{"id":54788,"date":"2020-01-04T00:00:00","date_gmt":"2020-01-03T21:00:00","guid":{"rendered":"https:\/\/prohoster.info\/blog\/blog_prohoster\/jpeg-algoritm-szhatiya"},"modified":"2020-02-18T14:02:50","modified_gmt":"2020-02-18T11:02:50","slug":"jpeg-algoritm-szhatiya","status":"publish","type":"post","link":"https:\/\/prohoster.info\/sk\/blog\/administrirovanie\/jpeg-algoritm-szhatiya","title":{"rendered":"JPEG. Algoritmus kompresie","gt_translate_keys":[{"key":"rendered","format":"text"}]},"content":{"rendered":"

Ahojte znova! Na\u0161iel som tento \u010dl\u00e1nok, nap\u00edsan\u00fd e\u0161te v m\u00e1ji 2019. Toto je pokra\u010dovanie s\u00e9rie \u010dl\u00e1nkov o WAVE a JPEG. Tu je prv\u00fd<\/a><\/noindex>. Tento pr\u00edspevok zahrnie inform\u00e1cie o algoritme k\u00f3dovania obr\u00e1zkov a o form\u00e1te samotnom.<\/p>\n

<\/p>\n

\u0160tipka hist\u00f3rie<\/h3>\n

<\/p>\n

Polievkov\u00e1 ly\u017eica z Wikip\u00e9die:<\/p>\n

<\/p>\n

JPEG (Joint Photographic Experts Group) \u2013 jeden z popul\u00e1rnych rastrov\u00fdch grafick\u00fdch form\u00e1tov, pou\u017e\u00edvan\u00fd na uchovanie fotografi\u00ed a podobn\u00fdch obr\u00e1zkov.<\/p><\/blockquote>\n

Tento \u0161tandard bol vyvinut\u00fd Spolo\u010dnou skupinou expertov na fotografiu u\u017e v roku 1991 na efekt\u00edvne kompresovanie obr\u00e1zkov.<\/p>\n

<\/a><\/noindex><\/p>\n

Ak\u00fa cestu prech\u00e1dzaj\u00fa obr\u00e1zky od surov\u00e9ho vzh\u013eadu po JPEG<\/h3>\n

<\/p>\n

Niektor\u00ed si myslia, \u017ee JPEG obr\u00e1zky s\u00fa komprimovan\u00e9 met\u00f3dou Huffmanovho k\u00f3dovania surov\u00e9 d\u00e1ta, ale to nie je pravda. Pred kontroln\u00fdm stla\u010den\u00edm prech\u00e1dzaj\u00fa d\u00e1ta dlh\u00fa cestu.<\/p>\n

<\/p>\n

Najprv sa farebn\u00e1 sch\u00e9ma zmen\u00ed z RGB na YCbCr. Na to existuje \u0161peci\u00e1lny algoritmus \u2013 tu<\/a><\/noindex>. Y sa nemen\u00ed, preto\u017ee zodpoved\u00e1 za jas a jeho zmena by bola vidite\u013en\u00e1.<\/p>\n

<\/p>\n

Prv\u00e1 vec, ktor\u00fa urobia s obr\u00e1zkom \u2013 je \u00abprerezovanie\u00bb<\/strong> (subsampling). Je to jednoduch\u00e9 pochopi\u0165: vezme sa 2x2 mrie\u017eka pixelov, \u010falej sa vezm\u00fa Cb a Cr \u2013 priemern\u00e9 hodnoty ka\u017ed\u00e9ho z komponentov YCbCr t\u00fdchto 4 pixelov. A tak, z\u00edskali sme 6 bajtov<\/em>, namiesto 4 Y, 4 Cb, 4 Cr sme dostali 4 Y a rovnak\u00e9 pre ka\u017ed\u00fd z nich Cb a Cr (4 + 4 + 4 = 12; 4 + 1 + 1 = 6; 12 \u2014 6 = 6). V r\u00e1mci aj 2\u00d72 kompresia s stratou s koeficientom kompresie 2:1 znie sol\u00eddne. To sa uplat\u0148uje na cel\u00e9 obr\u00e1zok. A tak \u2014 zn\u00ed\u017eili sme polovicu ve\u013ekosti. A tak\u00fato techniku m\u00f4\u017eeme pou\u017ei\u0165 v\u010faka na\u0161emu farebn\u00e9mu vn\u00edmaniu. \u010clovek si s \u013eahkos\u0165ou v\u0161imne rozdiel v jasnosti, ale nie v farbe, ak je priemern\u00fd v malej skupine px. Taktie\u017e prerezovanie m\u00f4\u017ee vykon\u00e1va\u0165 v l\u00ednii, 4 pixely horizont\u00e1lne a vertik\u00e1lne. Prv\u00fd variant sa pou\u017e\u00edva \u010dastej\u0161ie. Ak je kvalita obr\u00e1zka d\u00f4le\u017eit\u00e1, tak prerezovanie sa v\u00f4bec nevykon\u00e1va.
\nN\u00e1zorn\u00e1 ilustr\u00e1cia zjednodu\u0161enia (Habr neumo\u017enil vlo\u017ei\u0165 gif) \u2013 https:\/\/i.ibb.co\/Rg5Th9H\/150953010617579181.gif<\/a><\/noindex><\/p>\n

<\/p>\n

Hlavn\u00e1 \u010das\u0165 pr\u00edpravy<\/h5>\n

\nDCT<\/p>\n

Teraz naj\u0165a\u017e\u0161ia a potrebn\u00e1 \u010das\u0165. Cel\u00fd obr\u00e1zok sa rozdel\u00ed na bloky 8\u00d78 (pou\u017e\u00edvaj\u00fa sa vyplnenia v pr\u00edpade, ak rozl\u00ed\u0161enie nie je delite\u013en\u00e9 stranou bloku).<\/p>\n

<\/p>\n

Teraz sa k ka\u017ed\u00e9mu bloku aplikuje DCT (Diskr\u00e9tna kos\u00ednusov\u00e1 transform\u00e1cia)<\/strong>. V tejto \u010dasti sa z obr\u00e1zka vystrihne v\u0161etko zbyto\u010dn\u00e9. Pou\u017e\u00edvaj\u00fac DCP treba pochopi\u0165, popisuje tento blok (8\u00d78) nejak\u00fa monot\u00f3nnu \u010das\u0165 obr\u00e1zka: neba, steny; alebo obsahuje zlo\u017eit\u00fa \u0161trukt\u00faru (vlasy, symboly a pod.). Logicky, 64 podobn\u00fdch farebn\u00fdch pixelov sa m\u00f4\u017ee pop\u00edsa\u0165 len 1-\u00fdm, preto\u017ee ve\u013ekos\u0165 bloku je u\u017e zn\u00e1ma. Tu m\u00e1te kompresiu: 64 k 1.<\/p>\n

<\/p>\n

DCT premen\u00ed blok na spektrum, a tam, kde sa hodnoty ostro menia, koeficient sa st\u00e1va kladn\u00fdm, a \u010d\u00edm ostrej\u0161\u00ed je prechod, t\u00fdm vy\u0161\u0161\u00ed bude v\u00fdstup. Tam, kde je koeficient vy\u0161\u0161\u00ed, s\u00fa na obr\u00e1zku zobrazen\u00e9 jasn\u00e9 prechody v farbe a jasu, kde je ni\u017e\u0161\u00ed \u2014 slab\u00e9 (hladk\u00e9) zmeny hodn\u00f4t komponentov YCbCr v bloku.<\/p>\n

\nKvantovanie<\/p>\n

Tu sa u\u017e aplikuj\u00fa nastavenia kompresie. Ka\u017ed\u00fd z koeficientov v ka\u017edej z mat\u00edc 8\u00d78 sa del\u00ed na ur\u010dit\u00fd po\u010det. Ak kvalitu obr\u00e1zka po v\u0161etk\u00fdch jeho \u00faprav\u00e1ch viac zmen\u0161ova\u0165 nebudete, tak delite\u013e by mal by\u0165 jedni\u010dkou. Ak m\u00e1te v\u00e4\u010d\u0161\u00ed z\u00e1ujem o pam\u00e4\u0165, ktor\u00fa t\u00e1to fotografia zaber\u00e1, tak delite\u013e bude v\u00e4\u010d\u0161\u00ed ako 1, a podiel sa zaokr\u00fah\u013euje. Takto \u010dasto po zaokr\u00fahlen\u00ed vznik\u00e1 ve\u013ea n\u00fal.<\/p>\n

<\/p>\n

Kvantovanie sa rob\u00ed na vytvorenie mo\u017enosti e\u0161te v\u00e4\u010d\u0161ej kompresie. Takto to vyzer\u00e1 na pr\u00edklade kvantovania grafu y = sin(x):<\/p>\n

<\/p>\n

\"JPEG.<\/p>\n

\nKomprim\u00e1cia<\/p>\n

Najprv prech\u00e1dzame maticou zig-zagom:<\/p>\n

<\/p>\n

\"JPEG.<\/p>\n

<\/p>\n

Z\u00edskame jednorozmern\u00e9 pole s \u010d\u00edslami. Vid\u00edme, \u017ee v \u0148om je ve\u013ea n\u00fal, ktor\u00e9 m\u00f4\u017eeme odstr\u00e1ni\u0165. Na to namiesto sekvencie z mnoh\u00fdch n\u00fal zapisujeme 1 nulu a po nej \u010d\u00edslo, ktor\u00e9 ozna\u010duje ich po\u010det v sekvencii. T\u00fdmto sp\u00f4sobom m\u00f4\u017eeme zmen\u0161i\u0165 cel\u00e9 pole a\u017e na 1\/3 jeho ve\u013ekosti. A \u010falej jednoducho skomprimujeme toto pole met\u00f3dou Huffmana a zap\u00ed\u0161eme u\u017e do samotn\u00e9ho s\u00faboru.<\/p>\n

<\/p>\n

Kde sa pou\u017e\u00edva<\/h3>\n

<\/p>\n

V\u0161ade. Tak ako PNG, JPEG sa pou\u017e\u00edva vo fotoapar\u00e1tach, OS\u2019s (ako log\u00e1 firiem, ikony aplik\u00e1ci\u00ed, thumbnail\u2019y) a vo v\u0161etk\u00fdch mo\u017en\u00fdch oblastiach, kde je potrebn\u00e9 efekt\u00edvne uklada\u0165 obr\u00e1zky.<\/p>\n

<\/p>\n

Z\u00e1ver<\/h3>\n

<\/p>\n

V s\u00fa\u010dasnosti s\u00fa vedomosti o JPEG cenn\u00e9 iba na vzdel\u00e1vacie \u00fa\u010dely, preto\u017ee je u\u017e v\u0161ade zabudovan\u00fd a optimalizovan\u00fd ve\u013ek\u00fdmi skupinami \u013eud\u00ed, ale granit vedy je v\u0161ak st\u00e1le chutn\u00fd.<\/p>\n

\nZdroje<\/p>\n

\u010cl\u00e1nok o YCbCr na Wikip\u00e9dii<\/a><\/noindex>
\n\u010cl\u00e1nok o JPEG na Wikip\u00e9dii<\/a><\/noindex>
\nTrochu o DCT z pr\u00edspevku Pikabu<\/a><\/noindex>
\n\u010cl\u00e1nok o DCT na Wikip\u00e9dii<\/a><\/noindex><\/p>\n

Zdroj: habr.com<\/a><\/p>","protected":false,"gt_translate_keys":[{"key":"rendered","format":"html"}]},"excerpt":{"rendered":"

\u0418 \u0441\u043d\u043e\u0432\u0430 \u0437\u0434\u0440\u0430\u0432\u0441\u0442\u0432\u0443\u0439\u0442\u0435! \u042f \u043d\u0430\u0448\u0435\u043b \u044d\u0442\u0443 \u0441\u0442\u0430\u0442\u044c\u044e, \u043d\u0430\u043f\u0438\u0441\u0430\u043d\u043d\u0443\u044e \u0435\u0449\u0435 \u043c\u0430\u0435 2019-\u043e\u0433\u043e \u0433\u043e\u0434\u0430. \u042d\u0442\u043e \u2014 \u043f\u0440\u043e\u0434\u043e\u043b\u0436\u0435\u043d\u0438\u0435 \u0441\u0435\u0440\u0438\u0438 \u0441\u0442\u0430\u0442\u0435\u0439 \u043e WAVE \u0438 JPEG, \u0412\u043e\u0442 \u043f\u0435\u0440\u0432\u0430\u044f. \u042d\u0442\u0430 \u043f\u0443\u0431\u043b\u0438\u043a\u0430\u0446\u0438\u044f \u0432\u043a\u043b\u044e\u0447\u0438\u0442 \u0432 \u0441\u0435\u0431\u0435 \u0438\u043d\u0444\u043e\u0440\u043c\u0430\u0446\u0438\u044e \u043e\u0431 \u0430\u043b\u0433\u043e\u0440\u0438\u0442\u043c\u0435 \u043a\u043e\u0434\u0438\u0440\u043e\u0432\u0430\u043d\u0438\u044f \u0438\u0437\u043e\u0431\u0440\u0430\u0436\u0435\u043d\u0438\u0439 \u0438 \u043e \u0441\u0430\u043c\u043e\u043c \u0444\u043e\u0440\u043c\u0430\u0442\u0435 \u0432 \u0446\u0435\u043b\u043e\u043c. \u0429\u0435\u043f\u043e\u0442\u043a\u0443 \u0438\u0441\u0442\u043e\u0440\u0438\u0438 \u0421\u0442\u043e\u043b\u043e\u0432\u0443\u044e \u043b\u043e\u0436\u043a\u0443 \u0441\u0442\u0430\u0442\u044c\u0438 \u0438\u0437 \u0412\u0438\u043a\u0438\u043f\u0435\u0434\u0438\u0438: JPEG (Joint Photographic Experts Group) \u2014 \u043e\u0434\u0438\u043d \u0438\u0437 \u043f\u043e\u043f\u0443\u043b\u044f\u0440\u043d\u044b\u0445 […]<\/p>\n","protected":false,"gt_translate_keys":[{"key":"rendered","format":"html"}]},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[688],"tags":[],"class_list":["post-54788","post","type-post","status-publish","format-standard","hentry","category-administrirovanie"],"aioseo_notices":[],"aioseo_head":"\n\t\t\n\t\n\t\n\t\n\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\n","aioseo_head_json":{"title":"\ud83e\udd47JPEG. Algoritmus kompresie | ProHoster","description":"Ahoj znova! Na\u0161iel som tento \u010dl\u00e1nok, nap\u00edsan\u00fd e\u0161te v m\u00e1ji 2019. Toto je pokra\u010dovanie s\u00e9rie \u010dl\u00e1nkov o WAVE a JPEG. Tu je prv\u00fd.","canonical_url":"https:\/\/prohoster.info\/sk\/blog\/administrirovanie\/jpeg-algoritm-szhatiya","robots":"max-image-preview:large","keywords":"","webmasterTools":{"miscellaneous":""},"schema":null,"og:locale":"sk_SK","og:site_name":"ProHoster | \u041a\u0443\u043f\u0438\u0442\u044c \u043d\u0430\u0434\u0435\u0436\u043d\u044b\u0439 \u0445\u043e\u0441\u0442\u0438\u043d\u0433 \u0434\u043b\u044f \u0441\u0430\u0439\u0442\u043e\u0432 \u0441 \u0437\u0430\u0449\u0438\u0442\u043e\u0439 \u043e\u0442 DDoS, VPS VDS \u0441\u0435\u0440\u0432\u0435\u0440\u044b","og:type":"article","og:title":"\ud83e\udd47JPEG. \u0410\u043b\u0433\u043e\u0440\u0438\u0442\u043c \u0441\u0436\u0430\u0442\u0438\u044f | ProHoster","og:description":"\u0418 \u0441\u043d\u043e\u0432\u0430 \u0437\u0434\u0440\u0430\u0432\u0441\u0442\u0432\u0443\u0439\u0442\u0435! \u042f \u043d\u0430\u0448\u0435\u043b \u044d\u0442\u0443 \u0441\u0442\u0430\u0442\u044c\u044e, \u043d\u0430\u043f\u0438\u0441\u0430\u043d\u043d\u0443\u044e \u0435\u0449\u0435 \u043c\u0430\u0435 2019-\u043e\u0433\u043e \u0433\u043e\u0434\u0430. \u042d\u0442\u043e \u2014 \u043f\u0440\u043e\u0434\u043e\u043b\u0436\u0435\u043d\u0438\u0435 \u0441\u0435\u0440\u0438\u0438 \u0441\u0442\u0430\u0442\u0435\u0439 \u043e WAVE \u0438 JPEG, \u0412\u043e\u0442 \u043f\u0435\u0440\u0432\u0430\u044f.","og:url":"https:\/\/prohoster.info\/sk\/blog\/administrirovanie\/jpeg-algoritm-szhatiya","og:image":"https:\/\/prohoster.info\/wp-content\/uploads\/2021\/11\/logo-350.jpg","og:image:secure_url":"https:\/\/prohoster.info\/wp-content\/uploads\/2021\/11\/logo-350.jpg","og:image:width":350,"og:image:height":350,"article:published_time":"2020-01-03T21:00:00+00:00","article:modified_time":"2020-02-18T11:02:50+00:00","article:publisher":"https:\/\/www.facebook.com\/prohoster","article:author":"https:\/\/www.facebook.com\/prohoster"},"aioseo_meta_data":{"post_id":"54788","title":null,"description":null,"keywords":null,"keyphrases":null,"primary_term":null,"canonical_url":null,"og_title":null,"og_description":null,"og_object_type":"default","og_image_type":"default","og_image_url":null,"og_image_width":null,"og_image_height":null,"og_image_custom_url":null,"og_image_custom_fields":null,"og_video":null,"og_custom_url":null,"og_article_section":null,"og_article_tags":null,"twitter_use_og":false,"twitter_card":"default","twitter_image_type":"default","twitter_image_url":null,"twitter_image_custom_url":null,"twitter_image_custom_fields":null,"twitter_title":null,"twitter_description":null,"schema":{"blockGraphs":[],"customGraphs":[],"default":{"data":{"Article":[],"Course":[],"Dataset":[],"FAQPage":[],"Movie":[],"Person":[],"Product":[],"ProductReview":[],"Car":[],"Recipe":[],"Service":[],"SoftwareApplication":[],"WebPage":[]},"graphName":"","isEnabled":true},"graphs":[]},"schema_type":null,"schema_type_options":null,"pillar_content":false,"robots_default":true,"robots_noindex":false,"robots_noarchive":false,"robots_nosnippet":false,"robots_nofollow":false,"robots_noimageindex":false,"robots_noodp":false,"robots_notranslate":false,"robots_max_snippet":null,"robots_max_videopreview":null,"robots_max_imagepreview":"large","priority":null,"frequency":null,"local_seo":null,"seo_analyzer_scan_date":"2026-01-24 12:45:19","breadcrumb_settings":null,"limit_modified_date":false,"reviewed_by":null,"ai":null,"created":"2021-02-28 10:01:54","updated":"2026-01-24 12:45:19","focus_keyword":null,"additional_keywords":null,"truseo_locale":null},"gt_translate_keys":[{"key":"link","format":"url"}],"_links":{"self":[{"href":"https:\/\/prohoster.info\/sk\/wp-json\/wp\/v2\/posts\/54788","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/prohoster.info\/sk\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/prohoster.info\/sk\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/prohoster.info\/sk\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/prohoster.info\/sk\/wp-json\/wp\/v2\/comments?post=54788"}],"version-history":[{"count":0,"href":"https:\/\/prohoster.info\/sk\/wp-json\/wp\/v2\/posts\/54788\/revisions"}],"wp:attachment":[{"href":"https:\/\/prohoster.info\/sk\/wp-json\/wp\/v2\/media?parent=54788"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/prohoster.info\/sk\/wp-json\/wp\/v2\/categories?post=54788"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/prohoster.info\/sk\/wp-json\/wp\/v2\/tags?post=54788"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}