures to be optically scanned so that both the text and the chemical structures are recognized. The structures are converted directly into molecular structure files suitable for direct input into chemical databases, molecular modeling programs, image rendering programs and programs that perform real time manipulation of structures.

Apparatus and method for parsing a chemical string

Apparatus and method for optical recognition of chemical graphics

A machine-implemented method for extracting character data from printed forms of the type having generally straight lines and data regions separated by lines utilizes a processor coupled to an image scanning device and a data storage device. The method includes the steps of scanning a blank form so as to create a digital image of a first array of pixels, identifying the pixels in the array which form generally straight lines of connected line pixels, creating data masks in the array at locations separated by the identified lines, the data masks corresponding to the data regions in the printed form, scanning a filled-in form so as to create a digital image of a second array of pixels, identifying pixels in the second array which form generally straight lines of connected pixels, calculating the offset of the lines in the second array from the lines in the first array, locating the data masks created in the first array into the second array by use of the calculated offset, and extracting data corresponding to character pixels from the data masks in the second array.

Computer-implemented method for automatic extraction of data from

A method of creating a decision tree to enable a character recognition device to recognize characters in a new unknown font includes scanning a document printed in the unknown font to generate an array of pixels, with each pixel having a neighborhood state based on the colors of the surrounding pixels. A plurality of clusters of pixels, each representing a printed character, are identified. A pixel in one cluster is selected, and its neighborhood state determined. The neighborhood state is used to address a memory that has stored in it a probability table providing the probability, in a second font different from the new font being taught, that a pixel is black as a function of its neighborhood state. The stored probability of black associated with the neighborhood state of the select pixel is read from the memory and assigned as the probability of the selected pixel. The decision tree for the new font is generated using that assigned probability of black.

Method and apparatus for establishing pixel color probabilities for use

An improved method is described for transmitting facsimiles of scanned symbols. Prototype facsimiles of each symbol in a library are enhanced by averaging the representations of each scanned symbol with a respective previously created prototype facsimile for that symbol. The amount of white space at opposite sides of each symbol prototype is determined. The enhanced prototype facsimile for each scanned symbol is associated with positional parameters denoting the average white space at said opposite sides of each symbol.

Facsimile transmission using enhanced symbol prototypes with

Preliminary classification of a text character by examining a few of its PELs rather than requiring repeated comparisons of the entire pattern with a large set of prototypes is achieved by following a path through a decision tree in a root-to-leaf node direction. Each tree node specifies the location of a PEL in the character and the input pattern PEL value in turn determines the next branch to follow in the tree. The root node identifies the first PEL to be inspected while the final leaf indicates the class selected. The library pattern and inputs are compared. Upon match, an index number is generated and the next pattern processed. Upon a mismatch, the pattern will be registered as a new prototype in the library, the tree extended to include the new class, and the next pattern will be processed.

Adaptive facsimile compression using a dynamic extendable decision

There is disclosed a method and means for increasing the positional accuracy of operator controlled cursors engaged in the digitized encoding of graphic information such as line drawings. The method steps comprise those of digitizing the instantaneous contact position between the cursor and the data entry surface as reference coordinates; detecting any segment of a colored object upon the surface within a predetermined area about the cursor; ascertaining the location within the area of the centroid or the like of the detected segment; and digitizing said ascertained location as a displacement from the reference coordinates. Apparatus for practicing the method comprises an independently actuable cursor formed from a position encoder and an image scanner, the scanner generating a Boolean coded array of points counterpart to a preselected surface area. The scanned array is first buffered and then used to actuate an operator viewable LED display of the scanned array on one hand and said array is sent to a CPU on the other hand. In turn, the CPU calculates the coordinates of the centroid of that array area of contiguous points having the same designated Boolean (color) 0 or 1 value. Signals representing the calculated coordinates generated by the CPU then cause the LED display to differentially indicate the calculated centroid location to the operator such as by way of a flashing display element.

Growing community of inventors

Morgan Hill, CA, United States of America

Richard G Casey