{"id":11618,"date":"2025-05-25T23:15:55","date_gmt":"2025-05-25T21:15:55","guid":{"rendered":"https:\/\/geopard.tech\/?p=11618"},"modified":"2025-05-25T23:15:55","modified_gmt":"2025-05-25T21:15:55","slug":"didelio-tikslumo-dirbtinio-intelekto-modeliai-klasifikuoja-topografinius-zemelapius-greiciau-nei-tradiciniai","status":"publish","type":"post","link":"https:\/\/geopard.tech\/lt\/blog\/high-accuracy-ai-models-classifies-topographic-mapping-faster-than-traditional\/","title":{"rendered":"Didelio tikslumo dirbtinio intelekto modeliai klasifikuoja topografinius \u017eem\u0117lapius grei\u010diau nei tradiciniai"},"content":{"rendered":"

Indonezija, daugiau nei 17 000 sal\u0173, u\u017eiman\u010di\u0173 1,9 milijono kvadratini\u0173 kilometr\u0173 plot\u0105, turinti taut\u0105, susiduria su dideliu i\u0161\u0161\u016bkiu kurdama i\u0161samius \u017eem\u0117lapius, kurie pad\u0117t\u0173 \u012fgyvendinti jos vystymosi tikslus.<\/p>\n

Kadangi didelio mastelio topografiniai \u017eem\u0117lapiai (1:5000 masteliu) apima tik 3% \u0161alies teritorijos, tradiciniai metodai, tokie kaip rankinis stereografinis brai\u017eymas ir lauko tyrimai, yra per l\u0117ti, kad patenkint\u0173 skubius miest\u0173 planavimo, nelaimi\u0173 valdymo ir aplinkos apsaugos poreikius.<\/p>\n

Novatori\u0161kas tyrimas, paskelbtas m.\u00a0Nuotoliniai tyrimai<\/em> 2025 m. si\u016blo sprendim\u0105: gilaus mokymosi sistem\u0105, kuri automatizuoja \u017eem\u0117s dangos klasifikavim\u0105 naudojant labai didel\u0117s skiriamosios gebos palydovinius vaizdus.<\/p>\n

Indonezijos \u017eem\u0117lapi\u0173 sudarymo i\u0161\u0161\u016bkis Topografija<\/strong><\/h2>\n

D\u0117l Indonezijos dyd\u017eio ir sud\u0117tingumo kartografavimas yra mil\u017eini\u0161ka u\u017eduotis. U\u017e nacionalin\u012f kartografavim\u0105 atsakinga Geografin\u0117s informacijos agent\u016bra (BIG) \u0161iuo metu kasmet parengia 13 000 kvadratini\u0173 kilometr\u0173 topografini\u0173 \u017eem\u0117lapi\u0173.<\/p>\n

Tokiu tempu visos \u0161alies kartografavimas u\u017etrukt\u0173 daugiau nei \u0161imtmet\u012f. Net jei ne\u012ftrauktume mi\u0161k\u0173, kurie u\u017eima beveik pus\u0119 Indonezijos teritorijos, likusios teritorijos sudarymas vis tiek u\u017etrukt\u0173 60 met\u0173.<\/p>\n

\u0160i l\u0117ta pa\u017eanga prie\u0161tarauja nacionaliniams prioritetams, pvz.\u00a0Vieno \u017eem\u0117lapio politika<\/em>, pristatyta 2016 m., siekiant standartizuoti \u017eem\u0117lapius skirtinguose sektoriuose ir i\u0161vengti konflikt\u0173 d\u0117l \u017eem\u0117s naudojimo. \u0160ios politikos mastelio keitimas iki 1:5000 \u017eem\u0117lapi\u0173 yra b\u016btinas, ta\u010diau gerokai atsilieka nuo grafiko.<\/p>\n

Topografiniai \u017eem\u0117lapiai<\/strong>\u00a0yra i\u0161sam\u016bs nat\u016brali\u0173 ir \u017emogaus sukurt\u0173 \u017dem\u0117s pavir\u0161iaus darini\u0173, \u012fskaitant auk\u0161t\u012f (kalvas, sl\u0117nius), vandens telkinius, kelius, pastatus ir augmenij\u0105, vaizdai.<\/p>\n

Jie yra pagrindin\u0117s infrastrukt\u016bros planavimo, reagavimo \u012f nelaimes ir aplinkos steb\u0117senos priemon\u0117s. Indonezijoje \u0161i\u0173 \u017eem\u0117lapi\u0173 k\u016brimas masteliu 1:5000 (kai 1 cm \u017eem\u0117lapyje atitinka 50 metr\u0173 ant \u017eem\u0117s) yra labai svarbus tikslumui tokiuose projektuose kaip keli\u0173 tiesimas ar potvyni\u0173 modeliavimas.<\/p>\n

\"Indonezijos<\/p>\n

\u017dem\u0117s dangos duomenys<\/strong>, topografini\u0173 \u017eem\u0117lapi\u0173 pogrupis, rei\u0161kia fizin\u0119 med\u017eiag\u0105 \u017dem\u0117s pavir\u0161iuje, pavyzd\u017eiui, mi\u0161kus, miesto teritorijas ar vanden\u012f. Kitaip nei\u00a0\u017eem\u0117s naudojimas<\/em>\u00a0(kuris apib\u016bdina, kaip \u017emon\u0117s naudoja \u017eem\u0119, pvz., gyvenamosios ar pramonin\u0117s zonos), \u017eem\u0117 sutelkia d\u0117mes\u012f \u012f stebimus objektus.<\/p>\n

Tiksl\u016bs \u017eem\u0117s dangos \u017eem\u0117lapiai padeda vyriausyb\u0117ms steb\u0117ti mi\u0161k\u0173 naikinim\u0105, miest\u0173 pl\u0117tr\u0105 arba \u012fvertinti \u017eem\u0117s \u016bkio produktyvum\u0105. Tradici\u0161kai analitikai \u0161iuos objektus \u017eymi rankiniu b\u016bdu pikseliu po pikselio, naudodami aerofotonuotraukas arba palydovinius vaizdus \u2013 tai procesas, kuris u\u017eima daug laiko ir yra link\u0119s \u012f \u017emogi\u0161k\u0105sias klaidas.<\/p>\n

Pavyzd\u017eiui, keli\u0173 ar ma\u017e\u0173 pastat\u0173 identifikavimas tankiai apgyvendintose miesto teritorijose gali u\u017etrukti kruop\u0161taus darbo kelias dienas. 2025 m. atliktas tyrimas \u0161i\u0105 kli\u016bt\u012f sprend\u017eia rankin\u012f darb\u0105 pakeisdamas dirbtiniu intelektu, konkre\u010diai \u2013 giliuoju mokymusi, siekiant automatizuoti \u017eem\u0117s dangos klasifikavim\u0105.<\/p>\n

Dirbtinio intelekto valdoma palydovini\u0173 vaizd\u0173 analiz\u0117\u00a0<\/strong><\/h2>\n

Tyrimas buvo sutelktas \u012f Mataramo miest\u0105 \u2013 nedidel\u012f, bet \u012fvairiapus\u012f miesto rajon\u0105 Lomboko saloje, kaip bandom\u0105j\u012f atvej\u012f. Komanda naudojo\u00a0Plejad\u0173 palydoviniai vaizdai<\/strong>\u00a0nuo 2015 m., \u012fskaitant didel\u0117s skiriamosios gebos panchromatini\u0173 (0,5 metro) ir multispektrini\u0173 (2 metr\u0173) duomen\u0173.<\/p>\n

Panchromatiniai vaizdai fiksuoja smulkias erdvines detales pilkos spalvos tonu, o daugiaspektriniai vaizdai teikia spalv\u0173 ir infraraudon\u0173j\u0173 spinduli\u0173 informacij\u0105 konkre\u010diuose bangos ilgi\u0173 diapazonuose (pvz., raudona, \u017ealia, m\u0117lyna, artimoji infraraudonoji spinduliuot\u0117).<\/p>\n

Siekdami sujungti \u0161iuos privalumus, tyr\u0117jai pritaik\u0117 technik\u0105, vadinam\u0105 panoraminiu pary\u0161kinimu, kuri sujungia didel\u0117s skiriamosios gebos pilkos spalvos duomenis su ma\u017eesn\u0117s skiriamosios gebos spalvotais vaizdais. \u0160is procesas leido gauti ry\u0161kius, detalius vaizdus, kuri\u0173 skiriamoji geba yra 0,5 metro, idealiai tinkan\u010dius ma\u017eiems objektams, tokiems kaip keliai ar atskiri pastatai, aptikti.<\/p>\n

Panar\u0161inimas yra labai svarbus, nes jis i\u0161saugo i\u0161sami\u0105 daugiaspektrini\u0173 duomen\u0173 spektrin\u0119 informacij\u0105, kartu padidindamas erdvin\u012f ai\u0161kum\u0105 ir u\u017etikrindamas, kad spalvos tiksliai atitikt\u0173 fizines savybes.<\/p>\n

Toliau komanda i\u0161 vaizd\u0173 i\u0161trauk\u0117 papildomos informacijos, kad pagerint\u0173 klasifikavimo tikslum\u0105. Jie apskai\u010diavo normalizuot\u0105 skirtumin\u012f augalijos indeks\u0105 (NDVI) \u2013 augal\u0173 sveikatos mat\u0105, gaut\u0105 i\u0161 artimojo infraraudonojo spektro (NIR) ir raudonos \u0161viesos atspind\u017ei\u0173.<\/p>\n

Sveika augmenija atspindi daugiau artimojo infraraudonojo spinduliavimo ir sugeria daugiau raudonos \u0161viesos d\u0117l chlorofilo aktyvumo. Formul\u0117\u00a0NDVI = (NIR \u2212 raudona) \/ (NIR + raudona)<\/span>NDVI<\/span><\/span>=<\/span><\/span>(<\/span>artimasis infraraudon\u0173j\u0173 spinduli\u0173<\/span><\/span>\u2212<\/span><\/span>Raudona<\/span><\/span>)<\/span>\/<\/span>(<\/span>artimasis infraraudon\u0173j\u0173 spinduli\u0173<\/span><\/span>+<\/span><\/span>Raudona<\/span><\/span>)<\/span><\/span><\/span><\/span>\u00a0gaunamos vert\u0117s nuo -1 iki 1, kur didesn\u0117s vert\u0117s rodo tankesn\u0119, sveikesn\u0119 augmenij\u0105.<\/p>\n

NDVI yra ne\u012fkainojamas norint atskirti mi\u0161kus, dirbam\u0105 \u017eem\u0119 ir miesto \u017eali\u0105sias erdves. Pavyzd\u017eiui, \u0161iame tyrime NDVI pad\u0117jo atskirti ve\u0161lius plantacijas nuo plikos dirvos.<\/p>\n

Tekst\u016bros analiz\u0117<\/strong>\u00a0buvo dar vienas svarbus \u017eingsnis. Naudodami statistin\u012f metod\u0105, vadinam\u0105 pilkojo lygio bendro pasirei\u0161kimo matrica (GLCM), tyr\u0117jai kiekybi\u0161kai \u012fvertino vaizduose matomus modelius, tokius kaip \u017eem\u0117s \u016bkio lauk\u0173 \u0161iurk\u0161tumas, palyginti su asfaltuot\u0173 keli\u0173 lygumu.<\/p>\n

GLCM veikia analizuodamas, kaip da\u017enai paveiksl\u0117lyje pasitaiko pikseli\u0173 poros su konkre\u010diomis reik\u0161m\u0117mis ir erdviniais ry\u0161iais (pvz., horizontaliai gretimi). I\u0161 \u0161ios matricos galima gauti tokius rodiklius kaip\u00a0homogeni\u0161kumas<\/em>\u00a0(pikseli\u0173 ver\u010di\u0173 vienodumas),\u00a0kontrastas<\/em>\u00a0(vietiniai intensyvumo svyravimai) ir\u00a0entropija<\/em>\u00a0(pikseli\u0173 pasiskirstymo atsitiktinumas) yra apskai\u010diuojami.<\/p>\n

\u0160ie tekst\u016bros rodikliai pad\u0117jo dirbtinio intelekto modeliui atskirti pana\u0161iai atrodan\u010dius \u017eem\u0117s dangos tipus, pavyzd\u017eiui, asfaltuotus kelius nuo tamsi\u0173 dirvo\u017eemio plot\u0173.<\/p>\n

Siekdama supaprastinti duomenis, komanda pritaik\u0117\u00a0Pagrindini\u0173 komponent\u0173 analiz\u0117 (PCA)<\/strong>, technika, kuri identifikuoja reik\u0161mingiausius duomen\u0173 rinkinio modelius. PCA suma\u017eina pertekli\u0173, transformuodama koreliuotus kintamuosius (pvz., kelias tekst\u016bros juostas) \u012f ma\u017eesn\u012f nekoreliuot\u0173 komponent\u0173 rinkin\u012f.<\/p>\n

\u0160iame tyrime PCA sujung\u0117 penkias tekst\u016bros juostas \u012f du pagrindinius komponentus, i\u0161laikydama 95% pradin\u0117s informacijos. Tai supaprastino gilaus mokymosi modelio \u012fvest\u012f, pagerindama tiek tikslum\u0105, tiek skai\u010diavimo efektyvum\u0105.<\/p>\n

\u201eU-Net\u201c giluminis mokymasis \u017eem\u0117s dangos srityje <\/strong>Topografija<\/strong><\/h2>\n

Tyrimo esm\u0117 buvo gilaus mokymosi modelis, pagr\u012fstas \u201eU-Net\u201c architekt\u016bra \u2013 konvoliucinio neuroninio tinklo (CNN) tipu, pla\u010diai naudojamu vaizd\u0173 segmentavimo u\u017eduotims.<\/p>\n

Pavadintas d\u0117l U formos dizaino, U-Net sudaro dvi pagrindin\u0117s dalys: kodavimo \u012frenginys, kuris analizuoja vaizd\u0105, kad i\u0161skirt\u0173 hierarchines ypatybes (pvz., kra\u0161tus, tekst\u016bras), ir dekodavimo \u012frenginys, kuris rekonstruoja vaizd\u0105 su pikseli\u0173 \u017eym\u0117mis.<\/p>\n

Kodavimo \u012frenginys naudoja konvoliucinius sluoksnius ir telkim\u0105, kad suma\u017eint\u0173 vaizdo diskretizavim\u0105, u\u017efiksuodamas pla\u010dius modelius, o dekodavimo \u012frenginys padidina duomen\u0173 diskretizavim\u0105, kad atkurt\u0173 erdvin\u0119 skiriam\u0105j\u0105 geb\u0105. Praleid\u017eiant jungtis tarp kodavimo ir dekodavimo sluoksni\u0173, i\u0161saugomos smulkios detal\u0117s, leid\u017eiant tiksliai aptikti ribas \u2013 tai labai svarbi funkcija, sudarant siaur\u0173 gatvi\u0173 ar netaisyklingos formos pastat\u0173 \u017eem\u0117lapius.<\/p>\n

\"\u017dem\u0117s<\/p>\n

Modelyje buvo naudojamas \u201cResNet34\u201d magistral\u0117 \u2013 i\u0161 anksto apmokytas tinklas, \u017einomas d\u0117l savo gylio ir efektyvumo. \u201eResNet34\u201c priklauso likutini\u0173 tinkl\u0173 \u0161eimai, kuri \u012fdiegia \u201etrumpuosius ry\u0161ius\u201c, kad apeit\u0173 sluoksnius ir suma\u017eint\u0173 nykstan\u010dio gradiento problem\u0105 (kai giliems tinklams sunku mokytis d\u0117l ma\u017e\u0117jan\u010di\u0173 atnaujinim\u0173 mokymo metu).<\/p>\n

Pasinaudojant \u2019ResNet34\u201c geb\u0117jimu atpa\u017einti sud\u0117tingus modelius i\u0161 \u201eImageNet\u201c (did\u017eiul\u0117s vaizd\u0173 duomen\u0173 baz\u0117s), modeliui reik\u0117jo ma\u017eiau mokymo duomen\u0173 ir laiko prisitaikyti prie palydovini\u0173 vaizd\u0173.<\/p>\n

Modelio apmokymui reik\u0117jo 1440 vaizdo plyteli\u0173, kuri\u0173 kiekviena buvo 512 \u00d7 512 pikseli\u0173, apiman\u010di\u0173 \u0161e\u0161ias \u017eem\u0117s dangos klases: pastatus, kelius, \u017eem\u0117s \u016bkio paskirties \u017eem\u0119, plik\u0105 \u017eem\u0119, plantacijas ir vandens telkinius.<\/p>\n

Duomen\u0173 rinkinyje buvo b\u016bding\u0173 disbalans\u0173; keliai ir vandens telkiniai sudar\u0117 atitinkamai tik 3,7% ir 4,2% im\u010di\u0173, o pastatai ir \u017eem\u0117s \u016bkio paskirties \u017eem\u0117 \u2013 daugiau nei 25%. Nepaisant \u0161io i\u0161\u0161\u016bkio, modelis buvo apmokytas per 200 epoch\u0173 \u2013 tikslumo ir skai\u010diavimo s\u0105naud\u0173 pusiausvyra \u2013 naudojant 2 partij\u0173 dyd\u012f d\u0117l atminties apribojim\u0173.<\/p>\n

An\u00a0epocha<\/strong>\u00a0rei\u0161kia vien\u0105 visi\u0161k\u0105 mokymo duomen\u0173 perdavim\u0105 per model\u012f, tuo tarpu\u00a0partijos dydis<\/strong>\u00a0nustato, kiek pavyzd\u017ei\u0173 apdorojama prie\u0161 atnaujinant modelio parametrus. Ma\u017eesni paket\u0173 dyd\u017eiai suma\u017eina atminties naudojim\u0105, bet gali sul\u0117tinti mokym\u0105.<\/p>\n

\u017dem\u0117lapi\u0173 tobulinimas morfologiniu apdorojimu<\/strong><\/h2>\n

Net ir geriausi dirbtinio intelekto modeliai sukuria klaid\u0173, pavyzd\u017eiui, neteisingai klasifikuoja izoliuotus pikselius arba sukuria nelygius kra\u0161tus aplink elementus. Nor\u0117dami tai i\u0161spr\u0119sti, tyr\u0117jai pritaik\u0117 morfologin\u012f apdorojim\u0105 \u2013 technik\u0105, kuri i\u0161lygina tr\u016bkumus, naudodama tokias operacijas kaip erozija ir i\u0161pl\u0117timas.<\/p>\n

Erozija pa\u0161alina plonus pikseli\u0173 sluoksnius nuo objekt\u0173 rib\u0173, panaikindama ma\u017ey\u010dius neteisingai klasifikuotus plotus, o i\u0161pl\u0117timas prideda pikseli\u0173, kad i\u0161pl\u0117st\u0173 objekt\u0173 ribas ir u\u017epildyt\u0173 tarpus linijiniuose elementuose, tokiuose kaip keliai.<\/p>\n

\u0160ios operacijos remiasi strukt\u016brizuojan\u010diu elementu (ma\u017ea matrica), kuris slenka per vaizd\u0105 ir kei\u010dia pikseli\u0173 vertes. Optimalus branduolio dydis \u0161ioms operacijoms (5 \u00d7 5 pikseliai) buvo nustatytas naudojant pusiau dispersin\u0119 analiz\u0119 \u2013 geostatistin\u012f metod\u0105, kuris kiekybi\u0161kai \u012fvertino erdvinius vaizduose matomus modelius.<\/p>\n

Pusdispersija matuoja, kiek pikseli\u0173 vert\u0117s skiriasi skirtingais atstumais, ir padeda nustatyti mastel\u012f, kuriame tekst\u016bros elementai (pvz., pastat\u0173 sankaupos) yra ry\u0161kiausi.<\/p>\n

Dirbtinis intelektas padidina \u017eem\u0117lapi\u0173 sudarymo greit\u012f ir tikslum\u0105<\/strong><\/h2>\n

Modelis pasiek\u0117 pradin\u012f 84% tikslum\u0105 (kappa balas<\/strong>\u00a0= 0,79), kuris po papildomo apdorojimo padid\u0117jo iki 86% (kappa = 0,81).\u00a0kappa balas<\/strong>\u00a0(Koeno kapa) matuoja numatytos ir faktin\u0117s klasifikacijos atitikim\u0105, pakoreguojant pagal atsitiktinum\u0105.<\/p>\n

0,81 balo rodiklis rodo \u201cbeveik visi\u0161k\u0105\u201d atitikim\u0105, vir\u0161ijant\u012f 0,61\u20130,80 diapazon\u0105, kuris laikomas \u201cdideliu\u201d. Vandens telkiniai ir plantacijos buvo klasifikuoti beveik idealiu tikslumu (atitinkamai 97% ir 96%), o keliai, kuriems k\u0117l\u0117 problem\u0173 d\u0117l plonos, linijin\u0117s formos ir \u0161e\u0161\u0117li\u0173, pasiek\u0117 85%.<\/p>\n

\"Dirbtinis<\/p>\n

Pastatai ir \u017eem\u0117s \u016bkio paskirties \u017eem\u0117 taip pat pasirod\u0117 gerai \u2013 j\u0173 F1 balai buvo 88% ir 83%. F1 balas, harmoninis tikslumo ir atk\u016brimo vidurkis, subalansuoja klaidingai teigiamus ir klaidingai neigiamus rezultatus, tod\u0117l idealiai tinka nesubalansuot\u0173 duomen\u0173 rinkini\u0173 vertinimui.<\/p>\n

Efektyvumo padid\u0117jimas buvo dar ry\u0161kesnis. Tradicinis stereografinis brai\u017eymas, kai elementai trima\u010diuose aerofotonuotraukose \u017eymimi rankiniu b\u016bdu, pastatams ir augmenijai su\u017eym\u0117ti trunka devynias dienas vienam \u017eem\u0117lapio lapui (5,29 km\u00b2).<\/p>\n

Dirbtiniu intelektu pagr\u012fstas metodas suma\u017eino \u0161\u012f laik\u0105 iki 43 minu\u010di\u0173 vienam lapui \u2013 250 kart\u0173 geresnis. I\u0161 prad\u017ei\u0173 modelio apmokymas u\u017etruko 17 valand\u0173, ta\u010diau apmokytas jis gal\u0117t\u0173 klasifikuoti did\u017eiulius plotus su minimaliu \u017emogaus \u012fsiki\u0161imu. \u0160ios sistemos mastas leist\u0173 Indonezijai kasmet sudaryti 9000 km\u00b2 \u017eem\u0117lap\u012f, sutrumpinant numatom\u0105 u\u017ebaigimo laik\u0105 nuo daugiau nei \u0161imtme\u010dio iki vos 15 met\u0173.<\/p>\n

Dirbtinio intelekto \u017eem\u0117lapi\u0173 sudarymas skatina pasaulin\u012f tvarum\u0105<\/strong><\/h2>\n

Pasekm\u0117s apima daug daugiau nei Indonezij\u0105. Automatizuotas \u017eem\u0117s dangos klasifikavimas padeda siekti pasaulini\u0173 pastang\u0173, toki\u0173 kaip JT darnaus vystymosi tikslai (DVT). Pavyzd\u017eiui, mi\u0161k\u0173 naikinimo (DVT 15) arba miest\u0173 pl\u0117tros (DVT 11) steb\u0117jimas tampa greitesnis ir tikslesnis.<\/p>\n

Nelaimi\u0173 paveiktuose regionuose, pavyzd\u017eiui, potvyni\u0173 paveiktose teritorijose, atnaujinami \u017eem\u0117lapiai gali pad\u0117ti nustatyti pa\u017eeid\u017eiamas bendruomenes ir suplanuoti evakuacijos mar\u0161rutus.<\/p>\n

\u016akininkai taip pat gauna naudos; tiksl\u016bs \u017eem\u0117s dangos duomenys leid\u017eia taikyti tiksli\u0105j\u0105 \u017eemdirbyst\u0119, optimizuoti vandens naudojim\u0105 ir pas\u0117li\u0173 derli\u0173, stebint dirvo\u017eemio sveikat\u0105 ir augmenijos stres\u0105 naudojant NDVI.<\/p>\n

Ta\u010diau i\u0161\u0161\u016bki\u0173 vis dar i\u0161lieka. Modelio veikimas nepakankamai atstovaujamose klas\u0117se, tokiose kaip keliai, pabr\u0117\u017eia subalansuot\u0173 mokymo duomen\u0173 poreik\u012f. B\u016bsimuose darbuose b\u016bt\u0173 galima \u012ftraukti perk\u0117limo mokym\u0105si \u2013 technik\u0105, kai modelis, i\u0161 anksto apmokytas vienai u\u017eduo\u010diai (pvz., bendram vaizd\u0173 atpa\u017einimui), yra tiksliai pritaikomas konkre\u010diai programai (pvz., keli\u0173 aptikimui palydoviniuose vaizduose).<\/p>\n

Tai suma\u017eina dideli\u0173, brangiai kainuojan\u010di\u0173, pa\u017eenklint\u0173 duomen\u0173 rinkini\u0173 poreik\u012f. Tikslum\u0105 b\u016bt\u0173 galima dar labiau padidinti testuojant pa\u017eangias architekt\u016bras, tokias kaip \u201eU-Net3+\u201c, kuri pagerina objekt\u0173 agregavim\u0105 \u012fvairiais masteliais, arba transformatoriais pagr\u012fstus modelius (kurie puikiai fiksuoja tolimojo nuotolio priklausomybes vaizduose).<\/p>\n

Ta\u010diau Lidaro (\u0161viesos aptikimo ir diapazono matavimo) arba radaro duomen\u0173 integravimas taip pat gal\u0117t\u0173 pagerinti rezultatus, ypa\u010d debesuotuose regionuose, kur optiniai palydovai sunkiai veikia.<\/p>\n

I\u0161vada: nauja geoprini\u0173 moksl\u0173 era<\/h2>\n

\u0160is tyrimas \u017eymi l\u016b\u017eio ta\u0161k\u0105 topografinio kartografavimo srityje. Automatizuodamos \u017eem\u0117s dangos klasifikavim\u0105, \u0161alys gali grei\u010diau ir pigiau nei bet kada anks\u010diau parengti tikslius \u017eem\u0117lapius. Indonezijai \u0161i technologija yra ne tik patogumas \u2013 tai b\u016btinyb\u0117 norint valdyti spar\u010di\u0105 urbanizacij\u0105, apsaugoti mi\u0161kus ir pasiruo\u0161ti su klimatu susijusioms nelaim\u0117ms.<\/p>\n

Tobul\u0117jant dirbtiniam intelektui ir palydovin\u0117ms technologijoms, didel\u0117s skiriamosios gebos realaus laiko \u017eem\u0117lapi\u0173 sudarymo vizija tampa ranka pasiekiama, suteikiant vyriausyb\u0117ms ir bendruomen\u0117ms gali\u0173 kurti tvaresn\u0119 ateit\u012f.<\/p>\n

Nuoroda<\/strong>Hakim, YF; Tsai, F. Giliuoju mokymusi pagr\u012fstas \u017eem\u0117s dangos i\u0161skyrimas i\u0161 labai didel\u0117s skiriamosios gebos palydovini\u0173 vaizd\u0173, siekiant palengvinti didelio masto topografini\u0173 \u017eem\u0117lapi\u0173 k\u016brim\u0105. Remote Sens. 2025, 17, 473. https:\/\/doi.org\/10.3390\/rs17030473<\/a><\/p>","protected":false},"excerpt":{"rendered":"

Indonezija, daugiau nei 17 000 sal\u0173, u\u017eiman\u010di\u0173 1,9 milijono kvadratini\u0173 kilometr\u0173 plot\u0105, turinti tauta, susiduria su dideliu i\u0161\u0161\u016bkiu kurdama i\u0161samius \u017eem\u0117lapius, kurie pad\u0117t\u0173 \u012fgyvendinti jos vystymosi tikslus...<\/p>","protected":false},"author":210249433,"featured_media":11626,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"_coblocks_attr":"","_coblocks_dimensions":"","_coblocks_responsive_height":"","_coblocks_accordion_ie_support":"","_eb_attr":"","_crdt_document":"","content-type":"","jetpack_post_was_ever_published":false,"_jetpack_newsletter_access":"","_jetpack_dont_email_post_to_subs":false,"_jetpack_newsletter_tier_id":0,"_jetpack_memberships_contains_paywalled_content":false,"_jetpack_memberships_contains_paid_content":false,"footnotes":"","jetpack_publicize_message":"","jetpack_publicize_feature_enabled":true,"jetpack_social_post_already_shared":true,"jetpack_social_options":{"image_generator_settings":{"template":"highway","default_image_id":0,"font":"","enabled":false},"version":2},"_wpas_customize_per_network":false},"categories":[1661,1366],"tags":[],"class_list":["post-11618","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-satellite-imagery","category-topography"],"acf":[],"yoast_head":"\nHigh-Accuracy AI Models Classifies Topographic Mapping Faster Than Traditional - GeoPard Agriculture<\/title>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/geopard.tech\/lt\/tinklarastis\/didelio-tikslumo-dirbtinio-intelekto-modeliai-klasifikuoja-topografinius-zemelapius-greiciau-nei-tradiciniai\/\" \/>\n<meta property=\"og:locale\" content=\"lt_LT\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"High-Accuracy AI Models Classifies Topographic Mapping Faster Than Traditional\" \/>\n<meta property=\"og:description\" content=\"Indonesia, a nation of over 17,000 islands spanning 1.9 million square kilometers, faces a critical challenge in creating detailed maps to support its development goals....\" \/>\n<meta property=\"og:url\" content=\"https:\/\/geopard.tech\/lt\/tinklarastis\/didelio-tikslumo-dirbtinio-intelekto-modeliai-klasifikuoja-topografinius-zemelapius-greiciau-nei-tradiciniai\/\" \/>\n<meta property=\"og:site_name\" content=\"GeoPard - Precision agriculture Mapping software\" \/>\n<meta property=\"article:publisher\" content=\"https:\/\/www.facebook.com\/geopardAgriculture\/\" \/>\n<meta property=\"article:published_time\" content=\"2025-05-25T21:15:55+00:00\" \/>\n<meta property=\"og:image\" content=\"https:\/\/geopard.tech\/wp-content\/uploads\/2025\/05\/High-Accuracy-AI-Models-Classifies-Topographic-Mapping-Faster-Than-Traditional-1024x576.png\" \/>\n\t<meta property=\"og:image:width\" content=\"1024\" \/>\n\t<meta property=\"og:image:height\" content=\"576\" \/>\n\t<meta property=\"og:image:type\" content=\"image\/png\" \/>\n<meta name=\"author\" content=\"Muhammad Farjad\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:creator\" content=\"@geopardagri\" \/>\n<meta name=\"twitter:site\" content=\"@geopardagri\" \/>\n<meta name=\"twitter:label1\" content=\"Written by\" \/>\n\t<meta name=\"twitter:data1\" content=\"Muhammad Farjad\" \/>\n\t<meta name=\"twitter:label2\" content=\"Est. reading time\" \/>\n\t<meta name=\"twitter:data2\" content=\"9 minut\u0117s\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\\\/\\\/schema.org\",\"@graph\":[{\"@type\":\"Article\",\"@id\":\"https:\\\/\\\/geopard.tech\\\/blog\\\/high-accuracy-ai-models-classifies-topographic-mapping-faster-than-traditional\\\/#article\",\"isPartOf\":{\"@id\":\"https:\\\/\\\/geopard.tech\\\/blog\\\/high-accuracy-ai-models-classifies-topographic-mapping-faster-than-traditional\\\/\"},\"author\":{\"name\":\"Muhammad Farjad\",\"@id\":\"https:\\\/\\\/geopard.tech\\\/#\\\/schema\\\/person\\\/123c5562fb47aa8cf3aa81ae91e5e935\"},\"headline\":\"High-Accuracy AI Models Classifies Topographic Mapping Faster Than Traditional\",\"datePublished\":\"2025-05-25T21:15:55+00:00\",\"mainEntityOfPage\":{\"@id\":\"https:\\\/\\\/geopard.tech\\\/blog\\\/high-accuracy-ai-models-classifies-topographic-mapping-faster-than-traditional\\\/\"},\"wordCount\":1728,\"commentCount\":0,\"publisher\":{\"@id\":\"https:\\\/\\\/geopard.tech\\\/#organization\"},\"image\":{\"@id\":\"https:\\\/\\\/geopard.tech\\\/blog\\\/high-accuracy-ai-models-classifies-topographic-mapping-faster-than-traditional\\\/#primaryimage\"},\"thumbnailUrl\":\"https:\\\/\\\/i0.wp.com\\\/geopard.tech\\\/wp-content\\\/uploads\\\/2025\\\/05\\\/High-Accuracy-AI-Models-Classifies-Topographic-Mapping-Faster-Than-Traditional.png?fit=1920%2C1080&ssl=1\",\"articleSection\":[\"Satellite Imagery\",\"Topography\"],\"inLanguage\":\"lt-LT\",\"potentialAction\":[{\"@type\":\"CommentAction\",\"name\":\"Comment\",\"target\":[\"https:\\\/\\\/geopard.tech\\\/blog\\\/high-accuracy-ai-models-classifies-topographic-mapping-faster-than-traditional\\\/#respond\"]}]},{\"@type\":\"WebPage\",\"@id\":\"https:\\\/\\\/geopard.tech\\\/blog\\\/high-accuracy-ai-models-classifies-topographic-mapping-faster-than-traditional\\\/\",\"url\":\"https:\\\/\\\/geopard.tech\\\/blog\\\/high-accuracy-ai-models-classifies-topographic-mapping-faster-than-traditional\\\/\",\"name\":\"High-Accuracy AI Models Classifies Topographic Mapping Faster Than Traditional - GeoPard Agriculture\",\"isPartOf\":{\"@id\":\"https:\\\/\\\/geopard.tech\\\/#website\"},\"primaryImageOfPage\":{\"@id\":\"https:\\\/\\\/geopard.tech\\\/blog\\\/high-accuracy-ai-models-classifies-topographic-mapping-faster-than-traditional\\\/#primaryimage\"},\"image\":{\"@id\":\"https:\\\/\\\/geopard.tech\\\/blog\\\/high-accuracy-ai-models-classifies-topographic-mapping-faster-than-traditional\\\/#primaryimage\"},\"thumbnailUrl\":\"https:\\\/\\\/i0.wp.com\\\/geopard.tech\\\/wp-content\\\/uploads\\\/2025\\\/05\\\/High-Accuracy-AI-Models-Classifies-Topographic-Mapping-Faster-Than-Traditional.png?fit=1920%2C1080&ssl=1\",\"datePublished\":\"2025-05-25T21:15:55+00:00\",\"breadcrumb\":{\"@id\":\"https:\\\/\\\/geopard.tech\\\/blog\\\/high-accuracy-ai-models-classifies-topographic-mapping-faster-than-traditional\\\/#breadcrumb\"},\"inLanguage\":\"lt-LT\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\\\/\\\/geopard.tech\\\/blog\\\/high-accuracy-ai-models-classifies-topographic-mapping-faster-than-traditional\\\/\"]}]},{\"@type\":\"ImageObject\",\"inLanguage\":\"lt-LT\",\"@id\":\"https:\\\/\\\/geopard.tech\\\/blog\\\/high-accuracy-ai-models-classifies-topographic-mapping-faster-than-traditional\\\/#primaryimage\",\"url\":\"https:\\\/\\\/i0.wp.com\\\/geopard.tech\\\/wp-content\\\/uploads\\\/2025\\\/05\\\/High-Accuracy-AI-Models-Classifies-Topographic-Mapping-Faster-Than-Traditional.png?fit=1920%2C1080&ssl=1\",\"contentUrl\":\"https:\\\/\\\/i0.wp.com\\\/geopard.tech\\\/wp-content\\\/uploads\\\/2025\\\/05\\\/High-Accuracy-AI-Models-Classifies-Topographic-Mapping-Faster-Than-Traditional.png?fit=1920%2C1080&ssl=1\",\"width\":1920,\"height\":1080,\"caption\":\"High-Accuracy AI Models Classifies Topographic Mapping Faster Than Traditional\"},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\\\/\\\/geopard.tech\\\/blog\\\/high-accuracy-ai-models-classifies-topographic-mapping-faster-than-traditional\\\/#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"Home\",\"item\":\"https:\\\/\\\/geopard.tech\\\/\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"High-Accuracy AI Models Classifies Topographic Mapping Faster Than Traditional\"}]},{\"@type\":\"WebSite\",\"@id\":\"https:\\\/\\\/geopard.tech\\\/#website\",\"url\":\"https:\\\/\\\/geopard.tech\\\/\",\"name\":\"GeoPard - Precision agriculture software\",\"description\":\"Precision agriculture Mapping software\",\"publisher\":{\"@id\":\"https:\\\/\\\/geopard.tech\\\/#organization\"},\"alternateName\":\"GeoPard\",\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\\\/\\\/geopard.tech\\\/?s={search_term_string}\"},\"query-input\":{\"@type\":\"PropertyValueSpecification\",\"valueRequired\":true,\"valueName\":\"search_term_string\"}}],\"inLanguage\":\"lt-LT\"},{\"@type\":\"Organization\",\"@id\":\"https:\\\/\\\/geopard.tech\\\/#organization\",\"name\":\"GeoPard Agriculture\",\"alternateName\":\"GeoPard\",\"url\":\"https:\\\/\\\/geopard.tech\\\/\",\"logo\":{\"@type\":\"ImageObject\",\"inLanguage\":\"lt-LT\",\"@id\":\"https:\\\/\\\/geopard.tech\\\/#\\\/schema\\\/logo\\\/image\\\/\",\"url\":\"https:\\\/\\\/i0.wp.com\\\/geopard.tech\\\/wp-content\\\/uploads\\\/2022\\\/03\\\/favicon.png?fit=200%2C200&ssl=1\",\"contentUrl\":\"https:\\\/\\\/i0.wp.com\\\/geopard.tech\\\/wp-content\\\/uploads\\\/2022\\\/03\\\/favicon.png?fit=200%2C200&ssl=1\",\"width\":200,\"height\":200,\"caption\":\"GeoPard Agriculture\"},\"image\":{\"@id\":\"https:\\\/\\\/geopard.tech\\\/#\\\/schema\\\/logo\\\/image\\\/\"},\"sameAs\":[\"https:\\\/\\\/www.facebook.com\\\/geopardAgriculture\\\/\",\"https:\\\/\\\/x.com\\\/geopardagri\",\"https:\\\/\\\/www.linkedin.com\\\/company\\\/geopard-agriculture\\\/\",\"https:\\\/\\\/www.instagram.com\\\/geopardagriculture\\\/\"]},{\"@type\":\"Person\",\"@id\":\"https:\\\/\\\/geopard.tech\\\/#\\\/schema\\\/person\\\/123c5562fb47aa8cf3aa81ae91e5e935\",\"name\":\"Muhammad Farjad\",\"image\":{\"@type\":\"ImageObject\",\"inLanguage\":\"lt-LT\",\"@id\":\"https:\\\/\\\/secure.gravatar.com\\\/avatar\\\/51b9af87d9b0801ae1c0ab294a0f06eb669f60d5168f33394aeeca9de86537bb?s=96&d=identicon&r=g\",\"url\":\"https:\\\/\\\/secure.gravatar.com\\\/avatar\\\/51b9af87d9b0801ae1c0ab294a0f06eb669f60d5168f33394aeeca9de86537bb?s=96&d=identicon&r=g\",\"contentUrl\":\"https:\\\/\\\/secure.gravatar.com\\\/avatar\\\/51b9af87d9b0801ae1c0ab294a0f06eb669f60d5168f33394aeeca9de86537bb?s=96&d=identicon&r=g\",\"caption\":\"Muhammad Farjad\"},\"url\":\"#\"}]}<\/script>\n<!-- \/ Yoast SEO Premium plugin. -->","yoast_head_json":{"title":"Didelio tikslumo dirbtinio intelekto modeliai klasifikuoja topografinius \u017eem\u0117lapius grei\u010diau nei tradiciniai \u2013 \u201eGeoPard Agriculture\u201c","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/geopard.tech\/lt\/tinklarastis\/didelio-tikslumo-dirbtinio-intelekto-modeliai-klasifikuoja-topografinius-zemelapius-greiciau-nei-tradiciniai\/","og_locale":"lt_LT","og_type":"article","og_title":"High-Accuracy AI Models Classifies Topographic Mapping Faster Than Traditional","og_description":"Indonesia, a nation of over 17,000 islands spanning 1.9 million square kilometers, faces a critical challenge in creating detailed maps to support its development goals....","og_url":"https:\/\/geopard.tech\/lt\/tinklarastis\/didelio-tikslumo-dirbtinio-intelekto-modeliai-klasifikuoja-topografinius-zemelapius-greiciau-nei-tradiciniai\/","og_site_name":"GeoPard - Precision agriculture Mapping software","article_publisher":"https:\/\/www.facebook.com\/geopardAgriculture\/","article_published_time":"2025-05-25T21:15:55+00:00","og_image":[{"width":1024,"height":576,"url":"https:\/\/geopard.tech\/wp-content\/uploads\/2025\/05\/High-Accuracy-AI-Models-Classifies-Topographic-Mapping-Faster-Than-Traditional-1024x576.png","type":"image\/png"}],"author":"Muhammad Farjad","twitter_card":"summary_large_image","twitter_creator":"@geopardagri","twitter_site":"@geopardagri","twitter_misc":{"Written by":"Muhammad Farjad","Est. reading time":"9 minut\u0117s"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"Article","@id":"https:\/\/geopard.tech\/blog\/high-accuracy-ai-models-classifies-topographic-mapping-faster-than-traditional\/#article","isPartOf":{"@id":"https:\/\/geopard.tech\/blog\/high-accuracy-ai-models-classifies-topographic-mapping-faster-than-traditional\/"},"author":{"name":"Muhammad Farjad","@id":"https:\/\/geopard.tech\/#\/schema\/person\/123c5562fb47aa8cf3aa81ae91e5e935"},"headline":"High-Accuracy AI Models Classifies Topographic Mapping Faster Than Traditional","datePublished":"2025-05-25T21:15:55+00:00","mainEntityOfPage":{"@id":"https:\/\/geopard.tech\/blog\/high-accuracy-ai-models-classifies-topographic-mapping-faster-than-traditional\/"},"wordCount":1728,"commentCount":0,"publisher":{"@id":"https:\/\/geopard.tech\/#organization"},"image":{"@id":"https:\/\/geopard.tech\/blog\/high-accuracy-ai-models-classifies-topographic-mapping-faster-than-traditional\/#primaryimage"},"thumbnailUrl":"https:\/\/i0.wp.com\/geopard.tech\/wp-content\/uploads\/2025\/05\/High-Accuracy-AI-Models-Classifies-Topographic-Mapping-Faster-Than-Traditional.png?fit=1920%2C1080&ssl=1","articleSection":["Satellite Imagery","Topography"],"inLanguage":"lt-LT","potentialAction":[{"@type":"CommentAction","name":"Comment","target":["https:\/\/geopard.tech\/blog\/high-accuracy-ai-models-classifies-topographic-mapping-faster-than-traditional\/#respond"]}]},{"@type":"WebPage","@id":"https:\/\/geopard.tech\/blog\/high-accuracy-ai-models-classifies-topographic-mapping-faster-than-traditional\/","url":"https:\/\/geopard.tech\/blog\/high-accuracy-ai-models-classifies-topographic-mapping-faster-than-traditional\/","name":"Didelio tikslumo dirbtinio intelekto modeliai klasifikuoja topografinius \u017eem\u0117lapius grei\u010diau nei tradiciniai \u2013 \u201eGeoPard Agriculture\u201c","isPartOf":{"@id":"https:\/\/geopard.tech\/#website"},"primaryImageOfPage":{"@id":"https:\/\/geopard.tech\/blog\/high-accuracy-ai-models-classifies-topographic-mapping-faster-than-traditional\/#primaryimage"},"image":{"@id":"https:\/\/geopard.tech\/blog\/high-accuracy-ai-models-classifies-topographic-mapping-faster-than-traditional\/#primaryimage"},"thumbnailUrl":"https:\/\/i0.wp.com\/geopard.tech\/wp-content\/uploads\/2025\/05\/High-Accuracy-AI-Models-Classifies-Topographic-Mapping-Faster-Than-Traditional.png?fit=1920%2C1080&ssl=1","datePublished":"2025-05-25T21:15:55+00:00","breadcrumb":{"@id":"https:\/\/geopard.tech\/blog\/high-accuracy-ai-models-classifies-topographic-mapping-faster-than-traditional\/#breadcrumb"},"inLanguage":"lt-LT","potentialAction":[{"@type":"ReadAction","target":["https:\/\/geopard.tech\/blog\/high-accuracy-ai-models-classifies-topographic-mapping-faster-than-traditional\/"]}]},{"@type":"ImageObject","inLanguage":"lt-LT","@id":"https:\/\/geopard.tech\/blog\/high-accuracy-ai-models-classifies-topographic-mapping-faster-than-traditional\/#primaryimage","url":"https:\/\/i0.wp.com\/geopard.tech\/wp-content\/uploads\/2025\/05\/High-Accuracy-AI-Models-Classifies-Topographic-Mapping-Faster-Than-Traditional.png?fit=1920%2C1080&ssl=1","contentUrl":"https:\/\/i0.wp.com\/geopard.tech\/wp-content\/uploads\/2025\/05\/High-Accuracy-AI-Models-Classifies-Topographic-Mapping-Faster-Than-Traditional.png?fit=1920%2C1080&ssl=1","width":1920,"height":1080,"caption":"High-Accuracy AI Models Classifies Topographic Mapping Faster Than Traditional"},{"@type":"BreadcrumbList","@id":"https:\/\/geopard.tech\/blog\/high-accuracy-ai-models-classifies-topographic-mapping-faster-than-traditional\/#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https:\/\/geopard.tech\/"},{"@type":"ListItem","position":2,"name":"High-Accuracy AI Models Classifies Topographic Mapping Faster Than Traditional"}]},{"@type":"WebSite","@id":"https:\/\/geopard.tech\/#website","url":"https:\/\/geopard.tech\/","name":"GeoPard \u2013 tiksliosios \u017eemdirbyst\u0117s programin\u0117 \u012franga","description":"Tiksliosios \u017eemdirbyst\u0117s kartografavimo programin\u0117 \u012franga","publisher":{"@id":"https:\/\/geopard.tech\/#organization"},"alternateName":"GeoPard","potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/geopard.tech\/?s={search_term_string}"},"query-input":{"@type":"PropertyValueSpecification","valueRequired":true,"valueName":"search_term_string"}}],"inLanguage":"lt-LT"},{"@type":"Organization","@id":"https:\/\/geopard.tech\/#organization","name":"GeoPard \u017dem\u0117s \u016bkis","alternateName":"GeoPard","url":"https:\/\/geopard.tech\/","logo":{"@type":"ImageObject","inLanguage":"lt-LT","@id":"https:\/\/geopard.tech\/#\/schema\/logo\/image\/","url":"https:\/\/i0.wp.com\/geopard.tech\/wp-content\/uploads\/2022\/03\/favicon.png?fit=200%2C200&ssl=1","contentUrl":"https:\/\/i0.wp.com\/geopard.tech\/wp-content\/uploads\/2022\/03\/favicon.png?fit=200%2C200&ssl=1","width":200,"height":200,"caption":"GeoPard Agriculture"},"image":{"@id":"https:\/\/geopard.tech\/#\/schema\/logo\/image\/"},"sameAs":["https:\/\/www.facebook.com\/geopardAgriculture\/","https:\/\/x.com\/geopardagri","https:\/\/www.linkedin.com\/company\/geopard-agriculture\/","https:\/\/www.instagram.com\/geopardagriculture\/"]},{"@type":"Person","@id":"https:\/\/geopard.tech\/#\/schema\/person\/123c5562fb47aa8cf3aa81ae91e5e935","name":"Muhamadas Farjadas","image":{"@type":"ImageObject","inLanguage":"lt-LT","@id":"https:\/\/secure.gravatar.com\/avatar\/51b9af87d9b0801ae1c0ab294a0f06eb669f60d5168f33394aeeca9de86537bb?s=96&d=identicon&r=g","url":"https:\/\/secure.gravatar.com\/avatar\/51b9af87d9b0801ae1c0ab294a0f06eb669f60d5168f33394aeeca9de86537bb?s=96&d=identicon&r=g","contentUrl":"https:\/\/secure.gravatar.com\/avatar\/51b9af87d9b0801ae1c0ab294a0f06eb669f60d5168f33394aeeca9de86537bb?s=96&d=identicon&r=g","caption":"Muhammad Farjad"},"url":"#"}]}},"jetpack_publicize_connections":[],"jetpack_featured_media_url":"https:\/\/i0.wp.com\/geopard.tech\/wp-content\/uploads\/2025\/05\/High-Accuracy-AI-Models-Classifies-Topographic-Mapping-Faster-Than-Traditional.png?fit=1920%2C1080&ssl=1","jetpack_likes_enabled":true,"jetpack_sharing_enabled":true,"jetpack_shortlink":"https:\/\/wp.me\/pdiCPa-31o","_links":{"self":[{"href":"https:\/\/geopard.tech\/lt\/wp-json\/wp\/v2\/posts\/11618","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/geopard.tech\/lt\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/geopard.tech\/lt\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/geopard.tech\/lt\/wp-json\/wp\/v2\/users\/210249433"}],"replies":[{"embeddable":true,"href":"https:\/\/geopard.tech\/lt\/wp-json\/wp\/v2\/comments?post=11618"}],"version-history":[{"count":0,"href":"https:\/\/geopard.tech\/lt\/wp-json\/wp\/v2\/posts\/11618\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/geopard.tech\/lt\/wp-json\/wp\/v2\/media\/11626"}],"wp:attachment":[{"href":"https:\/\/geopard.tech\/lt\/wp-json\/wp\/v2\/media?parent=11618"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/geopard.tech\/lt\/wp-json\/wp\/v2\/categories?post=11618"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/geopard.tech\/lt\/wp-json\/wp\/v2\/tags?post=11618"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}