WO2002039372A1 - Procede et dispositif d'enregistrement d'informations - Google Patents

Procede et dispositif d'enregistrement d'informations Download PDF

Info

Publication number
WO2002039372A1
WO2002039372A1 PCT/SE2001/002504 SE0102504W WO0239372A1 WO 2002039372 A1 WO2002039372 A1 WO 2002039372A1 SE 0102504 W SE0102504 W SE 0102504W WO 0239372 A1 WO0239372 A1 WO 0239372A1
Authority
WO
WIPO (PCT)
Prior art keywords
images
information
lines
image
characters
Prior art date
Application number
PCT/SE2001/002504
Other languages
English (en)
Inventor
Markus Andreasson
Per ÅSTRAND
Original Assignee
C Technologies Ab
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by C Technologies Ab filed Critical C Technologies Ab
Priority to AU2002214490A priority Critical patent/AU2002214490A1/en
Publication of WO2002039372A1 publication Critical patent/WO2002039372A1/fr

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
    • G06V30/10—Character recognition
    • G06V30/14—Image acquisition
    • G06V30/142—Image acquisition using hand-held instruments; Constructional details of the instruments

Definitions

  • An object of the present invention is to provide a scanner pen which enable fast recording of text in real time.
  • Another object of the invention is to provide a scanner pen which may be used at a distance from the text and may scan several lines of text in a single stroke as well as discrete pictures.
  • the method comprises converting the information in each of the images to a coded representation,, comparing the coded representation of said images for determining an overlap position, and assemblying the images to form a composite image.
  • the coded representation may be a character code, such as ASCII.
  • the coded representation may comprise a division of the information inside boarders, such as rectangles, each comprising portions of the information, such as words included in said information.
  • the composite image may be converted to a character code format, such as ASCII-code.
  • each image may be separetly converted into character code format, such as ASCII, before assemblying.
  • the method may further comprise determining structures in each of said images, such as direction of lines or text line directions in each • image . This may be accomplished by means of a Hough transformation of each image. This information may be used for adjusting the rotational position and/or perspective of each image in dependence of the direction of lines. The information may also be used for the division of the image in reactangles .
  • a concept of the present invention is to record a
  • the device may comprise a memory adapted to store the recorded images, which are to be converted into sets of characters, in the memory, and to convert the stored images into sets of characters after completion of the recording of the images .
  • a memory adapted to store the recorded images, which are to be converted into sets of characters, in the memory, and to convert the stored images into sets of characters after completion of the recording of the images .
  • the device is advantageously designed in such a man- ner that a user can hold it by hand and at a distance from a substrate to record text on the substrate.
  • the device may be adapted to correct the images for rotation before they are converted into sets of characters.
  • the correction for rotation can be carried out in a relatively simple way.
  • some optical character recognition programs can process also rotated images, in which case the rotation is not required.
  • the device may correct the images for perspective before they are converted into sets of characters since the optical character recognition may be facilitated if all letters have the same size in the images .
  • the device may be designed as a reading head which is connected to a calculating unit in which the recorded images are processed.
  • the device may be adapted to assemble the sets of characters by comparing the sets of characters in pairs, the sets of characters being compared in a number of relative positions displaced relative to each other.
  • a first character in a first set of characters is compared with characters in the second set of characters until correspondence is achieved or until the first character in the first set of characters has been compared with all characters in the second set of char- acters .
  • the correspondence of the second characters in the first set of characters is compared with the characters in the second set of characters .
  • the device may be adapted to store the recorded images that are to be converted into sets of characters along with a serial number indicating in which order the images have been recorded, and to assemble the sets of characters with the aid of the serial number for the images corresponding to the sets of characters.
  • the serial numbers may be used for the recorded images since then a large number of images are to be assembled.
  • a method for recording of text on a substrate comprising the step of imaging and digitally recording images of the text, the images having partly overlapping contents.
  • the method is characterized in that it comprises the steps of converting at least two of the recorded images into a set of characters, each with a plurality of characters, by means of optical character recognition, and putting together the sets of characters with the aid of the characters in the sets of characters.
  • the area recorded by the sensor may be arranged so that a plurality of lines of text are imaged in a recorded image.
  • a computer-readable storage medium in which a computer program is stored which is adapted to be used for conversion of digital images, which are recorded by an image sensor, into text.
  • the storage medium is characterized in that the computer program comprises instructions for making the computer receive digital images as input signals, convert the digital images into sets of characters, with a plurality of characters, by means of character recognition, and put together the sets of characters with the aid of the characters in the sets of characters .
  • Fig. 1 is a schematic view of a device according to a first embodiment of the present invention.
  • Fig. 2 is a schematic block diagram of parts of the embodiment of Fig l .
  • Fig. 3 is a schematic view of images of text on a sheet of paper, which are recorded according to the invention.
  • Figs. 4a and 4b are diagrams, illustrating the principle of the Hough transformation.
  • Fig. 5 is a diagram, which shows maximum points for the Hough transform of two different images.
  • Fig. 6 is a diagram and a histogram for illustrating the detection of start and end points of words.
  • Figs. 7a and 7b are diagrams for illustrating the division of the text images into words.
  • Fig. 8 is a diagram of two images converted to characters for assembling.
  • Fig. 9 is a diagram for illustrating lines of text in a recorded image.
  • Fig. 10 is a flow chart of the operation of a computer program according to the invention.
  • Fig. 11a, lib and lie are shematic representations of a text, division thereof in rectangles and display on a small display.
  • Fig. 1 discloses a scanner pen, comprises a casing 1 having approximately the same shape as a conventional highlighter. In one short side of the casing there is an opening 2, which is intended to be directed at an area on a substrate which a user wants to image.
  • the information carrier can be a sheet of paper.
  • the casing 1 essentially contains an optics part 3, an electronic circuitry part 4 and a power supply part 5.
  • the optics part 3 comprises a lens system 7, light- emitting diodes 6, and an optical sensor 8 constituting an interface with the electronic circuitry part 4.
  • the light-emitting diodes 6 may be used to increase the illumination.
  • Fig. 9 shows a first line 32 and a second line 33. Lines having approximately the same direction as the first line 32 will intersect a larger number of points than lines having approximately the same inclination as the second line 33 since the dark pixels in the recorded image 15 are positioned along lines of text having approximately the same direction as the first line 32.
  • Fig. 5 illustrates that the maximum number of intersections can be used to determine the orientation of the lines of text.
  • the circles 73 correspond to points in the Hough transform diagram where several Hough transform curves of points in Fig. 9 intersect, i.e. maximum of intersections.
  • the circles 73 corresponds to the image 15 and are positioned along a straight line> 34. From the distance between the circles 73, the distance between the lines of text can be determined.
  • the position of the intersecting line 34 along the ⁇ axis indicates the rotation of the recorded image.
  • the second line 35 in Fig. 5 corresponds to the image 14 and extends through a plurality of maximum indicated by crosses 36 in the diagram. The inclination of the second line 35 indicates that the image has a perspective, i.e.
  • the lines of text have different rotations.
  • the sligth different distances between the crosses 36 indicate that the image has a perspective with larger distances between the lines at the lower portion of the line 36.
  • the displacement of the second intersecting line 35 in relation to the first line along the ⁇ axis indicates that the lines of text are rotated in the recorded image.
  • the image may be adjusted for perspective and rotated, for example to the horizontal direction, which is the same direction as image 15.
  • image 15 corresponding to line 34 need no adjustment, while image 14 corresponding to line 35 needs adjustment of the perspective to make line 36 vertical and with approximately equidistant crosses 36 and adjustment as to the rotational position, to move line 35 to the same angle ⁇ as image 15, which may correspond to zero angle.
  • Fig. 6 indicates how the letter "e” 37 is detected.
  • the number of dark pixels 7 are counted in the vertical direction perpendicular to the line 38, which may be calculated as described above.
  • the number of dark pixels is zero up to the start 39 of the letter "e” and will again be zero at the end 40 of the letter “e” .
  • the number of dark pixels has been zero for a predetermined period, this is detected as the end of a word.
  • Fig. 7a corresponds to a first recorded image 14 and Fig. 7b corresponds to a second recorded image 15.
  • a length of a first word 43 in Fig. 7a has correspondence in a length of a second word 44 in Fig. 7b.
  • a length of a third word 45 in Fig. 7a has correspondence in a length of a fourth word 46 in Fig. 7b.
  • the length of the words for each line it is possible to carry out a rough putting- together or stitching or assembling of the two images.
  • the images are roughly assembled so that a sequence of word lengths in the first recorded image corresponds to a sequence of word lengths in the second recorded image.
  • the word lengths along different lines in the first recorded image should thus correspond to word lengths along corresponding lines in the second recorded image. In this way, it is possible to determine how the images are displaced relative to each other.
  • Fig. 3 shows how this may be used to sort out unnecessary images so that optical character recognition need not be carried out as regards all images that are recorded.
  • a first image 47 and a second image 49 completely overlap the area which is covered by a third recorded image 48 which is indicated by dashed lines.
  • the third recorded image 48 can be completely omitted, without optical character recognition being carried out.
  • Fig. 3 is only schematic, since the images often overlap each other to a much larger extent. Indeed, if the scanner pen is held approximately still, all images overlap more or less, and a substantial saving of computing power may be obtained by omitting images with duplicate information.
  • a partial overlap may be used for sorting out errors in the optical character recognition process or the assembling process, in which case overlapping images are not omitted or discarded.
  • FIG. 8 shows a first set of characters 50 with a plurality of characters 60 corresponding to the first recorded image 47 in Fig. 3 and a second set of characters 51 with a plurality of characters 61 corresponding to the second recorded image 49 in Fig. 3.
  • the text in the first set of characters and the second set of char- acters follows the direction of the lines of• text 62, which may have been obtained by the Hough transform process described above.
  • the first set of characters 50 and the second set of characters 51 are put together or assembled by comparing characters in the two sets of characters. Firstly, the first character 52 in the first set of characters is compared with each of the characters in the second set of characters 51.
  • the operation . X CQ ⁇ CQ cd X X
  • Every new image is compared with previous information in order to find out its orientation therein.
  • each image is processed both in the vertical direction as well as in the horizontal direction, both forwards and backwards.
  • it is possible to scan in two dimensions by the present invention. This is possible because of the division of the image in a coded representation which is less time consuming to process, either in the form of rectangles or similar or in the form of ASCII code.
  • Fig. 11a is a text that is to be scanned.
  • Fig. lib is the division of this text in rectangles.

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Multimedia (AREA)
  • Theoretical Computer Science (AREA)
  • Character Discrimination (AREA)
  • Character Input (AREA)

Abstract

L'invention concerne un dispositif et un procédé d'enregistrement d'un texte, consistant à former une image du texte sur un capteur photosensible (8). Ce dispositif permet de transformer l'image (14-17, 47-49) en une série de caractères (50, 51) par reconnaissance de caractères, puis d'assembler ces séries de caractères (50, 51) à l'aide des caractères.
PCT/SE2001/002504 2000-11-13 2001-11-13 Procede et dispositif d'enregistrement d'informations WO2002039372A1 (fr)

Priority Applications (1)

Application Number Priority Date Filing Date Title
AU2002214490A AU2002214490A1 (en) 2000-11-13 2001-11-13 Method and device for recording of information

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
SE0004144-2 2000-11-13
SE0004144A SE0004144L (sv) 2000-11-13 2000-11-13 Textpussling

Publications (1)

Publication Number Publication Date
WO2002039372A1 true WO2002039372A1 (fr) 2002-05-16

Family

ID=20281797

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/SE2001/002504 WO2002039372A1 (fr) 2000-11-13 2001-11-13 Procede et dispositif d'enregistrement d'informations

Country Status (3)

Country Link
AU (1) AU2002214490A1 (fr)
SE (1) SE0004144L (fr)
WO (1) WO2002039372A1 (fr)

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US4379282A (en) * 1979-06-01 1983-04-05 Dest Corporation Apparatus and method for separation of optical character recognition data
US5216725A (en) * 1990-10-31 1993-06-01 Environmental Research Institute Of Michigan Apparatus and method for separating handwritten characters by line and word
WO1998020446A1 (fr) * 1996-11-01 1998-05-14 C Technologies Ab Procede et dispositif d'enregistrement

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US4379282A (en) * 1979-06-01 1983-04-05 Dest Corporation Apparatus and method for separation of optical character recognition data
US5216725A (en) * 1990-10-31 1993-06-01 Environmental Research Institute Of Michigan Apparatus and method for separating handwritten characters by line and word
WO1998020446A1 (fr) * 1996-11-01 1998-05-14 C Technologies Ab Procede et dispositif d'enregistrement

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
EL GOWELY, K. ET AL.: "Multi-phase recognition of multifont photoscript arabic text", 10TH INTERNATIONAL CONFERENCE ON PATTERN RECOGNITION, 1990, vol. 1, 16 June 1990 (1990-06-16) - 21 June 1990 (1990-06-21), pages 700 - 702 *

Also Published As

Publication number Publication date
SE0004144L (sv) 2002-05-14
AU2002214490A1 (en) 2002-05-21
SE0004144D0 (sv) 2000-11-13

Similar Documents

Publication Publication Date Title
JP2986383B2 (ja) ライン走査画像のためのスキューを補正する方法及びその装置
KR100792106B1 (ko) 포착 이미지 투사장치 및 포착 이미지 정정방법
US6904183B2 (en) Image capture systems
US7221810B2 (en) Method and device for recording of information
KR100402043B1 (ko) 비접촉형 화상 판독 장치 및 그것을 이용한 시스템
US5611033A (en) Apparatus and method for automerging images by matching features and aligning images
US7313289B2 (en) Image processing method and apparatus and computer-readable storage medium using improved distortion correction
EP0701225B1 (fr) Système pour transcrire des images sur un tableau avec un scanner de tableau à caméra
US5581637A (en) System for registering component image tiles in a camera-based scanner device transcribing scene images
US6768813B1 (en) Photogrammetric image processing apparatus and method
US5892854A (en) Automatic image registration using binary moments
CA2227289C (fr) Rotation d'images codees sur la longueur d'un passage
US20020149808A1 (en) Document capture
US5164996A (en) Optical character recognition by detecting geo features
JP4419269B2 (ja) 二次元コード抽出方法
US20060215232A1 (en) Method and apparatus for processing selected images on image reproduction machines
WO2005059819A1 (fr) Procede, dispositif, programme informatique et support d'enregistrement de mouvement d'unite d'utilisateur
US20090167884A1 (en) Self-Similar Capture Systems
US6618497B1 (en) Photogrammetric image processing apparatus and method
JP4169464B2 (ja) 画像処理方法と画像処理装置及びコンピュータ読み取り可能な記録媒体
WO2002039372A1 (fr) Procede et dispositif d'enregistrement d'informations
JPH1115951A (ja) ずれ検出装置および画像合成装置
WO2007146129A2 (fr) Systèmes de capture d'image autosemblable
JP3240899B2 (ja) 文書画像入力装置
JP2007318341A (ja) 印刷物上の位置算出装置、印刷物上の位置算出システム、印刷物上の位置算出方法及び印刷物

Legal Events

Date Code Title Description
AK Designated states

Kind code of ref document: A1

Designated state(s): AE AG AL AM AT AT AU AZ BA BB BG BR BY BZ CA CH CN CO CR CU CZ CZ DE DE DK DK DM DZ EC EE EE ES FI FI GB GD GE GH GM HR HU ID IL IN IS JP KE KG KP KR KZ LC LK LR LS LT LU LV MA MD MG MK MN MW MX MZ NO NZ OM PH PL PT RO RU SD SE SG SI SK SK SL TJ TM TR TT TZ UA UG US UZ VN YU ZA ZW

AL Designated countries for regional patents

Kind code of ref document: A1

Designated state(s): GH GM KE LS MW MZ SD SL SZ TZ UG ZW AM AZ BY KG KZ MD RU TJ TM AT BE CH CY DE DK ES FI FR GB GR IE IT LU MC NL PT SE TR BF BJ CF CG CI CM GA GN GQ GW ML MR NE SN TD TG

121 Ep: the epo has been informed by wipo that ep was designated in this application
DFPE Request for preliminary examination filed prior to expiration of 19th month from priority date (pct application filed before 20040101)
REG Reference to national code

Ref country code: DE

Ref legal event code: 8642

122 Ep: pct application non-entry in european phase
NENP Non-entry into the national phase

Ref country code: JP

WWW Wipo information: withdrawn in national office

Country of ref document: JP