Changes between Version 4 and Version 5 of OCR_evaluation


Ignore:
Timestamp:
Jul 20, 2011, 3:04:37 PM (13 years ago)
Author:
Klaus Thoden
Comment:

--

Legend:

Unmodified
Added
Removed
Modified
  • OCR_evaluation

    v4 v5  
    44
    55The documents of the previous [wiki:OverviewWorkOrders2008 workflows] were assessed in terms of how well they might perform being OCRed.
    6 
    76 * Easy
    8   * Bernstein, 1897 (Fraktur)
    9   * [http://mpdl-proto.mpiwg-berlin.mpg.de/mpdl/interface/echo/echoDocuView.xql?document=/echo/fr/Berzelius_1819_WCWY69V2.xml Berzelius 1819]
    10   * [http://mpdl-proto.mpiwg-berlin.mpg.de/mpdl/interface/echo/echoDocuView.xql?document=/echo/de/Ampere_1844_FA4H1833.xml Ampère 1844]
     7  * [http://mpdl-dev.mpiwg-berlin.mpg.de/ECHOdocuViewfull?tocMode=thumbs&url=/mpiwg/online/permanent/einstein_exhibition/sources/GGAGCX1B&viewMode=images&tocPN=1&searchPN=1&mode=texttool&characterNormalization=reg&query=&pn=300&queryType= Bernstein, 1897] (Fraktur)
     8  * [http://mpdl-dev.mpiwg-berlin.mpg.de/ECHOdocuViewfull?url=%2Fmpiwg%2Fonline%2Fpermanent%2Flibrary%2FWCWY69V2&viewMode=images&tocMode=thumbs&tocPN=1&query=&searchPN=1&queryType=&mode=texttool&characterNormalization=reg&pn=20 Berzelius 1819]
     9  * [http://mpdl-dev.mpiwg-berlin.mpg.de/ECHOdocuViewfull?url=%2Fmpiwg%2Fonline%2Fpermanent%2Flibrary%2FFA4H1833&viewMode=images&tocMode=thumbs&tocPN=1&query=&searchPN=1&queryType=&mode=texttool&characterNormalization=reg&pn=20 Ampère 1844]
    1110 * Medium
    12   * [http://mpdl-proto.mpiwg-berlin.mpg.de/mpdl/interface/echo/echoDocuView.xql?document=/echo/la/Vitruvius_1511_XS9KA6WS.xml Vitruvius 1511]
    13   * [http://mpdl-proto.mpiwg-berlin.mpg.de/mpdl/interface/echo/echoDocuView.xql?document=/echo/la/Cataneo_1600.xml Cataneo 1600]
    14   * [http://mpdl-proto.mpiwg-berlin.mpg.de/mpdl/interface/echo/echoDocuView.xql?document=/echo/la/Aristoteles_1547_731P9Q9Y.xml Aristoteles 1547] (italics)
    15   * [http://mpdl-proto.mpiwg-berlin.mpg.de/mpdl/interface/echo/echoDocuView.xql?document=/echo/la/Archimedes_1565_4E7V2WGH.xml Archimedes 1565] (many pictures (problematic?))
    16   * [http://mpdl-proto.mpiwg-berlin.mpg.de/mpdl/interface/echo/echoDocuView.xql?document=/echo/it/Cataneo_1572_ZBAS6ZM1.xml Cataneo 1572] (bad printing)
    17   * [http://mpdl-proto.mpiwg-berlin.mpg.de/mpdl/interface/echo/echoDocuView.xql?document=/echo/la/Viviani_1659_QN4GHYBF.xml Viviani 1659]
    18   * [http://mpdl-proto.mpiwg-berlin.mpg.de/mpdl/interface/echo/echoDocuView.xql?document=/echo/it/Bianconi_1746_0347ZVRW.xml Bianconi 1746]
    19   * [http://mpdl-proto.mpiwg-berlin.mpg.de/mpdl/interface/echo/echoDocuView.xql?document=/echo/it/Zanotti_1752_16HBZHF5.xml Zanotti 1752]
    20   * [http://mpdl-proto.mpiwg-berlin.mpg.de/mpdl/interface/echo/echoDocuView.xql?document=/echo/de/Bion_1765_TGXUZC1H.xml Bion 1765] (blackletter)
    21   * [http://mpdl-proto.mpiwg-berlin.mpg.de/mpdl/interface/echo/echoDocuView.xql?document=/echo/la/Vitruvius_1800_V82APKX9.xml Vitruvius 1800] (clear, but weak printing, can see printing of the other side)
    22   * [http://mpdl-proto.mpiwg-berlin.mpg.de/mpdl/interface/echo/echoDocuView.xql?document=/echo/it/Gallaccini_1767_D09WWP72.xml Gallaccini 1767] (small font, but quite clear)
    23   * [http://mpdl-dev.mpiwg-berlin.mpg.de/ECHOdocuViewfull?url=/mpiwg/online/permanent/library/54K7BX2W Angeli 1668] (contains italics, otherwise very clear)
    24   * [http://mpdl-proto.mpiwg-berlin.mpg.de/mpdl/interface/echo/echoDocuView.xql?document=/echo/la/Trigault_1639_QA0BRSXP.xml Trigault 1639]
    25   * [http://mpdl-proto.mpiwg-berlin.mpg.de/mpdl/interface/echo/echoDocuView.xql?document=/echo/la/Bernoulli_1738_AZ870BWE.xml Bernoulli 1738]
     11  * [http://mpdl-dev.mpiwg-berlin.mpg.de/ECHOdocuViewfull?url=%2Fmpiwg%2Fonline%2Fpermanent%2Flibrary%2FXS9KA6WS&viewMode=images&tocMode=thumbs&tocPN=1&query=&searchPN=1&queryType=&mode=texttool&characterNormalization=reg&pn=20 Vitruvius 1511]
     12  * [http://mpdl-dev.mpiwg-berlin.mpg.de/ECHOdocuViewfull?url=%2Fmpiwg%2Fonline%2Fpermanent%2Farchimedes%2Fcatan_belli_502_la_1600&viewMode=images&tocMode=thumbs&tocPN=1&query=&searchPN=1&queryType=&mode=texttool&characterNormalization=reg&pn=20 Cataneo 1600]
     13  * [http://mpdl-dev.mpiwg-berlin.mpg.de/ECHOdocuViewfull?url=%2Fmpiwg%2Fonline%2Fpermanent%2Flibrary%2F731P9Q9Y&viewMode=images&tocMode=thumbs&tocPN=1&query=&searchPN=1&queryType=&mode=texttool&characterNormalization=reg&pn=20 Aristoteles 1547] (italics)
     14  * [http://mpdl-dev.mpiwg-berlin.mpg.de/ECHOdocuViewfull?url=%2Fmpiwg%2Fonline%2Fpermanent%2Flibrary%2F4E7V2WGH&viewMode=images&tocMode=thumbs&tocPN=1&query=&searchPN=1&queryType=&mode=texttool&characterNormalization=reg&pn=20 Archimedes 1565] (many pictures (problematic?))
     15  * [http://mpdl-dev.mpiwg-berlin.mpg.de/ECHOdocuViewfull?url=%2Fmpiwg%2Fonline%2Fpermanent%2Flibrary%2FZBAS6ZM1&viewMode=images&tocMode=thumbs&tocPN=1&query=&searchPN=1&queryType=&mode=texttool&characterNormalization=reg&pn=20 Cataneo 1572] (bad printing)
     16  * [http://mpdl-dev.mpiwg-berlin.mpg.de/ECHOdocuViewfull?url=%2Fmpiwg%2Fonline%2Fpermanent%2Flibrary%2FQN4GHYBF&viewMode=images&tocMode=thumbs&tocPN=1&query=&searchPN=1&queryType=&mode=texttool&characterNormalization=reg&pn=20 Viviani 1659]
     17  * [http://mpdl-dev.mpiwg-berlin.mpg.de/ECHOdocuViewfull?url=%2Fmpiwg%2Fonline%2Fpermanent%2Flibrary%2F0347ZVRW&viewMode=images&tocMode=thumbs&tocPN=1&query=&searchPN=1&queryType=&mode=texttool&characterNormalization=reg&pn=20 Bianconi 1746]
     18  * [http://mpdl-dev.mpiwg-berlin.mpg.de/ECHOdocuViewfull?url=%2Fmpiwg%2Fonline%2Fpermanent%2Flibrary%2F16HBZHF5&viewMode=images&tocMode=thumbs&tocPN=1&query=&searchPN=1&queryType=&mode=texttool&characterNormalization=reg&pn=20 Zanotti 1752]
     19  * [http://mpdl-dev.mpiwg-berlin.mpg.de/ECHOdocuViewfull?url=%2Fmpiwg%2Fonline%2Fpermanent%2Flibrary%2FTGXUZC1H&viewMode=images&tocMode=thumbs&tocPN=1&query=&searchPN=1&queryType=&mode=texttool&characterNormalization=reg&pn=20 Bion 1765] (blackletter)
     20  * [http://mpdl-dev.mpiwg-berlin.mpg.de/ECHOdocuViewfull?url=%2Fmpiwg%2Fonline%2Fpermanent%2Flibrary%2FV82APKX9&viewMode=images&tocMode=thumbs&tocPN=1&query=&searchPN=1&queryType=&mode=texttool&characterNormalization=reg&pn=20 Vitruvius 1800] (clear, but weak printing, can see printing of the other side)
     21  * [http://mpdl-dev.mpiwg-berlin.mpg.de/ECHOdocuViewfull?url=%2Fmpiwg%2Fonline%2Fpermanent%2Flibrary%2FD09WWP72&viewMode=images&tocMode=thumbs&tocPN=1&query=&searchPN=1&queryType=&mode=texttool&characterNormalization=reg&pn=20 Gallaccini 1767] (small font, but quite clear)
     22  * [http://mpdl-dev.mpiwg-berlin.mpg.de/ECHOdocuViewfull?tocMode=thumbs&url=/mpiwg/online/permanent/library/54K7BX2W&viewMode=images&tocPN=1&searchPN=1&mode=texttool&characterNormalization=reg&query=&pn=20&queryType= Angeli 1668] (contains italics, otherwise very clear)
     23  * [http://mpdl-dev.mpiwg-berlin.mpg.de/ECHOdocuViewfull?url=%2Fmpiwg%2Fonline%2Fpermanent%2Flibrary%2FQA0BRSXP&viewMode=images&tocMode=thumbs&tocPN=1&query=&searchPN=1&queryType=&mode=texttool&characterNormalization=reg&pn=20 Trigault 1639]
     24  * [http://mpdl-dev.mpiwg-berlin.mpg.de/ECHOdocuViewfull?url=%2Fmpiwg%2Fonline%2Fpermanent%2Flibrary%2FAZ870BWE&viewMode=images&tocMode=thumbs&tocPN=1&query=&searchPN=1&queryType=&mode=texttool&characterNormalization=reg&pn=20 Bernoulli 1738]
    2625 * Hard
    27   * [http://mpdl-proto.mpiwg-berlin.mpg.de/mpdl/interface/echo/echoDocuView.xql?document=/echo/la/Vitruvius_1544_2UZM8E2N.xml Vitruvius 1544]
    28   * [http://mpdl-proto.mpiwg-berlin.mpg.de/mpdl/interface/echo/echoDocuView.xql?document=/echo/de/Vitruvius_1757_DNBVYAV4.xml Vitruvius 1757] (Mixed blackletter and antiqua)
    29   * [http://mpdl-proto.mpiwg-berlin.mpg.de/mpdl/interface/echo/echoDocuView.xql?document=/echo/it/Zonca_1656_UR271U6Y.xml Zonca 1656] (bad printing)
    30   * [http://mpdl-proto.mpiwg-berlin.mpg.de/mpdl/interface/echo/echoDocuView.xql?document=/echo/en/Bacon_1670_WX8HY2V2.xml Bacon 1670] (contains italics, bad printing)
    31   * [http://mpdl-proto.mpiwg-berlin.mpg.de/mpdl/interface/echo/echoDocuView.xql?document=/echo/la/Clavius_1606_FBVYV7EH.xml Clavius 1606]
    32   * [http://mpdl-proto.mpiwg-berlin.mpg.de/mpdl/interface/echo/echoDocuView.xql?document=/echo/la/Barrow_1674_U19ERSE3.xml Barrow 1674] (bad printing, italics and Greek)
    33   * [http://mpdl-proto.mpiwg-berlin.mpg.de/mpdl/interface/echo/echoDocuView.xql?document=/echo/la/Gravesande_1721_KN9XTZRQ.xml Gravesande 1721]
    34   * [http://mpdl-proto.mpiwg-berlin.mpg.de/mpdl/interface/echo/echoDocuView.xql?document=/echo/fr/Vitruvius_1618_3XFC5KGV.xml Vitruvius 1618] (Thesaurus has columns)
    35   * [http://mpdl-proto.mpiwg-berlin.mpg.de/mpdl/interface/echo/echoDocuView.xql?document=/echo/fr/Mersenne_1635_508_fr.xml Mersenne 1635] (Microfilm)
    36   * [http://mpdl-proto.mpiwg-berlin.mpg.de/mpdl/interface/echo/echoDocuView.xql?document=/echo/la/Aristoteles_1548_9NN63YC9.xml Aristoteles 1548] (contains Greek)
    37   * [http://mpdl-proto.mpiwg-berlin.mpg.de/mpdl/interface/echo/echoDocuView.xql?document=/echo/it/Vitruvius_1556_XYTWCGV1.xml Vitruvius 1556] (extremly small font)
    38   * [http://mpdl-proto.mpiwg-berlin.mpg.de/mpdl/interface/echo/echoDocuView.xql?document=/echo/la/Aristoteles_1585_XSY685ZD.xml Aristoteles 1585] (bad printing)
    39   * [http://mpdl-proto.mpiwg-berlin.mpg.de/mpdl/interface/echo/echoDocuView.xql?document=/echo/de/Specklin_1599_SSM0YQED.xml Specklin 1599] (bad printing, blackletter)
    40   * [http://mpdl-proto.mpiwg-berlin.mpg.de/mpdl/interface/echo/echoDocuView.xql?document=/echo/la/Biancani_1635_GWS4WXH4.xml Biancani 1635] (very small font)
    41   * [http://mpdl-proto.mpiwg-berlin.mpg.de/mpdl/interface/echo/echoDocuView.xql?document=/echo/la/Vitruvius_1567_514_la.xml Vitruvius 1567] (very small font, mixed italics and upright)
    42   * [http://mpdl-proto.mpiwg-berlin.mpg.de/mpdl/interface/echo/echoDocuView.xql?document=/echo/la/Archimedes_1565_YS05QMU8.xml Archimedes 1565] (mixed italics and upright)
     26  * [http://mpdl-dev.mpiwg-berlin.mpg.de/ECHOdocuViewfull?url=%2Fmpiwg%2Fonline%2Fpermanent%2Flibrary%2F2UZM8E2N&viewMode=images&tocMode=thumbs&tocPN=1&query=&searchPN=1&queryType=&mode=texttool&characterNormalization=reg&pn=20 Vitruvius 1544]
     27  * [http://mpdl-dev.mpiwg-berlin.mpg.de/ECHOdocuViewfull?url=%2Fmpiwg%2Fonline%2Fpermanent%2Flibrary%2FDNBVYAV4&viewMode=images&tocMode=thumbs&tocPN=1&query=&searchPN=1&queryType=&mode=texttool&characterNormalization=reg&pn=20 Vitruvius 1757] (Mixed blackletter and antiqua)
     28  * [http://mpdl-dev.mpiwg-berlin.mpg.de/ECHOdocuViewfull?url=%2Fmpiwg%2Fonline%2Fpermanent%2Flibrary%2FUR271U6Y&viewMode=images&tocMode=thumbs&tocPN=1&query=&searchPN=1&queryType=&mode=texttool&characterNormalization=reg&pn=20 Zonca 1656] (bad printing)
     29  * [http://mpdl-dev.mpiwg-berlin.mpg.de/ECHOdocuViewfull?url=%2Fmpiwg%2Fonline%2Fpermanent%2Flibrary%2FWX8HY2V2&viewMode=images&tocMode=thumbs&tocPN=1&query=&searchPN=1&queryType=&mode=texttool&characterNormalization=reg&pn=20 Bacon 1670] (contains italics, bad printing)
     30  * [http://mpdl-dev.mpiwg-berlin.mpg.de/ECHOdocuViewfull?url=%2Fmpiwg%2Fonline%2Fpermanent%2Flibrary%2FFBVYV7EH&viewMode=images&tocMode=thumbs&tocPN=1&query=&searchPN=1&queryType=&mode=texttool&characterNormalization=reg&pn=20 Clavius 1606]
     31  * [http://mpdl-dev.mpiwg-berlin.mpg.de/ECHOdocuViewfull?url=%2Fmpiwg%2Fonline%2Fpermanent%2Flibrary%2FU19ERSE3&viewMode=images&tocMode=thumbs&tocPN=1&query=&searchPN=1&queryType=&mode=texttool&characterNormalization=reg&pn=20 Barrow 1674] (bad printing, italics and Greek)
     32  * [http://mpdl-dev.mpiwg-berlin.mpg.de/ECHOdocuViewfull?url=%2Fmpiwg%2Fonline%2Fpermanent%2Flibrary%2FKN9XTZRQ&viewMode=images&tocMode=thumbs&tocPN=1&query=&searchPN=1&queryType=&mode=texttool&characterNormalization=reg&pn=20 Gravesande 1721]
     33  * [http://mpdl-dev.mpiwg-berlin.mpg.de/ECHOdocuViewfull?url=%2Fmpiwg%2Fonline%2Fpermanent%2Flibrary%2F3XFC5KGV&viewMode=images&tocMode=thumbs&tocPN=1&query=&searchPN=1&queryType=&mode=texttool&characterNormalization=reg&pn=20 Vitruvius 1618] (Thesaurus has columns)
     34  * [http://mpdl-dev.mpiwg-berlin.mpg.de/ECHOdocuViewfull?tocMode=thumbs&url=/mpiwg/online/permanent/archimedes/merse_quest_508_fr_1635&viewMode=images&tocPN=1&searchPN=1&mode=texttool&characterNormalization=reg&query=&pn=20&queryType= Mersenne 1635] (Microfilm)
     35  * [http://mpdl-dev.mpiwg-berlin.mpg.de/ECHOdocuViewfull?url=%2Fmpiwg%2Fonline%2Fpermanent%2Flibrary%2F9NN63YC9&viewMode=images&tocMode=thumbs&tocPN=1&query=&searchPN=1&queryType=&mode=texttool&characterNormalization=reg&pn=20 Aristoteles 1548] (contains Greek)
     36  * [http://mpdl-dev.mpiwg-berlin.mpg.de/ECHOdocuViewfull?url=%2Fmpiwg%2Fonline%2Fpermanent%2Flibrary%2FXYTWCGV1&viewMode=images&tocMode=thumbs&tocPN=1&query=&searchPN=1&queryType=&mode=texttool&characterNormalization=reg&pn=20 Vitruvius 1556] (extremly small font)
     37  * [http://mpdl-dev.mpiwg-berlin.mpg.de/ECHOdocuViewfull?url=%2Fmpiwg%2Fonline%2Fpermanent%2Flibrary%2FXSY685ZD&viewMode=images&tocMode=thumbs&tocPN=1&query=&searchPN=1&queryType=&mode=texttool&characterNormalization=reg&pn=20 Aristoteles 1585] (bad printing)
     38  * [http://mpdl-dev.mpiwg-berlin.mpg.de/ECHOdocuViewfull?url=%2Fmpiwg%2Fonline%2Fpermanent%2Flibrary%2FSSM0YQED&viewMode=images&tocMode=thumbs&tocPN=1&query=&searchPN=1&queryType=&mode=texttool&characterNormalization=reg&pn=20 Specklin 1599] (bad printing, blackletter)
     39  * [http://mpdl-dev.mpiwg-berlin.mpg.de/ECHOdocuViewfull?url=%2Fmpiwg%2Fonline%2Fpermanent%2Flibrary%2FGWS4WXH4&viewMode=images&tocMode=thumbs&tocPN=1&query=&searchPN=1&queryType=&mode=texttool&characterNormalization=reg&pn=20 Biancani 1635] (very small font)
     40  * [http://mpdl-dev.mpiwg-berlin.mpg.de/ECHOdocuViewfull?url=%2Fmpiwg%2Fonline%2Fpermanent%2Farchimedes%2Fvitru_archi_514_la_1567&viewMode=images&tocMode=thumbs&tocPN=1&query=&searchPN=1&queryType=&mode=texttool&characterNormalization=reg&pn=80 Vitruvius 1567] (very small font, mixed italics and upright)
     41  * [http://mpdl-dev.mpiwg-berlin.mpg.de/ECHOdocuViewfull?url=%2Fmpiwg%2Fonline%2Fpermanent%2Flibrary%2FYS05QMU8&viewMode=images&tocMode=thumbs&tocPN=1&query=&searchPN=1&queryType=&mode=texttool&characterNormalization=reg&pn=20 Archimedes 1565] (mixed italics and upright)
    4342
    4443== Command overview ==
     
    4948 1. {{{ocropus-align -l OCRopus/ocropy/data/default.fst book/0001/??????.fst}}}
    5049 1. {{{ocropus-hocr book/}}}
     50
     51== First results ==
     52A page of Bacon 1670 ([http://mpdl-dev.mpiwg-berlin.mpg.de/ECHOdocuViewfull?tocMode=thumbs&url=/mpiwg/online/permanent/library/WX8HY2V2&viewMode=text&tocPN=1&searchPN=1&mode=texttool&characterNormalization=reg&query=&pn=150&queryType= link]) was OCRed with above commands. The output does not look too good (xhtml file attached):
     53
     54{{{
     55116
     56tha| yet it cannot be for other - Kcafo+: - ~r f1r(l, it is found 9tit upon
     57ccrta;n Trccs; and thofe Trces bear no fuch Fruit, os may allure that |rd to
     58B|rrics, and fo is of(cn found thcrc| which may have givcn occaf1on tothc
     59tale. But that which maketh an cnd of th|quc(lion is, that M1(Tcltoe hath
     60w|nld go into a Bv@gh( ?nd bcf1dcs/it feemcrh to bc morcfat and un@uous,
     61th1n the ordinary Sap of thc T!5c| both py thc Bcrry which is clammy, and
     62by that it connnucth grccn winter and Summcr, which Ihc Trcc do(h
     63not.
     64tryal would bc m-c,b|riPPing vf thc Bough of aCrab~cc inthe Ba~and
     65likely, to try it with fomeothcr watering or anointing, . that w7c notfonatu-
     66things askill not me Bongh.
     671t wcrc good to try, what T\&nM would put forth, @ thcy bc fo|biddcn
     68Io p@tfottb thcir natn|l Boughs: powl thcreforc a Tree; .andcover it,(omc
     69will Dut forth Roots 9 for fo will a Cions, bcir@ turncd down into clav.
     70fhat is .not fo f1atnral to thcplant as Clay isg tryit with Lcathcr,or Clotb, or
     71becn knodn t|gtow o\&t of apol9rd-
     72.A Man may count the prickcs / Trccs to bcakinde d Excrcfccnce,for
     73the/will ncvcr be@onghs| nor be Lcaves. @c plants that have p.ricklcs,
     74are Thorns, Black and whitc. s Bryer, Rofc, Ltmmon"trecsv Crab-trccq
     75Pricklcs in thc Lc2f .are, H9lly5 Juniper, vhin .bu(h, ~hi(bc i Ncttlcs alfo
     76have a fiall vencmous Fr~Ici fo hath Borrage, but harmlefs. Thc . @ufc
     77For the ha@c of the5pirit to put "Th, a@d the w|9t ot nourimmcty toput
     78forth a Bough, and thc clof~fs of thc Balks caufc Frickles .in Bonghsg 0nd
     79the Leav@s othcrw.!fcarc Wygh,as Burragc and Nettles are. As (or thc Leavcs
     80of Holl)|, they arefmooth, but nevcr Plain, bnr asit wcrc with folds for thc2
     81amc caufc.
     8256o. Therc be alfo Tlots, thavhough the? havc no ?~kles, yet they havea
     83kinde of Downey 9r Velvct Rine upop their Leafes; as Refe-C\<mpi0@ |te~
     84Cdrh2isns,,
     85}}}