Showing papers on "Optical character recognition published in 1991"

Handwritten alphanumeric character recognition by the neocognitron

Abstract: A pattern recognition system which works with the mechanism of the neocognitron, a neural network model for deformation-invariant visual pattern recognition, is discussed. The neocognition was developed by Fukushima (1980). The system has been trained to recognize 35 handwritten alphanumeric characters. The ability to recognize deformed characters correctly depends strongly on the choice of the training pattern set. Some techniques for selecting training patterns useful for deformation-invariant recognition of a large number of characters are suggested. >

Optical character recognition — a survey

Abstract: In order to highlight the interesting problems and actual results on the state of the art in optical character recognition (OCR), this paper describes and compares preprocessing, feature extraction and postprocessing techniques for commercial reading machines. Problems related to handwritten and printed character recognition are pointed out, and the functions and operations of the major components of an OCR system are described. Historical background on the development of character recognition is briefly given and the working of an optical scanner is explained. The specifications of several recognition systems that are commercially available are reported and compared.

Detection of highlighted regions

Abstract: A method and apparatus for detection of highlighted regions of a document. A document containing highlighted regions is scanned using a gray scale scanner. Morphology and threshold reduction techniques are used to separate highlighted and non-highlighted portions of the document. Having separated the highlighted and non-highlighted portions, optical character recognition (OCR) techniques can then be used to extract text from the highlighted regions.

Noise tolerant optical character recognition system

Abstract: Disclosed is a method of optical character recognition that first segments a graphical page image into word images. The method obtains a set of features by extracting smaller outlines of the dark regions in the word images, and then further dissecting each of the smaller outlines into small sections called micro-features. Micro-features are simply sections of character outlines, therefore, they can easily be extracted from the outlines of an entire word without any knowledge about character segmentation boundaries. Micro-features are extracted from an outline by finding the local extremities of the outline and then defining a micro-feature between each pair of sequential extremities. Once extracted, the micro-features are compared to micro-features from an ideal character in order to classify a character, and convert it into a character code.

Optical character recognition system and method

Abstract: A system and method are disclosed for optical scanning and recognition of alphanumeric characters of different sizes and fonts such as E13B, OCRA, OCRB, and Farington 7B type fonts recorded on documents such as bank drafts and checks. The system comprises an optical scanner, a microprocessor, a read-only-memory (ROM), a random-access-memory (RAM, with a stored program) for storing data and a plurality of predetermined character-identification patterns (templates). Each template includes three configurations: an actual pixel patterns (bit configuration) or a character, a configuration representing significant portions of the character which remain the same and do not change despite changes in size of the character, and a configuration representing portions of the character which are given added weight in distinguishing between similar characters. The scanner optically scans each document and produces a plurality of gray-scale pixel values which are stored in the RAM under control of the microprocessor. The microprocessor processes the stored pixel values, and effectively locates and segments each character on the document. The microprocessor then converts the segmented-character pixel value to binary data, and selects, from the plurality of templates, the template that matches the binary data, the matching template serving to identify (recognize) the segmented character.

A polygon-based technique for the automatic classification of text and graphics components from digitized paper-based forms

Abstract: A polygon-based graphics/text separation method is comprised of two sequential processes. First a raster to contour vector conversion step is used to convert a digitized bitmap into a collection of simple polygons. Next a component classification process is used to extract six particularly defined features of each of the individual polygon-based components to enable the separation of graphics and text polygons. Graphical polygons are further classified into four subclasses. Textual polygons are grouped into polygon strings (text strings). Each string contains a sequence of segmented character contour polygons which is ready for an optical character recognition algorithm to convert them into computer understandable ASCII characters.

High-speed document verification system

Abstract: A high-speed document verification system includes a document which is printed with a pattern having a predetermined arrangement of different reflectivity due to varying densities, line resolutions, or fluorescence. The arrangement represents information about the document. The document is fed into a high-speed document scanner sensitive to the varying ink densities or line resolutions. A graphic image of the document is produced by the scanner and this image or a graphic file of the image is checked to see if the proper pattern exists. A comparison unit, such as an optical character recognition system may be used to compare the scanned document's image with known density arrangements of valid documents to determine what information, if any, is represented by the arrangement. The graphic image may be sent to an operator's work station to be visually checked rather than being compared by the comparison unit or the image may be sent to the operator after it has been rejected by the comparison unit.

Transformation-ring-projection (trp) algorithm and its vlsi implementation

Abstract: The size-orientation-invariance characteristic plays an important role in pattern recognition. It has many applications in computer vision, optical character recognition (OCR), office automation, electronic publication, graphics, etc. In this paper, a new method called transformation-ring-projection (TRP) is proposed to achieve this characteristic. In TRP, shape transformation technique is employed to center the pattern image and normalize its size; the ring-projection scheme is used to handle the orientation problem. An experiment was conducted to verify the proposed method in character recognition. The TRP algorithm requires only simple and regular operations, and provides the feasibility or VLSI implementation to speed up computation for real-time processing. A study on VLSI architecture with extensive parallel processing and pipelining capabilities for the proposed TRP algorithm is also presented.

Character segmentation using an associative memory for optical character recognition

Abstract: Character segmentation apparatus for an optical character recognition system for segmenting individual character images in an image of a document having many characters prior to performing character identification, including a movable kernel for capturing a sub-image framed within a window having an area corresponding to an area occupied by an individual character, the window being movable in the document image in pixel-by-pixel steps to capture a sub-image for each step of the window, an associative memory for responding to the captured sub-image by producing a corresponding one of a set of images of known characters with which the associative memory has been trained, and a sensor responsive to the behavior of the associative memory for determining whether the sub-image is the image of an individual character or a non-character.

Adaptive post-processing of OCR text via knowledge acquisition

Abstract: Optical Character Recognit ion (OCR) is a convenient and eff ic ient tool for off ice automation and information retrieval, and is becoming more and more important in today's off ice and library environment. Depending on the text to be recognized, and the software and hardware employed, OCR software produces various types of errors in the recognized texts. The error types and their distributions are environment-dependent . In this paper, we first provide a classification for the types of errors that occur. An eff icient approach for the post-processing of recognition errors in OCR text is proposed so that errors can be e f f i ciently detected and corrected with the aid of a computer. The approach also allows for the correction of OCR errors to be partially automated. The major contribution of this approach is the capability of knowledge acquisition for OCR postprocessing which facilitates the eff ic ient correction of OCR errors. Through self-learning, the postprocessor is able to perform better and more accurately as processing proceeds. Experimental results are provided to demonstrate the eff ic iency and the effect iveness of this approach.

Research on Chinese OCR in Taiwan

Abstract: This paper describes typical research on Chinese optical character recognition in Taiwan. Chinese characters can be represented by a set of basic line segments called strokes. Several approaches to the recognition of handwritten Chinese characters by stroke analysis are described here. A typical optical character recognition (OCR) system consists of four main parts: image preprocessing, feature extraction, radical extraction and matching. Image preprocessing is used to provide the suitable format for data processing. Feature extraction is used to extract stable features from the Chinese character. Radical extraction is used to decompose the Chinese character into radicals. Finally, matching is used to recognize the Chinese character. The reasons for using strokes as the features for Chinese character recognition are the following. First, all Chinese characters can be represented by a combination of strokes. Second, the algorithms developed under the concept of strokes do not have to be modified when the number of characters increases. Therefore, the algorithms described in this paper are suitable for recognizing large sets of Chinese characters.

Rotation of images for optical character recognition

Abstract: A method and apparatus for properly orienting an text in order to perform optical character recognition (OCR). The text is digitized and placed into an image. The image is subsampled to determine an initial "guess" about the orientation of the image. If there are are specified number of sets of lines between lines having no black-to-white or white-to-black transitions, then the image is assumed to be oriented correctly. Otherwise, the image is assumed to be perpendicular to the line-of-sight of the OCR engine and the image is rotated 90 degrees counterclockwise in a preferred embodiment. A combination of rotations and trial OCR scans for the image is performed until the best results for the trial OCR is obtained or the maximum number of iterations is exceeded. Then, the remainder of OCR is performed on the image.

Rotationally impervious feature extraction for optical character recognition

Abstract: A feature-based optical character recognition system, employing a feature-based recognition device such as a neural network or an absolute distance measure device, extracts a set of features from segmented character images in a document, at least some of the extracted features being at least nearly impervious to rotation or skew of the document image, so as to enhance the reliability of the system. One rotationally invariant feature extracted by the system is the number of intercepts between boundary transitions in the image with at least a selected one of a plurality of radii centered at the centroid of the character in the image.

Rotation and position invariant optical character recognition

Abstract: A character recognition system wherein the flexibility of the recognition task is expanded for orientation independence, position independence, and orientation and position independence. The system also provides a technique for implementing concurrency in the processing to achieve high speed without sacrificing performance. The system is readily implemented on conventional machine vision computing systems.

OCR for sequentially displaying document layout according to recognition process

Abstract: In an optical character recognition device for optically recognizing characters written on a sheet, the sheet size, the location of the characters on the sheet and the like are determined based on a sheet image detected from the sheet by use of a scanner. Further, in the character recognition device according to a preferred embodiment of this invention, the sheet size, the location of the characters on the sheet and the like can be displayed on a layout displaying screen. With the character recognition device, the location of the processed character and the recognition result or the recognized character can be independently displayed on a substantially real time basis in the recognition process.

GLR Parsing With Probability

Abstract: Structured recursive hierarchical models such as context-free grammars have applications in many areas and the attractions of the LR parsing strategy are such that generalisations which enable this strategy to be employed in new areas should be warmly welcomed by many researchers. An area of particular interest to us is the recognition of continuous speech and it was with this application in mind that the probabilistic LR parser was developed. However, there are likely to be other applications as well, perhaps in handwriting and optical character recognition and in certain compilers.

Optical character recognition method and apparatus using context analysis and a parsing algorithm which constructs a text data tree

Abstract: A system for recognition of characters on a medium. The system includes a scanner for scanning a medium such as a page of printed text and graphics and producing a bit-mapped representation of the page. The bit-mapped representation of the page is then stored in a memory means such as the memory of a computer system. A processor processes the bit-mapped image to produce an output comprising coded character representations of the text on the page. The present invention discloses parsing a page to allow for production of the output characters in a logical sequence, a combination of feature detection methods and template matching methods for recognition of characters and a number of methods for feature detection such as use of statistical data and polygon fitting.

Survey: omnifont-printed character recognition

Abstract: This paper presents an overview of methods for recognition of omnifont printed Roman alphabet characters with various fonts, sizes and formats (plain, bold, etc.) from OCR system perspectives. First, it summarizes the current needs for optical printed character recognition (OPCR) in general, and then describes its importance for conversion between paper and electronic media. Current status of commercially available software and products for OPCR are briefly reviewed. Analysis indicates that the challenge we face in OPCR is far from being solved, and there is still a great gap between human needs and machine reading capabilities. Second, OPCR systems and algorithms are briefly reviewed and compared from the context of digital document processing for the following four stages: preprocessing of images, segmentation, recognition, and post-processing. Finally, possible research directions to improve the performance of OPCR systems are suggested, such as using an approach based on the combination of template matching and varieties of feature-based algorithms to recognize isolated characters, the use of multilayered architectures for OPCR, and parallel processing- based high-performance architectures.

Handwritten digits recognition by a supervised Kohonen-like learning algorithm

Abstract: The authors describe the application of a supervised learning algorithm, based on Kohonen's self-organizing feature maps, to pattern recognition. They adopt an idea previously used for semantic map organization and discuss its adaptation to pattern recognition. The basic motivation is to organize the map by the patterns and their association targets simultaneously. A by-product of this process is that the class labeling of neurons on the map emerges during the learning phase. The algorithm and results obtained for a handwritten zip code database are presented. >

Character recognition system

Abstract: An optical character recognition system which automatically reads handwritten characters and the like which do not have to be printed in a special format. Recognition tables derived from the pattern bit map and from a skeleton pixel matrix describe the character in terms of the relative position of the pixels along the borders of the character and provide a plurality of recognition strings, one in each table, for the front and rear views of the character and for the shape of the holes in the character which are opened from the top (as in the numeral four) or opened from the bottom (as in the numeral seven). From the recognition tables, the characters are recognized by searching recognition files containing blocks of successions of lines of code corresponding selectively to the codes in the recognition tables. The recognition file is arranged in hierarchal order so that the blocks in the file which represent characters having the lowest level of recognition difficulty in the character set to be recognized are searched first, the next highest level next and so forth. Recognition blocks for the next character in the group of blocks for the same difficulty of recognition level or to blocks for the next level of recognition difficulty. In this manner characters are recognized with a high degree of reliability and an indication of failure to recognize the character occurs rather than misrecognition.

Recognition and verification of hardwritten and hand-printer british postal addresses

Abstract: An algorithmic architecture for a high-performance optical character recognition (OCR) system for hand-printed and handwritten addresses is proposed. The architecture integrates syntactic and contextual post-processing with character recognition to optimise postcode recognition performance, and verifies the postcode against simple features extracted from the remainder of the address to ensure a low error rate. An enhanced version of the characteristic loci character recognition algorithm was chosen for the system to make it tolerant of variations in writing style. Feature selection for the classifier is performed automatically using the B/W algorithm. Syntactic and contextual information for hand-printed British postcodes have been integrated into the system by combining low-level postcode syntax information with a dictionary trie structure. A full implementation of the postcode dictionary trie is described. Features which define the town name effectively, and can easily be extracted from a handwritten or hand-printed town name are used for postcode verification. A database totalling 3473 postcode/address image has used to evaluate the performance of the complete postcode recognition process. The basic character recognition rate for the full unconstrained alphanumeric character set is 63.1%, compared with an expected maximum attainable 75–80%. The addition of the syntactic and contextual knowledge stages produces an overall postcode recognition rate which is equivalent to an alphanumeric character recognition rate of 86–90%. Separate verification experiments on a subset of 820 address images show that, with the first-order features chosen, an overall correct address feature code extraction rate of around 35% is achieved.

A multi-resolution approach to extract the address block on flat mail pieces

Abstract: A new approach is presented for dealing with the extraction of block address destination (BAD) on flat mail pieces. It is based on the construction of a pyramidal data structure which represents the image with a multi-resolution approach, so that much independence is provided with respect to variable parameters such as the size of characters, number of lines, and handwritten or machine printed. A top-down analysis is then performed on the pyramidal structure to interpret the various blocks segmented and their relationship between every level of the pyramid. Results are given in order to demonstrate the performances of this method. >

A handwritten Arabic character recognition technique for machine reader

Abstract: A technique for real-time recognition of unconstrained Arabic characters is presented. The proposed technique does not require any constraints of the character forms other than limiting them to a reasonable size and orientation. Structural features, which are more suitable for handwritten character recognition, are selected. Structural features that are independent of the writer style, which are called stable features, use a list of integer values (vector) to describe the character. On the other hand CHAIN CODE is used for other structural features (decisive) that are suitable for more variation of the writer style. A suitable clustering technique is chosen to accomplish the classifier procedure. The algorithm can be extended to cursive words after introducing the additional segmentation stage. >

New insights into correlation-based template matching

Abstract: Correction-based template matching has been used extensively in computer vision for object recognition and also for other tasks such as edge detection, stereo, motion and inspection. It has also found wide application in character recognition. A deeper understanding of the performance of this technique for such tasks would help predict when it will succeed or fail. Previous work on this problem has examined correlation-based template matching using signal processing techniques. Our approach is different: we dissect it employing concepts from geometry and physics. This leads to new insights into correlation-based template matching. We study the performance of correlation between images for different lighting conditions, viewpoints and scales of a scene, obtaining new results for scale variation and viewpoint change for binary images. We analyze gray level images for changes in lighting alone and obtain useful and novel formulae. Knowing how correlation behaves with these changes helps to strategically distribute templates for a given recognition task. We then develop a method to compute the probability of confusion for recognition by template matching. We obtain a closed form solution for the probability of confusion in the two template case. We conclude by noting that template matching encounters difficulties in tasks such as object recognition because of its strong dependence on viewing conditions, although it can be useful in some situations when templates are chosen and positioned judiciously.© (1991) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only.

Text-to-speech converter using optical character recognition - reads scanned text into memory for reproduction by loudspeaker or on video screen at discretion of user

Abstract: The text is scanned optically and read by a microprocessor into a digital memory, with conversion of the optical signals into ASCII code by an OCR computer program. The microprocessor is also in a position to render the ASCII encoded test into acoustic output. The scanned text can be also be reproduced on a screen with any desired deg. of enlargement, and scrolled under the control of the reader. USE/ADVANTAGE - By visually handicapped or reluctant readers, text can be understood more rapidly, accurately and economically than with conventional enlarging TV reprodn. equipment.

Table recognition for automated document entry system

Abstract: Most documents include various layout objects such as headlines text lines charts and tables. In particular tables are powerful tools that allow large quantities of data to be easily understood. An automated document entry system is needed that can recognize the document layout objects and extract the information from tables. In this paper an effective table recognition method is described. The proposed method is composed of three steps: (1) document layout structure recognition (2) table layout structure recognition (3) table content recognition. To develop the table layout structure recognition step we first examined the layout structure of tables in existing documents and classified several common structures. As a result of the examination we created ten rules and designed a ruled line and box extraction algorithm based on these rules. The effectiveness of the proposed method has been confirmed in experiments. Accordingly the proposed method will greatly contribute to the creation of an automated document entry system to allow faster document recognition and permit the data in tables to be extracted.

Recognition and verification of handwritten and hand-printed British postal addresses

Abstract: An algorithmic architecture for a high-performance optical character recognition (OCR) system for hand-printed and handwritten addresses is proposed. The architecture integrates syntactic and contextual post-processing with character recognition to optimise postcode recognition performance, and verifies the postcode against simple features extracted from the remainder of the address to ensure a low error rate

Optical character recognition apparatus

Abstract: An optical character recognition apparatus is comprised of an original document reading out section for generating an original character signal corresponding to light and shade of an original document, a character slice section for sequentially generating a slice signal corresponding to one character from the original character signal, a character recognition section for recognizing a character corresponding to the slice character signal, and a display section for displaying a recognized character in correspondence to the original document, wherein when a correction target character is found from the displayed characters, the correction target character and characters around the correction target character are displayed on the display section by utilizing the original character signal. Therefore, the operator can efficiently correct the correction target character without consulting the original document, and also the operator can easily confirm by an image of that original document that a separate character or the like is erroneously recognized.