{"id":181944,"date":"2026-06-08T02:48:45","date_gmt":"2026-06-08T00:48:46","guid":{"rendered":"https:\/\/prohoster.info\/blog\/novosti-interneta\/opencv-5-0-0"},"modified":"2026-06-08T02:48:45","modified_gmt":"2026-06-08T00:48:46","slug":"opencv-5-0-0","status":"publish","type":"post","link":"https:\/\/prohoster.info\/hr\/blog\/news\/opencv-5-0-0","title":{"rendered":"OpenCV 5.0.0","gt_translate_keys":[{"key":"rendered","format":"text"}]},"content":{"rendered":"<p><img decoding=\"async\" src=\"\/wp-content\/uploads\/2026\/06\/0827c35d7663a25a36859f66c23893bc.jpg\" style=\"display:block;margin: 0 auto;\" \/><\/p>\n<p><noindex>      <noindex><\/p>\n<p>Objavljeno <strong><a rel=\"nofollow\" href=\"https:\/\/github.com\/opencv\/opencv\">OpenCV 5.0.0<\/a><\/strong> \u2014 velika objava otvorene biblioteke za ra\u010dunalni vid i obradu slika. Izdanje je objavljeno <strong>6. lipnja 2026. godine<\/strong> i ve\u0107 je ozna\u010deno kao aktualno u repozitoriju projekta na GitHubu. OpenCV se distribuira pod licencom <strong>Apache-2.0<\/strong>.<\/p>\n<p><\/noindex><\/noindex><\/p>\n<p>Svrha izdanja je prelazak OpenCV-a od \"klasi\u010dne biblioteke ra\u010dunalnog vida\" prema modernijoj platformi za CV\/AI zadatke: projekt je o\u010distio staro API, preradio modul za neuronsko umnogavanje, pro\u0161irio podr\u0161ku za ONNX, dodao temelje za VLM\/LLM scenarije i a\u017eurirao 3D\/kalibracijski stog. Razvijatelji opisuju OpenCV 5.0 kao granu koja se tako\u0111er sada smatra stabilnom, a nova funkcionalnost \u0107e ve\u0107inom biti uklju\u010dena tamo.<\/p>\n<h3>Glavne promjene<\/h3>\n<ul>\n<li>\n<p><strong>Novi DNN inferencijski motor.<\/strong><br \/>\nU modulu dnn uveden je novi motor koji su\u017eivotno djeluje sa starim. Bolje funkcionira s dinami\u010dkim oblicima tenzora, podgrafovima i modernim mogu\u0107nostima ONNX-a. Potvr\u0111eno pokrivanje vi\u0161e <strong>od 80% specifikacije ONNX<\/strong>, dok je u OpenCV 4.x pokriveno manje od 23%.<br \/>\nTo je vjerojatno glavna promjena u izdanju: OpenCV nastoji postati ne samo omota\u010d oko modela, ve\u0107 i potpuno ugra\u0111eni inferencijski motor za moderne neuronske mre\u017ee.<\/p>\n<\/li>\n<li>\n<p><strong>Odabir motora pri u\u010ditavanju modela.<\/strong><br \/>\nU cv::dnn::readNet() i srodnim funkcijama uveden je parametar ENGINE_AUTO. Po defaultu, OpenCV prvo poku\u0161ava novi motor, a u slu\u010daju neuspjeha vra\u0107a se na klasi\u010dni. Pona\u0161anje se tako\u0111er mo\u017ee odrediti putem varijable okru\u017eenja OPENCV_FORCE_DNN_ENGINE: koristiti stari motor, novi, automatski re\u017eim ili ONNX Runtime. Stari projekti nisu obavezni odmah se potpuno prebaciti na novi motor, ali nove modele je mogu\u0107e postupno prevesti na njega.<\/p>\n<\/li>\n<li>\n<p><strong>Integracija s ONNX Runtime.<\/strong><br \/>\nOpenCV se sada mo\u017ee kompilirati s uklju\u010denim ONNX Runtime. Zbog toga OpenCV koristi vlastiti ONNX parser za izgradnju ORT grafa, \u0161to ne zahtijeva puni paket ONNX. Za kompilaciju su predvi\u0111eni parametri CMake -DWITH_ONNXRUNTIME=ON i -DDOWNLOAD_ONNXRUNTIME=ON; za NVIDIA GPU \u2014 -DDOWNLOAD_ONNXRUNTIME_GPU=ON. Ova funkcija \u0107e biti korisna onima kojima je potreban GPU inferencijski stroj ve\u0107 sada, jer novi DNN motor OpenCV-a trenuta\u010dno radi samo na CPU-u.<\/p>\n<\/li>\n<li>\n<p><strong>Podr\u0161ka za modele vizija-i-jezika.<\/strong><br \/>\nNovi DNN engine uklju\u010duje tokenizatore, slojeve pa\u017enje, blokove dekodiranja, post-procesiranje i KV cache potrebne za pokretanje VLM modela \u00abiz kutije\u00bb. Biblioteka se historijski povezuje s obradom slika, a sada ulazi u podru\u010dje modela koji istodobno rade sa slikama i tekstom.<\/p>\n<\/li>\n<li>\n<p><strong>GPU podr\u0161ka novog DNN motora jo\u0161 nije spremna.<\/strong><br \/>\nNovi engine u OpenCV 5.0 radi samo na CPU. Nativna podr\u0161ka za GPU bit \u0107e dodana u sljede\u0107im izdanjima; za GPU se za sada preporu\u010duje kori\u0161tenje klasi\u010dnog motora ili verzije s ONNX Runtime i NVIDIA execution providers. OpenCV 5.0 sam po sebi jo\u0161 ne zamjenjuje specijalizirane runtimove za server ili desktop CUDA inferencu.<\/p>\n<\/li>\n<li>\n<p><strong>Stari Darknet i Caffe parseri su uklonjeni.<\/strong><br \/>\nRazvija\u010di su uklonili parser-e Darknet i Caffe, budu\u0107i da je ve\u0107ina modela prenesena u ONNX. TFLite jo\u0161 uvijek ostaje kroz klasi\u010dni motor, dok je prijenos na novi motor planiran za kasnije. OpenCV se odri\u010de starih formata i fokusira se na ONNX kao glavni format za razmjenu modela.<\/p>\n<\/li>\n<li>\n<p><strong>Veliko \u010di\u0161\u0107enje zastarjelog API-ja.<\/strong><br \/>\nUklonjen je C API OpenCV 1.x: funkcije poput cvCreateMat() i stare strukture poput CvMat vi\u0161e ne ulaze u glavni API. Tako\u0111er je uklonjena podr\u0161ka za OpenVX, G-API modul je preba\u010den u opencv_contrib, a klasi\u010dni ML modul tako\u0111er. Ovo je prelomna promjena za stare projekte, ali se odavno o\u010dekivala: biblioteka se osloba\u0111a tehni\u010dkog duga nakupljenog desetlje\u0107ima.<\/p>\n<\/li>\n<li>\n<p><strong>Features2D je preimenovan u Features.<\/strong><br \/>\nModul je pro\u0161iren za moderne vektore zna\u010dajki iz neuronskih mre\u017ea. U glavnom repozitoriju ostali su SIFT, ORB, FAST, GoodFeaturesToTrack i MSER, dok su zastarjeli detektori i deskriptori preba\u010deni u opencv_contrib. Dodani su ALIKED, DISK i LightGlue matcher. OpenCV se o\u010digledno prilago\u0111ava modernim pipeline-ovima, gdje se zna\u010dajke slika sve vi\u0161e izvode ne klasi\u010dnim algoritmima ve\u0107 neuronskim mre\u017eama.<\/p>\n<\/li>\n<li>\n<p><strong>Nove vrste podataka u Core.<\/strong><br \/>\nDodani su CV_16BF \/ cv::bfloat, CV_32U, CV_64U, CV_64S i CV_Bool. Tako\u0111er je pobolj\u0161ana podr\u0161ka za N-dimenzionalne nizove, 1D nizove i 0D skalarne vrijednosti. LAPACK je sada uvijek dostupan: ako ne postoji vanjska biblioteka, OpenCV gradi ugra\u0111eni minimalni podskup. Promjene su va\u017ene za DNN, numeri\u010dke algoriteme i kompatibilnost s modernim formatima podataka, gdje se bfloat16 i pro\u0161irene vrste sve vi\u0161e susre\u0107u.<\/p>\n<\/li>\n<li>\n<p><strong>Obrada slika je ubrzana.<\/strong><br \/>\nU imgproc su revidirani warpAffine, warpPerspective i remap. Pobolj\u0161anja performansi su od 10% do vi\u0161e od 300% ovisno o platformi, veli\u010dini slike, vrsti podataka i operativnim oznakama.<br \/>\nOvo vi\u0161e nije o AI modu \u2014 takve optimizacije su izravno va\u017ene za klasi\u010dne zadatke: video streamove, stabilizaciju, projekcijske transformacije i pripremu podataka prije inferencije.<\/p>\n<\/li>\n<li>\n<p><strong>Pobolj\u0161an je izlaz teksta.<\/strong><br \/>\nStari mehanizam renderiranja teksta zamijenjen je STB-based TrueType motorom s ugra\u0111enim varijabilnim fontom Rubik. Mogu\u0107e je povezivanje korisni\u010dkih fontova i uvedena je \u0161iroka podr\u0161ka za Unicode. Me\u0111utim, slo\u017eena pisma kao \u0161to su arapski i devnagari se jo\u0161 uvijek ispravno ne prikazuju bez Harfbuzz, a obojeni emoji nisu podr\u017eani. OpenCV postaje prikladniji za generiranje anotacija, dijagnosti\u010dkih slika i vizualizacija bez vanjskih biblioteka.<\/p>\n<\/li>\n<li>\n<p><strong>A\u017eurirani su 3D i kalibracija kamera.<\/strong><br \/>\ncalib3d je podijeljen na geometry, calib, stereo i ptcloud. USAC je postao standardni backend za stabilnu procjenu homografije, bitne\/fundamentalne matrice, PnP i druge zadatke. Dodan je pipeline za kalibraciju vi\u0161enamjenskih kamera, inicijalna podr\u0161ka za obradu mre\u017ea i oblaka to\u010daka, uklju\u010duju\u0107i TSDF, ICP, uvoz i izvoz .ply i .obj. Ovo je va\u017ean korak za robotiku, SLAM, AR\/VR i vi\u0161enamjenske sustave, gdje je sama 2D obrada odavno neadekvatna.<\/p>\n<\/li>\n<li>\n<p><strong>Revidirani su primjeri i dokumentacija.<\/strong><br \/>\nZastarjeli primjeri su uklonjeni, a preostali su a\u017eurirani. Za DNN primjere koristi se zajedni\u010dka kolekcija modela koja se mo\u017ee preuzeti skriptom download_models.py. Dodani su eksperimentalni primjeri za VLM i LDM, a dokumentacija je dobila modernu temu, brzu pretragu, svijetli\/tamni na\u010din rada i pobolj\u0161anu navigaciju. Za ovako veliku biblioteku ovo nije kozmeti\u010dka promjena: normalna dokumentacija i radni primjeri direktno utje\u010du na to ho\u0107e li se developeri stvarno mo\u0107i prebaciti na OpenCV 5.x.<\/p>\n<\/li>\n<\/ul>\n<p>Na kraju, OpenCV 5.0 nije samo jo\u0161 jedno izdanje s optimizacijama, ve\u0107 zna\u010dajno a\u017euriranje svega: manje naslije\u0111a iz doba OpenCV 1.x i vi\u0161e usmjerenosti prema ONNX-u, neuromre\u017enoj inferenciji, modernim zna\u010dajkama, 3D obradi i heterogenom ubrzanju. Pri tome, prijelaz za stare projekte mo\u017ee zahtijevati ispravke zbog uklanjanja C API-a i premje\u0161tanja dijela modula u opencv_contrib.<\/p>\n<p>Izvor: <a \ncontent=\"nofollow\" rel=\"nofollow\" href=\"https:\/\/www.linux.org.ru\/news\/development\/18313193\">linux.org.ru<\/a><\/p>","protected":false,"gt_translate_keys":[{"key":"rendered","format":"html"}]},"excerpt":{"rendered":"<p>\u041e\u043f\u0443\u0431\u043b\u0438\u043a\u043e\u0432\u0430\u043d OpenCV 5.0.0 \u2014 \u043a\u0440\u0443\u043f\u043d\u044b\u0439 \u0432\u044b\u043f\u0443\u0441\u043a \u043e\u0442\u043a\u0440\u044b\u0442\u043e\u0439 \u0431\u0438\u0431\u043b\u0438\u043e\u0442\u0435\u043a\u0438 \u043a\u043e\u043c\u043f\u044c\u044e\u0442\u0435\u0440\u043d\u043e\u0433\u043e \u0437\u0440\u0435\u043d\u0438\u044f \u0438 \u043e\u0431\u0440\u0430\u0431\u043e\u0442\u043a\u0438 \u0438\u0437\u043e\u0431\u0440\u0430\u0436\u0435\u043d\u0438\u0439. \u0420\u0435\u043b\u0438\u0437 \u0441\u043e\u0441\u0442\u043e\u044f\u043b\u0441\u044f 6 \u0438\u044e\u043d\u044f 2026 \u0433\u043e\u0434\u0430 \u0438 \u0443\u0436\u0435 \u043e\u0442\u043c\u0435\u0447\u0435\u043d \u043a\u0430\u043a \u0430\u043a\u0442\u0443\u0430\u043b\u044c\u043d\u044b\u0439 \u0432 \u0440\u0435\u043f\u043e\u0437\u0438\u0442\u043e\u0440\u0438\u0438 \u043f\u0440\u043e\u0435\u043a\u0442\u0430 \u043d\u0430 GitHub. OpenCV \u0440\u0430\u0441\u043f\u0440\u043e\u0441\u0442\u0440\u0430\u043d\u044f\u0435\u0442\u0441\u044f \u043f\u043e\u0434 \u043b\u0438\u0446\u0435\u043d\u0437\u0438\u0435\u0439 Apache-2.0. \u0421\u0443\u0442\u044c \u0440\u0435\u043b\u0438\u0437\u0430 \u2014 \u043f\u0435\u0440\u0435\u0445\u043e\u0434 OpenCV \u043e\u0442 \u00ab\u043a\u043b\u0430\u0441\u0441\u0438\u0447\u0435\u0441\u043a\u043e\u0439 \u0431\u0438\u0431\u043b\u0438\u043e\u0442\u0435\u043a\u0438 \u043a\u043e\u043c\u043f\u044c\u044e\u0442\u0435\u0440\u043d\u043e\u0433\u043e \u0437\u0440\u0435\u043d\u0438\u044f\u00bb \u043a \u0431\u043e\u043b\u0435\u0435 \u0441\u043e\u0432\u0440\u0435\u043c\u0435\u043d\u043d\u043e\u0439 \u043f\u043b\u0430\u0442\u0444\u043e\u0440\u043c\u0435 \u0434\u043b\u044f CV\/AI-\u0437\u0430\u0434\u0430\u0447: \u043f\u0440\u043e\u0435\u043a\u0442 \u0432\u044b\u0447\u0438\u0441\u0442\u0438\u043b \u0441\u0442\u0430\u0440\u043e\u0435 API, \u043f\u0435\u0440\u0435\u0440\u0430\u0431\u043e\u0442\u0430\u043b [&hellip;]<\/p>\n","protected":false,"gt_translate_keys":[{"key":"rendered","format":"html"}]},"author":8,"featured_media":181945,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[702],"tags":[],"class_list":["post-181944","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-news"],"aioseo_notices":[],"aioseo_head":"\n\t\t<!-- All in One SEO 5.0.4 - aioseo.com -->\n\t<meta name=\"description\" content=\"\u041e\u043f\u0443\u0431\u043b\u0438\u043a\u043e\u0432\u0430\u043d\" \/>\n\t<meta name=\"robots\" content=\"max-image-preview:large\" \/>\n\t<meta name=\"author\" content=\"Erik Peterson\"\/>\n\t<link rel=\"canonical\" href=\"https:\/\/prohoster.info\/hr\/blog\/news\/opencv-5-0-0\" \/>\n\t\t<meta name=\"generator\" content=\"All in One SEO (AIOSEO) 5.0.4\" \/>\n\t\t<meta property=\"og:locale\" content=\"hr_HR\" \/>\n\t\t<meta property=\"og:site_name\" content=\"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\" \/>\n\t\t<meta property=\"og:type\" content=\"article\" \/>\n\t\t<meta property=\"og:title\" content=\"\ud83e\udd47OpenCV 5.0.0 | ProHoster\" \/>\n\t\t<meta property=\"og:description\" content=\"\u041e\u043f\u0443\u0431\u043b\u0438\u043a\u043e\u0432\u0430\u043d\" \/>\n\t\t<meta property=\"og:url\" content=\"https:\/\/prohoster.info\/hr\/blog\/news\/opencv-5-0-0\" \/>\n\t\t<meta property=\"og:image\" content=\"https:\/\/prohoster.info\/wp-content\/uploads\/2021\/11\/logo-350.jpg\" \/>\n\t\t<meta property=\"og:image:secure_url\" content=\"https:\/\/prohoster.info\/wp-content\/uploads\/2021\/11\/logo-350.jpg\" \/>\n\t\t<meta property=\"og:image:width\" content=\"350\" \/>\n\t\t<meta property=\"og:image:height\" content=\"350\" \/>\n\t\t<meta property=\"article:published_time\" content=\"2026-06-08T00:48:46+00:00\" \/>\n\t\t<meta property=\"article:modified_time\" content=\"2026-06-08T00:48:46+00:00\" \/>\n\t\t<meta property=\"article:publisher\" content=\"https:\/\/www.facebook.com\/prohoster\" \/>\n\t\t<meta property=\"article:author\" content=\"https:\/\/www.facebook.com\/prohoster\" \/>\n\t\t<!-- All in One SEO -->\n\n","aioseo_head_json":{"title":"\ud83e\udd47OpenCV 5.0.0 | ProHoster","description":"Objavljeno","canonical_url":"https:\/\/prohoster.info\/hr\/blog\/news\/opencv-5-0-0","robots":"max-image-preview:large","keywords":"","webmasterTools":{"miscellaneous":""},"schema":null,"og:locale":"hr_HR","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\udd47OpenCV 5.0.0 | ProHoster","og:description":"\u041e\u043f\u0443\u0431\u043b\u0438\u043a\u043e\u0432\u0430\u043d","og:url":"https:\/\/prohoster.info\/hr\/blog\/news\/opencv-5-0-0","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":"2026-06-08T00:48:46+00:00","article:modified_time":"2026-06-08T00:48:46+00:00","article:publisher":"https:\/\/www.facebook.com\/prohoster","article:author":"https:\/\/www.facebook.com\/prohoster"},"aioseo_meta_data":[],"gt_translate_keys":[{"key":"link","format":"url"}],"_links":{"self":[{"href":"https:\/\/prohoster.info\/hr\/wp-json\/wp\/v2\/posts\/181944","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/prohoster.info\/hr\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/prohoster.info\/hr\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/prohoster.info\/hr\/wp-json\/wp\/v2\/users\/8"}],"replies":[{"embeddable":true,"href":"https:\/\/prohoster.info\/hr\/wp-json\/wp\/v2\/comments?post=181944"}],"version-history":[{"count":0,"href":"https:\/\/prohoster.info\/hr\/wp-json\/wp\/v2\/posts\/181944\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/prohoster.info\/hr\/wp-json\/wp\/v2\/media\/181945"}],"wp:attachment":[{"href":"https:\/\/prohoster.info\/hr\/wp-json\/wp\/v2\/media?parent=181944"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/prohoster.info\/hr\/wp-json\/wp\/v2\/categories?post=181944"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/prohoster.info\/hr\/wp-json\/wp\/v2\/tags?post=181944"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}