E-mabhuku uye mafomati awo: DjVu - nhoroondo yayo, zvakanakira, zvakaipa uye maficha

Mukutanga 70s, American munyori Michael Hart akakwanisa tora kuwana kusingagumi kune Xerox Sigma 5 komputa yakaiswa paYunivhesiti yeIllinois. Kuti ashandise zvakanaka zviwanikwa zvemuchina, akasarudza kugadzira bhuku rekutanga remagetsi, kudhindazve US Declaration of Independence.

Nhasi, mabhuku edhijitari apararira, zvakanyanya nekuda kwekuvandudzwa kwemidziyo inotakurika (mafoni efoni, e-readers, laptops). Izvi zvakakonzera kubuda kwenhamba huru yemafomati e-book. Ngatiedzei kunzwisisa maitiro avo uye tiudze nhoroondo yevanonyanya kufarirwa - ngatitange neiyo DjVu fomati.

E-mabhuku uye mafomati awo: DjVu - nhoroondo yayo, zvakanakira, zvakaipa uye maficha
/flickr/ Lane Pearman / CC

Kubuda kwechimiro

DjVu yakagadziridzwa muna 1996 neAT&T Labs nechinangwa chimwe - kupa vanogadzira webhu chishandiso chekuparadzira mifananidzo yakakwira-resolution paInternet.

Ichokwadi ndechokuti panguva iyoyo 90% yemashoko ose achiripo yakachengetwa pamapepa, uye akawanda emagwaro anokosha aiva nemifananidzo yemavara nemapikicha. Kuti uchengetedze kuverenga kwechinyorwa uye kunaka kwemifananidzo, zvaive zvakafanira kuita zviyero zvepamusoro-soro.

Classic web formats - JPEG, GIF uye PNG - yakaita kuti zvikwanise kushanda nemifananidzo yakadaro, asi nemutengo wehuwandu. Panyaya yeJPEG, kuitira kuti zvinyorwa yakaverengwa pachiratidziro chekutarisa, ndaifanira kuongorora gwaro racho nechisarudzo che300 dpi. Peji yeruvara rwemagazini yaitora 500 KB. Kudhawunirodha mafaera esaizi iyi kubva paInternet yaive basa rakawanda panguva iyoyo.

Imwe nzira yaive yekudhijita mapepa emapepa uchishandisa OCR tekinoroji, asi makore makumi maviri apfuura kurongeka kwavo kwaive kure nekunaka - mushure mekugadzirisa, mhedzisiro yaifanira kupepetwa nemaoko. Panguva imwecheteyo, magirafu nemifananidzo zvakaramba zviri "pamusoro". Uye kunyange dai zvaikwanisika kuisa chifananidzo chakaongororwa mugwaro rezvinyorwa, mamwe maonero ekuona akarasika, semuenzaniso, ruvara rwepepa, maitiro ayo, uye izvi zvakakosha zvikamu zvezvinyorwa zvekare.

Kugadzirisa matambudziko aya, AT&T yakagadzira DjVu. Zvakaita kuti zvikwanise kudzvanya zvinyorwa zvemavara akaongororwa ane resolution ye300 dpi kusvika 40-60 KB, ine saizi yepakutanga ye25 MB. DjVu yakaderedza saizi yemapeji matema nechena kusvika 10–30 KB.

DjVu inomanikidza sei magwaro

DjVu inogona kushanda nemapepa akaongororwa mapepa uye mamwe mafomati edhijitari, akadai sePDF. Iyo DjVu inoshanda sei nhema tekinoroji inotsemura mufananidzo muzvikamu zvitatu: kumberi, kumashure uye dema uye chena (bit) mask.

Iyo maski inochengetwa pakugadziriswa kwefaira rekutanga uye rine mufananidzo wezvinyorwa uye mamwe mashoko akajeka - mitsetse yakanaka uye madhayagiramu - pamwe chete nemifananidzo yakasiyana.

Iyo ine resolution ye300 dpi kuchengetedza mitsetse yakanaka uye mavara akajeka, uye inomanikidzwa uchishandisa JB2 algorithm, inova mutsauko weAT&T's JBIG2 algorithm yefax. Chimiro cheJB2 zviri chainoita ndechekuti inotsvaga mavara akapetwa pane peji uye inochengetedza mufananidzo wavo kamwe chete. Saka, muzvinyorwa zvemapeji akawanda, mapeji mashomanana akatevedzana anogovera "dictionary" yakafanana.

Iyo yekumashure ine manyorerwo epeji nemifananidzo, uye kugadzirisa kwayo kwakadzikira pane iyo yemask. Iyo isina kurasikirwa kumashure inochengetwa pa100 dpi.

Foreground zvitoro ruzivo rwemavara nezve mask, uye kugadzirisa kwayo kunowanzo kuderedzwa zvakanyanya, nekuti kazhinji mavara emavara akasviba uye akafanana kune imwe yakadhindwa hunhu. Inoshandiswa kumanikidza kumberi uye kumashure wavelet compression.

Nhanho yekupedzisira yekugadzira gwaro reDjVu ndeye entropy encoding, apo inoshandura arithmetic encoder inoshandura kutevedzana kwemavara akafanana kuita kukosha kwebhinari.

Zvakanakira chimiro

Basa raDjVu raive save "Zvinhu" zvegwaro repepa mune fomu yedhijitari, zvichibvumira kunyange makombiyuta asina simba kushanda nemagwaro akadaro. Naizvozvo, software yekuona mafaera eDjVu inokwanisa "kukurumidza kupa". Kutenda kwaari mundangariro loading Icho chete chidimbu cheiyo DjVu peji iyo inofanirwa kuratidzwa pachiratidziri.

Izvi zvakare zvinoita kuti zvikwanise kuona mafaera "asina kurodha", kureva, mapeji ega ega akawanda-mapeji DjVu gwaro. Muchiitiko ichi, kudhirowa kunofambira mberi kwemifananidzo kunoshandiswa, kana zvikamu zvichiita se "zvinooneka" sezvo faira rinotorwa (semuJPEG).

Makore makumi maviri apfuura, pakatangwa chimiro ichi, peji rakaiswa mumatanho matatu: kutanga chikamu chemavara chakaiswa, mushure memasekondi mashoma mavhezheni ekutanga emifananidzo uye kumashure akaiswa. Pashure pacho, peji yose yebhuku racho β€œyakaoneka.”

Kuvapo kwechimiro chetatu-nhanho inokubvumirawo kuti utsvage kuburikidza nemabhuku akaongororwa (sezvo paine chinyorwa chakakosha chinyorwa). Izvi zvakava nyore pakushanda nehunyanzvi mabhuku uye mabhuku ereferenzi, saka DjVu yakava hwaro hwemaraibhurari akati wandei emabhuku esainzi. Somuenzaniso, muna 2002 akasarudzwa Internet Archive seimwe yemafomati (pamwe neTIFF uye PDF) yepurojekiti yekuchengetedza mabhuku akaongororwa kubva kwakavhurika masosi.

Kuipa kwechimiro

Nekudaro, sematekinoroji ese, DjVu ine zvipingamupinyi zvayo. Semuyenzaniso, kana uchikodha scans emabhuku muiyo DjVu fomati, mamwe mavara ari mugwaro anogona kutsiviwa nemamwe akafanana pakuonekwa. Izvi zvinowanzoitika nemabhii "i" uye "n", ndicho chikonzero dambudziko iri kugamuchirwa zita "yin dambudziko". Izvo hazvibvi pamutauro wezvinyorwa uye zvinokanganisa, pakati pezvimwe zvinhu, nhamba uye mamwe madiki anodzokorora mavara.

Chikonzero chayo kukanganisa kwemaitiro muJB2 encoder. Iyo "inopatsanura" inoongorora mumapoka ezvikamu gumi kusvika makumi maviri uye inoita duramazwi rezviratidzo zvakajairika kuboka rega rega. Duramazwi rine mienzaniso yemavara akajairika nenhamba dzine mapeji uye zvinorongedzerwa zvekutaridzika kwavo. Paunoona bhuku reDjVu, mavara kubva muduramazwi anoiswa munzvimbo chaidzo.

Izvi zvinokutendera kuti uderedze saizi yefaira reDjVu, zvisinei, kana mavara maviri anoratidzwa akafanana, iyo encoder inogona kuvavhiringa kana kukanganisa zvakafanana. Dzimwe nguva izvi zvinotungamira mukukuvadzwa kune mafomula mugwaro rehunyanzvi. Kugadzirisa dambudziko iri, unogona kusiya compression algorithms, asi izvi zvichawedzera saizi yedhijitari kopi yebhuku.

Chimwe chinokanganisa chimiro ndechekuti haitsigirwe nekusarudzika mune akawanda emazuva ano masisitimu anoshanda (kusanganisira nharembozha). Naizvozvo, kushanda nayo iwe unofanirwa kuisa wechitatu-bato zvirongwa, zvakadai seDjVuReader, WinDjView, Evince, etc. Zvisinei, pano ndinoda kuona kuti vamwe vaverengi vemagetsi (somuenzaniso, ONYX BOOX) vanotsigira DjVu format "kunze kwebhokisi" - sezvo zvikumbiro zvinodiwa zvakatoiswa ipapo.

Nenzira, takataura nezve zvimwe zvikumbiro zve-Android-based vaverengi vanogona kuita mune imwe yapfuura zvinhu.

E-mabhuku uye mafomati awo: DjVu - nhoroondo yayo, zvakanakira, zvakaipa uye maficha
Muverengi ONYX BOOX Chronos

Rimwe dambudziko refomati rinoonekwa kana uchishanda neDjVu zvinyorwa pane zvidiki zvidzitiro zvenharembozha - smartphones, mahwendefa, vaverengi. Dzimwe nguva mafaira eDjVu anounzwa nenzira ye scan yebhuku rakapararira, uye mabhuku ehunyanzvi uye magwaro ekushanda anowanzo muA4 fomati, saka unofanirwa "kufambisa" mufananidzo uchitsvaga ruzivo.

Zvisinei, tinocherechedza kuti dambudziko iri rinogonawo kugadziriswa. Nzira iri nyore, hongu, ndeyekutsvaga gwaro mune imwe fomati - asi kana sarudzo iyi isingagoneki (semuenzaniso, iwe unofanirwa kushanda nehuwandu hukuru hwehunyanzvi mabhuku muDjVu), ipapo unogona kushandisa vaverengi vemagetsi. ine diagonal yakakura kubva 9,7 kusvika 13,3 inches, iyo yakanyatso "talored" yekushanda nemagwaro akadaro.

Semuenzaniso, mumutsara weONYX BOOX zvishandiso zvakadaro Chronos ΠΈ MAX 2 (nenzira, takagadzirira kuongororwa kwemuenzaniso uyu wekuverenga, uye tichakurumidza kuibudisa pa blog yedu), uyewo chitsamba, iyo ine E Ink Mobius Carta skrini ine diagonal ye 10,3 inches uye yakawedzera resolution. Zvishandiso zvakadaro zvinokutendera kuti uongorore zvakadzikama zvese zvemifananidzo muhukuru hwavo hwepakutanga uye zvakakodzera kune avo vanowanzo fanira kuverenga mabhuku edzidzo kana ehunyanzvi. Kuti uone DjVu uye mafaera ePDF inoshandiswa ne NEO Reader, iyo inokutendera kuti ugadzirise mutsauko uye ukobvu hwemafonti akadhindwa.

Pasinei nekukanganisa kwemafomati, nhasi DjVu inoramba iri imwe yeanonyanya kufarirwa mafomati e "kuchengetedza" mabhuku ezvinyorwa. Izvi zvinonyanya kukonzerwa nekuti iye zviri yakavhurika, uye zvimwe zvipimo zvetekinoroji nhasi zvinobvumira matekinoroji echizvino-zvino uye kusimukira kuti zvipfuure.

Muzvinyorwa zvinotevera tichaenderera mberi nenyaya yenhoroondo yekubuda kwemafomati e-book uye maitiro ebasa ravo.

PS Maseti akati wandei eONYX BOOX vaverengi:



Source: www.habr.com

Voeg