{"id":11549,"date":"2025-05-04T21:00:25","date_gmt":"2025-05-04T19:00:25","guid":{"rendered":"https:\/\/geopard.tech\/?p=11549"},"modified":"2025-05-04T21:00:25","modified_gmt":"2025-05-04T19:00:25","slug":"cmtnet-no-jauna-define-precizo-lauksaimniecibu-parspejot-tradicionalo-kulturaugu-klasifikaciju","status":"publish","type":"post","link":"https:\/\/geopard.tech\/lv\/blog\/cmtnet-redefines-precision-agriculture-by-outperforming-traditional-crop-classification\/","title":{"rendered":"CMTNet no jauna defin\u0113 prec\u012bzo lauksaimniec\u012bbu, p\u0101rsp\u0113jot tradicion\u0101lo kult\u016braugu klasifik\u0101ciju"},"content":{"rendered":"

Prec\u012bza kult\u016braugu klasifik\u0101cija ir b\u016btiska m\u016bsdienu prec\u012bzajai lauksaimniec\u012bbai, kas \u013cauj lauksaimniekiem uzraudz\u012bt kult\u016braugu vesel\u012bbu, prognoz\u0113t ra\u017eu un efekt\u012bvi sadal\u012bt resursus. Tom\u0113r tradicion\u0101l\u0101s metodes bie\u017ei vien c\u012bn\u0101s ar lauksaimniec\u012bbas vides sare\u017e\u0123\u012bt\u012bbu, kur kult\u016braugi iev\u0113rojami at\u0161\u0137iras p\u0113c veida, aug\u0161anas stadij\u0101m un spektr\u0101laj\u0101m iez\u012bm\u0113m.<\/p>\n

Kas ir hiperspektr\u0101l\u0101 att\u0113lveido\u0161ana un CMTNet ietvars?<\/h2>\n

Hiperspektr\u0101l\u0101 att\u0113lveido\u0161ana (HSI) \u2014 tehnolo\u0123ija, kas uztver datus simtiem \u0161auru, nep\u0101rtrauktu vi\u013c\u0146u garuma joslu, \u2014 ir k\u013cuvusi par revolucion\u0101ru tehnolo\u0123iju \u0161aj\u0101 jom\u0101. At\u0161\u0137ir\u012bb\u0101 no standarta RGB kamer\u0101m vai multispektr\u0101lajiem sensoriem, kas apkopo datus da\u017e\u0101s plat\u0101s josl\u0101s, HSI nodro\u0161ina detaliz\u0113tu \u201cspektr\u0101lo pirkstu nospiedumu\u201d katram pikselim.<\/p>\n

Piem\u0113ram, vesel\u012bga ve\u0123et\u0101cija hlorofila aktivit\u0101tes d\u0113\u013c sp\u0113c\u012bgi atstaro tuv\u0101 infrasarkan\u0101 starojuma gaismu, savuk\u0101rt stresa skart\u0101m kult\u016br\u0101m ir at\u0161\u0137ir\u012bgi absorbcijas mode\u013ci. Re\u0123istr\u0113jot \u0161\u012bs smalk\u0101s vari\u0101cijas (no 400 l\u012bdz 1000 nanometriem) ar augstu telpisko iz\u0161\u0137irtsp\u0113ju (pat 0,043 metri), HSI \u013cauj prec\u012bzi diferenc\u0113t kult\u016braugu sugas, noteikt slim\u012bbas un veikt augsnes anal\u012bzi.<\/p>\n

Neskatoties uz \u0161\u012bm priek\u0161roc\u012bb\u0101m, eso\u0161aj\u0101m metod\u0113m ir gr\u016bti l\u012bdzsvarot lok\u0101las deta\u013cas, piem\u0113ram, lapu tekst\u016bru vai augsnes mode\u013cus, ar glob\u0101liem mode\u013ciem, piem\u0113ram, liela m\u0113roga kult\u016braugu izplat\u012bbu. \u0160is ierobe\u017eojums k\u013c\u016bst \u012bpa\u0161i ac\u012bmredzams trok\u0161\u0146ainos vai nel\u012bdzsvarotos datu kopumos, kur nelielas spektr\u0101l\u0101s at\u0161\u0137ir\u012bbas starp kult\u016braugiem var izrais\u012bt nepareizu klasifik\u0101ciju.<\/p>\n

Lai risin\u0101tu \u0161\u012bs probl\u0113mas, p\u0113tnieki izstr\u0101d\u0101ja\u00a0CMTNet<\/strong>\u00a0(Convolutional Meets Transformer Network \u2014 konvolucion\u0101lais t\u012bkls satiek transformatoru t\u012bklu) \u2014 jauns dzi\u013c\u0101s m\u0101c\u012b\u0161an\u0101s ietvars, kas apvieno konvolucion\u0101lo neironu t\u012bklu (CNN) un transformatoru stipr\u0101s puses. CNN ir neironu t\u012bklu klase, kas paredz\u0113ta re\u017e\u0123veida datu, piem\u0113ram, att\u0113lu, apstr\u0101dei, izmantojot filtru sl\u0101\u0146us, kas nosaka telpisk\u0101s hierarhijas (piem\u0113ram, malas, tekst\u016bras).<\/p>\n

\"CMTNet<\/p>\n

Transformatori, kas s\u0101kotn\u0113ji tika izstr\u0101d\u0101ti dabisk\u0101s valodas apstr\u0101dei, izmanto pa\u0161nov\u0113ro\u0161anas meh\u0101nismus, lai model\u0113tu datu t\u0101la darb\u012bbas r\u0101diusa atkar\u012bbas, padarot tos prasm\u012bgus glob\u0101lu mode\u013cu tver\u0161an\u0101. At\u0161\u0137ir\u012bb\u0101 no iepriek\u0161\u0113jiem mode\u013ciem, kas sec\u012bgi apstr\u0101d\u0101 lok\u0101l\u0101s un glob\u0101l\u0101s paz\u012bmes, CMTNet izmanto paral\u0113lu arhitekt\u016bru, lai vienlaikus ieg\u016btu abu veidu inform\u0101ciju.<\/p>\n

\u0160\u012b pieeja ir izr\u0101d\u012bjusies \u013coti efekt\u012bva, sasniedzot visaugst\u0101ko precizit\u0101ti tr\u012bs galvenajos bezpilota lidapar\u0101tu (UAV) HSI datu kop\u0101s. Piem\u0113ram, WHU-Hi-LongKou datu kop\u0101 CMTNet sasniedza kop\u0113jo precizit\u0101ti (OA) 99,58%, p\u0101rsp\u0113jot iepriek\u0161\u0113jo lab\u0101ko modeli par 0,19%.<\/p>\n

Tradicion\u0101l\u0101s hiperspektr\u0101l\u0101s att\u0113lveido\u0161anas izaicin\u0101jumi lauksaimniec\u012bbas klasifik\u0101cij\u0101<\/h2>\n

Agr\u012bn\u0101s hiperspektr\u0101lo datu anal\u012bzes metodes bie\u017ei koncentr\u0113j\u0101s uz spektr\u0101laj\u0101m vai telpiskaj\u0101m iez\u012bm\u0113m, k\u0101 rezult\u0101t\u0101 rezult\u0101ti bija nepiln\u012bgi. Spektr\u0101l\u0101s metodes, piem\u0113ram, galveno komponentu anal\u012bze (PCA), samazin\u0101ja datu sare\u017e\u0123\u012bt\u012bbu, koncentr\u0113joties uz vi\u013c\u0146a garuma inform\u0101ciju, bet ignor\u0113ja telpisk\u0101s attiec\u012bbas starp pikse\u013ciem.<\/p>\n

Piem\u0113ram, PCA p\u0101rveido daudzdimension\u0101lus spektr\u0101los datus maz\u0101k\u0101 skait\u0101 komponentu, kas izskaidro visliel\u0101ko dispersiju, vienk\u0101r\u0161ojot anal\u012bzi. Tom\u0113r \u0161\u012b pieeja atmet telpisko kontekstu, piem\u0113ram, kult\u016braugu izvietojumu lauk\u0101. Turpret\u012b telpisk\u0101s metodes, piem\u0113ram, matem\u0101tiskie morfolo\u0123ijas operatori, izc\u0113la kult\u016braugu fizisk\u0101 izvietojuma mode\u013cus, bet ignor\u0113ja kritiskas spektr\u0101l\u0101s deta\u013cas.<\/p>\n

Matem\u0101tisk\u0101 morfolo\u0123ija izmanto t\u0101das darb\u012bbas k\u0101 dilat\u0101cija un erozija, lai no att\u0113liem ieg\u016btu formas un strukt\u016bras, piem\u0113ram, robe\u017eas starp laukiem. Laika gait\u0101 konvolucion\u0101lie neironu t\u012bkli (CNN) uzlaboja klasifik\u0101ciju, apstr\u0101d\u0101jot abu veidu datus.<\/p>\n

Tom\u0113r to fiks\u0113tie uztveres lauki \u2014 att\u0113la laukums, ko t\u012bkls var "redz\u0113t" vienlaikus \u2014 ierobe\u017eoja to sp\u0113ju uztvert t\u0101las darb\u012bbas atkar\u012bbas. Piem\u0113ram, 3D-CNN var\u0113tu b\u016bt gr\u016bti at\u0161\u0137irt divas sojas pupi\u0146u \u0161\u0137irnes ar l\u012bdz\u012bgiem spektr\u0101lajiem profiliem, bet at\u0161\u0137ir\u012bgiem aug\u0161anas mode\u013ciem liel\u0101 lauk\u0101.<\/p>\n

Transformatori \u2014 neironu t\u012bkla veids, kas s\u0101kotn\u0113ji tika izstr\u0101d\u0101ts dabisk\u0101s valodas apstr\u0101dei, pied\u0101v\u0101ja risin\u0101jumu \u0161ai probl\u0113mai. Izmantojot pa\u0161nov\u0113rs\u012bbas meh\u0101nismus, Transformatori izce\u013cas ar glob\u0101lu attiec\u012bbu model\u0113\u0161anu datos. Pa\u0161nov\u0113rs\u012bba \u013cauj modelim izv\u0113rt\u0113t da\u017e\u0101du ievades sec\u012bbas da\u013cu noz\u012bmi, \u013caujot tam koncentr\u0113ties uz atbilsto\u0161iem re\u0123ioniem (piem\u0113ram, slimu augu kopu), vienlaikus ignor\u0113jot troksni (piem\u0113ram, m\u0101ko\u0146u \u0113nas).<\/p>\n

Tom\u0113r tie bie\u017ei vien nepamana s\u012bkgraudainas lok\u0101las deta\u013cas, piem\u0113ram, lapu malas vai augsnes plaisas. Hibr\u012bdie mode\u013ci, piem\u0113ram, CTMixer, m\u0113\u0123in\u0101ja apvienot CNN un Transformer, bet dar\u012bja to sec\u012bgi, vispirms apstr\u0101d\u0101jot lok\u0101l\u0101s paz\u012bmes un v\u0113l\u0101k glob\u0101l\u0101s paz\u012bmes. \u0160\u012b pieeja noveda pie neefekt\u012bvas inform\u0101cijas sapludin\u0101\u0161anas un neoptim\u0101las veiktsp\u0113jas sare\u017e\u0123\u012bt\u0101s lauksaimniec\u012bbas vid\u0113s.<\/p>\n

K\u0101 darbojas CMTNet: lok\u0101lo un glob\u0101lo funkciju savieno\u0161ana<\/h2>\n

CMTNet p\u0101rvar \u0161os ierobe\u017eojumus, izmantojot unik\u0101lu tr\u012bsda\u013c\u012bgu arhitekt\u016bru, kas paredz\u0113ta spektr\u0101li telpisko, lok\u0101lo un glob\u0101lo elementu efekt\u012bvai ieg\u016b\u0161anai un apvieno\u0161anai.<\/p>\n

1.<\/strong> Pirm\u0101 sast\u0101vda\u013ca, spektr\u0101li telpisko paz\u012bmju ieguves modulis<\/strong>, apstr\u0101d\u0101 neapstr\u0101d\u0101tus HSI datus, izmantojot 3D un 2D konvolucion\u0101los sl\u0101\u0146us.<\/p>\n

3D konvolucion\u0101lie sl\u0101\u0146i vienlaikus analiz\u0113 gan telpisko (augstums \u00d7 platums), gan spektr\u0101lo (vi\u013c\u0146a garums) dimensiju, tverot t\u0101dus mode\u013cus k\u0101 noteiktu vi\u013c\u0146u garumu atstaro\u0161anos p\u0101ri kult\u016braugu vainagam. Piem\u0113ram, 3D grauds var\u0113tu noteikt, ka vesel\u012bga kukur\u016bza aug\u0161\u0113j\u0101s lap\u0101s atstaro vair\u0101k tuv\u0101 infrasarkan\u0101 starojuma sal\u012bdzin\u0101jum\u0101 ar apak\u0161\u0113j\u0101m lap\u0101m.<\/p>\n

P\u0113c tam 2D sl\u0101\u0146i preciz\u0113 \u0161\u012bs iez\u012bmes, koncentr\u0113joties uz telpisk\u0101m deta\u013c\u0101m, piem\u0113ram, augu izvietojumu lauk\u0101. \u0160is divpak\u0101pju process nodro\u0161ina gan spektr\u0101l\u0101s daudzveid\u012bbas (piem\u0113ram, hlorofila satura), gan telpisk\u0101 konteksta (piem\u0113ram, rindu atstarpes) saglab\u0101\u0161anu.<\/p>\n

2.<\/strong> Otr\u0101 sast\u0101vda\u013ca, lok\u0101li glob\u0101ls paz\u012bmju ieguves modulis<\/strong>, darbojas paral\u0113li. Viena atzara izmanto CNN, lai koncentr\u0113tos uz lok\u0101l\u0101m deta\u013c\u0101m, piem\u0113ram, atsevi\u0161\u0137u lapu tekst\u016bru vai augsnes plankumu formu. \u0160\u012bs paz\u012bmes ir kritiski svar\u012bgas, lai identific\u0113tu sugas ar l\u012bdz\u012bgiem spektr\u0101lajiem profiliem, piem\u0113ram, da\u017e\u0101das sojas pupi\u0146u \u0161\u0137irnes.<\/p>\n

Otra nozare izmanto Transformerus, lai model\u0113tu glob\u0101las attiec\u012bbas, piem\u0113ram, k\u0101 kult\u016braugi ir sadal\u012bti pla\u0161\u0101s plat\u012bb\u0101s vai k\u0101 tuvum\u0101 eso\u0161o koku \u0113nas ietekm\u0113 spektr\u0101los r\u0101d\u012bjumus. Apstr\u0101d\u0101jot \u0161\u012bs paz\u012bmes vienlaic\u012bgi, nevis sec\u012bgi, CMTNet nov\u0113r\u0161 inform\u0101cijas zudumu, kas nomoka agr\u0101kos hibr\u012bdmode\u013cus.<\/p>\n

Piem\u0113ram, kam\u0113r CNN atzars identific\u0113 kokvilnas lapu robain\u0101s malas, Transformer atzars atpaz\u012bst, ka \u0161\u012bs lapas ir da\u013ca no liel\u0101ka kokvilnas lauka, ko ierobe\u017eo sezama augi.<\/p>\n

3.<\/strong> Tre\u0161\u0101 sast\u0101vda\u013ca, vair\u0101ku izvadu ierobe\u017eojumu modulis<\/strong>, nodro\u0161ina l\u012bdzsvarotu m\u0101c\u012b\u0161anos lok\u0101laj\u0101s, glob\u0101laj\u0101s un apvienotaj\u0101s funkcij\u0101s. Apm\u0101c\u012bbas laik\u0101 katram funkciju veidam tiek piem\u0113rotas atsevi\u0161\u0137as zaud\u0113jumu funkcijas, piespie\u017eot t\u012bklu preciz\u0113t visus savas izpratnes aspektus.<\/p>\n

Zaud\u0113jumu funkcija kvantific\u0113 starp\u012bbu starp prognoz\u0113taj\u0101m un faktiskaj\u0101m v\u0113rt\u012bb\u0101m, vadot mode\u013ca korekcijas. Piem\u0113ram, lok\u0101lo paz\u012bmju zudums var sod\u012bt modeli par lapu malu nepareizu klasific\u0113\u0161anu, savuk\u0101rt glob\u0101lie zudumi labo k\u013c\u016bdas liela m\u0113roga kult\u016braugu izplat\u012bb\u0101.<\/p>\n

\u0160ie zudumi tiek apvienoti, izmantojot svarus, kas optimiz\u0113ti ar nejau\u0161as mekl\u0113\u0161anas pal\u012bdz\u012bbu \u2014 metodi, kas p\u0101rbauda da\u017e\u0101das svaru kombin\u0101cijas, lai maksim\u0101li palielin\u0101tu precizit\u0101ti. \u0160is process rada stabilu un piel\u0101gojamu modeli, kas sp\u0113j apstr\u0101d\u0101t da\u017e\u0101dus lauksaimniec\u012bbas scen\u0101rijus.<\/p>\n

CMTNet veiktsp\u0113jas nov\u0113rt\u0113\u0161ana bezpilota lidapar\u0101tu hiperspektr\u0101los datu kopumos<\/h2>\n

Lai nov\u0113rt\u0113tu CMTNet, p\u0113tnieki to test\u0113ja ar trim bezpilota lidapar\u0101tu (UAV) ieg\u016btiem hiperspektr\u0101lajiem datu kopumiem no Uha\u0146as Universit\u0101tes. \u0160ie datu kopumi tiek pla\u0161i izmantoti t\u0101lizp\u0113tes etaloni to augst\u0101s kvalit\u0101tes un daudzveid\u012bbas d\u0113\u013c:<\/p>\n

    \n
  1. WHU-Hi-LongKou<\/strong>\u0160is datu kopums aptver 550 \u00d7 400 pikse\u013cus ar 270 spektr\u0101laj\u0101m josl\u0101m un telpisko iz\u0161\u0137irtsp\u0113ju 0,463 metri. Telpisk\u0101 iz\u0161\u0137irtsp\u0113ja 0,463 metri noz\u012bm\u0113, ka katrs pikselis att\u0113lo 0,463 m \u00d7 0,463 m lielu plat\u012bbu uz zemes, kas \u013cauj identific\u0113t atsevi\u0161\u0137us augus. Taj\u0101 ir iek\u013cauti devi\u0146i kult\u016braugu veidi, piem\u0113ram, kukur\u016bza, kokvilna un r\u012bsi, ar 1019 apm\u0101c\u012bbas paraugiem un 203\u00a0523 testa paraugiem.<\/li>\n
  2. WHU-Hi-Hanchuan<\/strong>\u0160aj\u0101 datu kop\u0101, kas uztver 1217 \u00d7 303 pikse\u013cus ar 0,109 metru iz\u0161\u0137irtsp\u0113ju, ir iek\u013cauti 16 zemes seguma veidi, tostarp zemenes, sojas pupas un plastmasas pl\u0113ves. Augst\u0101ka iz\u0161\u0137irtsp\u0113ja (0,109 m) nodro\u0161ina s\u012bk\u0101ku inform\u0101ciju, piem\u0113ram, at\u0161\u0137ir\u012bbu starp jauniem un nobriedu\u0161iem sojas pupi\u0146u augiem. Apm\u0101c\u012bbas un testa paraugu kopsumma bija attiec\u012bgi 1289 un 256\u00a0241.<\/li>\n
  3. WHU-Hi-HongHu<\/strong>Ar 940 \u00d7 475 pikse\u013ciem un 270 josl\u0101m \u0161is augstas iz\u0161\u0137irtsp\u0113jas (0,043 metri) datu kopums ietver 22 klases, piem\u0113ram, kokvilnas, rap\u0161a un \u0137iploku asnus. Ar 0,043 m iz\u0161\u0137irtsp\u0113ju ir redzamas atsevi\u0161\u0137as lapas un augsnes plaisas, padarot to ide\u0101li piem\u0113rotu smalkgraudainai klasifik\u0101cijai. Tas satur 1925 apm\u0101c\u012bbas paraugus un 384\u00a0678 testa paraugus.<\/li>\n<\/ol>\n

    \"Augstas<\/p>\n

    Modelis tika apm\u0101c\u012bts NVIDIA TITAN Xp GPU, izmantojot PyTorch, ar m\u0101c\u012b\u0161an\u0101s \u0101trumu 0,001 un partijas lielumu 100. M\u0101c\u012b\u0161an\u0101s \u0101trums nosaka, cik liel\u0101 m\u0113r\u0101 modelis piel\u0101go savus parametrus apm\u0101c\u012bbas laik\u0101 \u2014 ja tas ir p\u0101r\u0101k augsts, tas var p\u0101rsniegt optim\u0101l\u0101s v\u0113rt\u012bbas; ja tas ir p\u0101r\u0101k zems, apm\u0101c\u012bba k\u013c\u016bst l\u0113na.<\/p>\n

    Katrs eksperiments tika atk\u0101rtots desmit reizes, lai nodro\u0161in\u0101tu uzticam\u012bbu, un ievades iel\u0101pi \u2014 nelieli pilna att\u0113la segmenti \u2014 tika optimiz\u0113ti l\u012bdz 13 \u00d7 13 pikse\u013ciem, izmantojot re\u017e\u0123a mekl\u0113\u0161anu \u2014 metodi, kas p\u0101rbauda da\u017e\u0101dus iel\u0101pu izm\u0113rus, lai atrastu visefekt\u012bv\u0101kos.<\/p>\n

    CMTNet sasniedz vismodern\u0101ko precizit\u0101ti kult\u016braugu klasifik\u0101cij\u0101<\/h2>\n

    CMTNet sasniedza iev\u0113rojamus rezult\u0101tus vis\u0101s datu kop\u0101s, p\u0101rsp\u0113jot eso\u0161\u0101s metodes gan kop\u0113j\u0101s precizit\u0101tes (OA), gan klases specifisk\u0101s veiktsp\u0113jas zi\u0146\u0101. OA m\u0113ra pareizi klasific\u0113to pikse\u013cu procentu\u0101lo daudzumu vis\u0101s klas\u0113s, savuk\u0101rt vid\u0113j\u0101 precizit\u0101te (AA) apr\u0113\u0137ina vid\u0113jo precizit\u0101ti katr\u0101 klas\u0113, nov\u0113r\u0161ot nel\u012bdzsvarot\u012bbu.<\/p>\n

    WHU-Hi-LongKou datu kop\u0101 CMTNet sasniedza OA 99,58%, p\u0101rsp\u0113jot CTMixer par 0,19%. Sare\u017e\u0123\u012bt\u0101m klas\u0113m ar ierobe\u017eotiem apm\u0101c\u012bbas datiem, piem\u0113ram, kokvilnai (41 paraugs), CMTNet joproj\u0101m sasniedza 99,53% precizit\u0101ti. L\u012bdz\u012bgi WHU-Hi-HanChuan datu kop\u0101 tas uzlaboja arb\u016bza (22 paraugs) precizit\u0101ti no 82,42% l\u012bdz 96,11%, demonstr\u0113jot sp\u0113ju apstr\u0101d\u0101t nel\u012bdzsvarotus datus, izmantojot efekt\u012bvu iez\u012bmju sapludin\u0101\u0161anu.<\/p>\n

    Klasifik\u0101cijas kar\u0161u vizu\u0101l\u0101 sal\u012bdzin\u0101\u0161ana atkl\u0101ja maz\u0101k fragment\u0113tu plankumu un vienm\u0113r\u012bg\u0101kas robe\u017eas starp laukiem, sal\u012bdzinot ar t\u0101diem mode\u013ciem k\u0101 3D-CNN un Vision Transformer (ViT). Piem\u0113ram, \u0113n\u0101m pak\u013cautaj\u0101 WHU-Hi-HanChuan datu kop\u0101 CMTNet samazin\u0101ja k\u013c\u016bdas, ko izrais\u012bja zems saules le\u0146\u0137is, savuk\u0101rt ResNet k\u013c\u016bdaini klasific\u0113ja sojas pupi\u0146as k\u0101 pel\u0113kus jumtus.<\/p>\n

    \"CMTNet<\/p>\n

    \u0112nas rada unik\u0101lu izaicin\u0101jumu, jo t\u0101s maina spektr\u0101los raksturlielumus \u2014 \u0113n\u0101 eso\u0161s sojas pupi\u0146u augs var atstarot maz\u0101k tuv\u0101s infrasarkan\u0101s gaismas, atg\u0101dinot neve\u0123et\u0101ciju. Izmantojot glob\u0101lo kontekstu, CMTNet atzina, ka \u0161ie \u0113nainie augi ir da\u013ca no liel\u0101ka sojas pupi\u0146u lauka, t\u0101d\u0113j\u0101di samazinot k\u013c\u016bdas.<\/p>\n

    WHU-Hi-HongHu datu kop\u0101 modelis izc\u0113l\u0101s, at\u0161\u0137irot spektr\u0101li l\u012bdz\u012bgas kult\u016bras, piem\u0113ram, da\u017e\u0101das krustzie\u017eu dzimtas augu \u0161\u0137irnes, sasniedzot 96,54% precizit\u0101ti.\u00a0Brassica parachinensis<\/em>.<\/p>\n

    Abl\u0101cijas p\u0113t\u012bjumi \u2014 eksperimenti, kuros komponenti tiek no\u0146emti, lai nov\u0113rt\u0113tu to ietekmi \u2014 apstiprin\u0101ja katra modu\u013ca noz\u012bmi. Pievienojot tikai vair\u0101ku izvades ierobe\u017eojumu moduli, OA palielin\u0101j\u0101s par 1,52% WHU-Hi-HongHu, uzsverot t\u0101 lomu iez\u012bmju sapludin\u0101\u0161anas uzlabo\u0161an\u0101. Bez \u0161\u012b modu\u013ca lok\u0101l\u0101s un glob\u0101l\u0101s paz\u012bmes tika apvienotas haotiski, k\u0101 rezult\u0101t\u0101 rad\u0101s nekonsekventas klasifik\u0101cijas.<\/p>\n

    Skait\u013co\u0161anas kompromisi un praktiski apsv\u0113rumi<\/h2>\n

    Lai gan CMTNet precizit\u0101te ir nep\u0101rsp\u0113jama, t\u0101 skait\u013co\u0161anas izmaksas ir augst\u0101kas nek\u0101 tradicion\u0101laj\u0101m metod\u0113m. Apm\u0101c\u012bba ar WHU-Hi-HongHu datu kopu aiz\u0146\u0113ma 1885 sekundes, sal\u012bdzinot ar 74 sekund\u0113m Random Forest (RF) \u2014 ma\u0161\u012bnm\u0101c\u012b\u0161an\u0101s algoritmam, kas apm\u0101c\u012bbas laik\u0101 veido l\u0113mumu kokus.<\/p>\n

    Tom\u0113r \u0161is kompromiss ir pamatots prec\u012bzaj\u0101 lauksaimniec\u012bb\u0101, kur precizit\u0101te tie\u0161i ietekm\u0113 ra\u017eas prognozes un resursu sadali. Piem\u0113ram, slimas kult\u016bras nepareiza klasific\u0113\u0161ana k\u0101 vesel\u012bga var izrais\u012bt nekontrol\u0113tus kait\u0113k\u013cu uzliesmojumus, izn\u012bcinot veselus laukus.<\/p>\n

    Re\u0101llaika lietojumprogramm\u0101m turpm\u0101kaj\u0101 darb\u0101 var\u0113tu izp\u0113t\u012bt mode\u013cu saspie\u0161anas metodes, piem\u0113ram, lieko neironu apgrie\u0161anu vai svaru kvant\u0113\u0161anu (samazinot skaitlisko precizit\u0101ti), lai samazin\u0101tu izpildes laiku, nezaud\u0113jot veiktsp\u0113ju. Apgrie\u0161ana no\u0146em no neironu t\u012bkla maz\u0101k svar\u012bgus savienojumus, l\u012bdz\u012bgi k\u0101 zaru apgrie\u0161ana no koka, lai uzlabotu t\u0101 formu, savuk\u0101rt kvant\u0113\u0161ana vienk\u0101r\u0161o skaitliskos apr\u0113\u0137inus, pa\u0101trinot apstr\u0101di.<\/p>\n

    Hiperspektr\u0101l\u0101s kult\u016braugu klasifik\u0101cijas n\u0101kotne ar CMTNet<\/h2>\n

    Neskatoties uz pan\u0101kumiem, CMTNet saskaras ar ierobe\u017eojumiem. Veiktsp\u0113ja nedaudz pasliktin\u0101s stipri \u0113notos apgabalos, k\u0101 redzams WHU-Hi-HanChuan datu kop\u0101 (97.29% OA pret 99.58% labi apgaismot\u0101 LongKou). \u0112nas sare\u017e\u0123\u012b klasifik\u0101ciju, jo t\u0101s samazina atstarot\u0101s gaismas intensit\u0101ti, mainot spektra profilus.<\/p>\n

    Turkl\u0101t klases ar \u0101rk\u0101rt\u012bgi maziem apm\u0101c\u012bbas paraugiem, piem\u0113ram, \u0161aurlapu sojas pupi\u0146as (20 paraugi), atpaliek no t\u0101m, kur\u0101m ir daudz datu. Mazs paraugu lielums ierobe\u017eo mode\u013ca sp\u0113ju apg\u016bt da\u017e\u0101das vari\u0101cijas, piem\u0113ram, lapu formas at\u0161\u0137ir\u012bbas augsnes kvalit\u0101tes d\u0113\u013c.<\/p>\n

    Turpm\u0101kajos p\u0113t\u012bjumos var\u0113tu integr\u0113t multimod\u0101lus datus, piem\u0113ram, LiDAR augstuma kartes vai termisko att\u0113lveido\u0161anu, lai uzlabotu notur\u012bbu pret \u0113n\u0101m un aizsegumiem. LiDAR (gaismas noteik\u0161ana un diapazona noteik\u0161ana) izmanto l\u0101zera impulsus, lai izveidotu 3D reljefa mode\u013cus, kas var\u0113tu pal\u012bdz\u0113t at\u0161\u0137irt kult\u016braugus no \u0113n\u0101m, analiz\u0113jot augstuma at\u0161\u0137ir\u012bbas.<\/p>\n

    Turkl\u0101t termisk\u0101 att\u0113lveido\u0161ana uztver siltuma sign\u0101lus, sniedzot papildu nor\u0101des par augu vesel\u012bbu \u2014 stresa skart\u0101m kult\u016br\u0101m bie\u017ei ir augst\u0101ka lapotnes temperat\u016bra samazin\u0101tas transpir\u0101cijas d\u0113\u013c. Da\u013c\u0113ji uzraudz\u012btas m\u0101c\u012b\u0161an\u0101s metodes, kas izmanto nemar\u0137\u0113tus datus (piem\u0113ram, bezpilota lidapar\u0101tu att\u0113lus bez manu\u0101l\u0101m anot\u0101cij\u0101m), var\u0113tu ar\u012b uzlabot retu kult\u016braugu veidu veiktsp\u0113ju.<\/p>\n

    Izmantojot konsekvences regulariz\u0101ciju \u2014 apm\u0101cot modeli, lai ieg\u016btu stabilas prognozes nedaudz main\u012bt\u0101s viena un t\u0101 pa\u0161a att\u0113la versij\u0101s \u2014, p\u0113tnieki var izmantot nemar\u0137\u0113tus datus, lai uzlabotu visp\u0101rin\u0101\u0161anu.<\/p>\n

    Visbeidzot, CMTNet izvieto\u0161ana perif\u0113rijas ier\u012bc\u0113s, piem\u0113ram, dronos, kas apr\u012bkoti ar ieb\u016bv\u0113tiem grafiskajiem procesoriem, var\u0113tu nodro\u0161in\u0101t re\u0101llaika uzraudz\u012bbu att\u0101los apst\u0101k\u013cos. Perif\u0113rijas izvieto\u0161ana samazina atkar\u012bbu no m\u0101ko\u0146dato\u0161anas, samazinot latentumu un datu p\u0101rraides izmaksas. Tom\u0113r tas prasa mode\u013ca optimiz\u0101ciju ierobe\u017eotai atmi\u0146ai un apstr\u0101des jaudai, iesp\u0113jams, izmantojot vieglas arhitekt\u016bras, piem\u0113ram, MobileNet vai zin\u0101\u0161anu destil\u0101ciju, kur maz\u0101ks \u201cstudenta\u201d modelis atdarina liel\u0101ku \u201cskolot\u0101ja\u201d modeli.<\/p>\n

    Secin\u0101jums<\/h2>\n

    CMTNet ir iev\u0113rojams solis uz priek\u0161u hiperspektr\u0101lo kult\u016braugu klasifik\u0101cij\u0101. Saska\u0146ojot CNN un Transformerus, tas risina ilgsto\u0161as probl\u0113mas iez\u012bmju ieg\u016b\u0161an\u0101 un sapludin\u0101\u0161an\u0101, pied\u0101v\u0101jot lauksaimniekiem un agronomiem jaud\u012bgu instrumentu prec\u012bzajai lauksaimniec\u012bbai.<\/p>\n

    Pielietojumi aptver visu, s\u0101kot no slim\u012bbu atkl\u0101\u0161anas re\u0101llaik\u0101 l\u012bdz ap\u016bde\u0146o\u0161anas grafiku optimiz\u0113\u0161anai, un tas viss ir \u013coti svar\u012bgi ilgtsp\u0113j\u012bgai lauksaimniec\u012bbai klimata p\u0101rmai\u0146u un iedz\u012bvot\u0101ju skaita pieauguma apst\u0101k\u013cos. T\u0101 k\u0101 bezpilota lidapar\u0101tu tehnolo\u0123ija k\u013c\u016bst pieejam\u0101ka, t\u0101diem mode\u013ciem k\u0101 CMTNet b\u016bs iz\u0161\u0137iro\u0161a loma glob\u0101laj\u0101 p\u0101rtikas nodro\u0161in\u0101jum\u0101.<\/p>\n

    N\u0101kotnes sasniegumi, piem\u0113ram, viegl\u0101kas arhitekt\u016bras un multimod\u0101la datu sapludin\u0101\u0161ana, var\u0113tu v\u0113l vair\u0101k uzlabot to praktiskumu. Ar past\u0101v\u012bgu inov\u0101ciju CMTNet var\u0113tu k\u013c\u016bt par vied\u0101s lauksaimniec\u012bbas sist\u0113mu st\u016brakmeni vis\u0101 pasaul\u0113, nodro\u0161inot efekt\u012bvu zemes izmanto\u0161anu un notur\u012bgu p\u0101rtikas ra\u017eo\u0161anu n\u0101kamaj\u0101m paaudz\u0113m.<\/p>\n

    Atsauce: <\/strong>Guo, X., Feng, Q. un Guo, F. CMTNet: hibr\u012bds CNN transformatoru t\u012bkls bezpilota lidapar\u0101tu (UAV) hiperspektr\u0101lai kult\u016braugu klasifik\u0101cijai prec\u012bzaj\u0101 lauksaimniec\u012bb\u0101. Sci Rep 15, 12383 (2025). https:\/\/doi.org\/10.1038\/s41598-025-97052-w<\/a><\/p>","protected":false},"excerpt":{"rendered":"

    Prec\u012bza kult\u016braugu klasifik\u0101cija ir b\u016btiska m\u016bsdienu prec\u012bzajai lauksaimniec\u012bbai, \u013caujot lauksaimniekiem uzraudz\u012bt kult\u016braugu vesel\u012bbu, prognoz\u0113t ra\u017eu un efekt\u012bvi sadal\u012bt resursus. Tom\u0113r tradicion\u0101l\u0101s metodes bie\u017ei vien\u2026<\/p>","protected":false},"author":210157960,"featured_media":11556,"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":"","content-type":"","_jetpack_newsletter_access":"","_jetpack_dont_email_post_to_subs":false,"_jetpack_newsletter_tier_id":0,"_jetpack_memberships_contains_paywalled_content":false,"_wpcom_ai_launchpad_first_post":false,"_jetpack_feature_clip_id":0,"_jetpack_memberships_contains_paid_content":false,"footnotes":"","jetpack_publicize_message":"{title}\n\n{excerpt}\n\n{url}","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,"jetpack_post_was_ever_published":false},"categories":[1657,1377],"tags":[],"class_list":["post-11549","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-precision-farming","category-crop-monitoring"],"acf":[],"yoast_head":"\nCMTNet Redefines Precision Agriculture By Outperforming Traditional Crop Classification - 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\/lv\/emuars\/cmtnet-no-jauna-define-precizo-lauksaimniecibu-parspejot-tradicionalo-kulturaugu-klasifikaciju\/\" \/>\n<meta property=\"og:locale\" content=\"lv_LV\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"CMTNet Redefines Precision Agriculture By Outperforming Traditional Crop Classification - GeoPard Agriculture\" \/>\n<meta property=\"og:description\" content=\"Accurate crop classification is essential for modern precision agriculture, enabling farmers to monitor crop health, predict yields, and allocate resources efficiently. Traditional methods, however, often...\" \/>\n<meta property=\"og:url\" content=\"https:\/\/geopard.tech\/lv\/emuars\/cmtnet-no-jauna-define-precizo-lauksaimniecibu-parspejot-tradicionalo-kulturaugu-klasifikaciju\/\" \/>\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=\"2025-05-04T19:00:25+00:00\" \/>\n<meta property=\"og:image\" content=\"https:\/\/geopard.tech\/wp-content\/uploads\/2025\/05\/CMTNet-Redefines-Precision-Agriculture-By-Outperforming-Traditional-Crop-Classification-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=\"dementievgeopard\" \/>\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=\"dementievgeopard\" \/>\n\t<meta name=\"twitter:label2\" content=\"Est. reading time\" \/>\n\t<meta name=\"twitter:data2\" content=\"10 min\u016b\u0161u\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\\\/\\\/schema.org\",\"@graph\":[{\"@type\":\"Article\",\"@id\":\"https:\\\/\\\/geopard.tech\\\/blog\\\/cmtnet-redefines-precision-agriculture-by-outperforming-traditional-crop-classification\\\/#article\",\"isPartOf\":{\"@id\":\"https:\\\/\\\/geopard.tech\\\/blog\\\/cmtnet-redefines-precision-agriculture-by-outperforming-traditional-crop-classification\\\/\"},\"author\":{\"name\":\"dementievgeopard\",\"@id\":\"https:\\\/\\\/geopard.tech\\\/#\\\/schema\\\/person\\\/dd217733c742620adc57befbbcd84a8a\"},\"headline\":\"CMTNet Redefines Precision Agriculture By Outperforming Traditional Crop Classification\",\"datePublished\":\"2025-05-04T19:00:25+00:00\",\"mainEntityOfPage\":{\"@id\":\"https:\\\/\\\/geopard.tech\\\/blog\\\/cmtnet-redefines-precision-agriculture-by-outperforming-traditional-crop-classification\\\/\"},\"wordCount\":2064,\"commentCount\":0,\"publisher\":{\"@id\":\"https:\\\/\\\/geopard.tech\\\/#organization\"},\"image\":{\"@id\":\"https:\\\/\\\/geopard.tech\\\/blog\\\/cmtnet-redefines-precision-agriculture-by-outperforming-traditional-crop-classification\\\/#primaryimage\"},\"thumbnailUrl\":\"https:\\\/\\\/i0.wp.com\\\/geopard.tech\\\/wp-content\\\/uploads\\\/2025\\\/05\\\/CMTNet-Redefines-Precision-Agriculture-By-Outperforming-Traditional-Crop-Classification.png?fit=3600%2C2025&ssl=1\",\"articleSection\":[\"Precision Farming\",\"Crop monitoring\"],\"inLanguage\":\"lv-LV\",\"potentialAction\":[{\"@type\":\"CommentAction\",\"name\":\"Comment\",\"target\":[\"https:\\\/\\\/geopard.tech\\\/blog\\\/cmtnet-redefines-precision-agriculture-by-outperforming-traditional-crop-classification\\\/#respond\"]}]},{\"@type\":\"WebPage\",\"@id\":\"https:\\\/\\\/geopard.tech\\\/blog\\\/cmtnet-redefines-precision-agriculture-by-outperforming-traditional-crop-classification\\\/\",\"url\":\"https:\\\/\\\/geopard.tech\\\/blog\\\/cmtnet-redefines-precision-agriculture-by-outperforming-traditional-crop-classification\\\/\",\"name\":\"CMTNet Redefines Precision Agriculture By Outperforming Traditional Crop Classification - GeoPard Agriculture\",\"isPartOf\":{\"@id\":\"https:\\\/\\\/geopard.tech\\\/#website\"},\"primaryImageOfPage\":{\"@id\":\"https:\\\/\\\/geopard.tech\\\/blog\\\/cmtnet-redefines-precision-agriculture-by-outperforming-traditional-crop-classification\\\/#primaryimage\"},\"image\":{\"@id\":\"https:\\\/\\\/geopard.tech\\\/blog\\\/cmtnet-redefines-precision-agriculture-by-outperforming-traditional-crop-classification\\\/#primaryimage\"},\"thumbnailUrl\":\"https:\\\/\\\/i0.wp.com\\\/geopard.tech\\\/wp-content\\\/uploads\\\/2025\\\/05\\\/CMTNet-Redefines-Precision-Agriculture-By-Outperforming-Traditional-Crop-Classification.png?fit=3600%2C2025&ssl=1\",\"datePublished\":\"2025-05-04T19:00:25+00:00\",\"breadcrumb\":{\"@id\":\"https:\\\/\\\/geopard.tech\\\/blog\\\/cmtnet-redefines-precision-agriculture-by-outperforming-traditional-crop-classification\\\/#breadcrumb\"},\"inLanguage\":\"lv-LV\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\\\/\\\/geopard.tech\\\/blog\\\/cmtnet-redefines-precision-agriculture-by-outperforming-traditional-crop-classification\\\/\"]}]},{\"@type\":\"ImageObject\",\"inLanguage\":\"lv-LV\",\"@id\":\"https:\\\/\\\/geopard.tech\\\/blog\\\/cmtnet-redefines-precision-agriculture-by-outperforming-traditional-crop-classification\\\/#primaryimage\",\"url\":\"https:\\\/\\\/i0.wp.com\\\/geopard.tech\\\/wp-content\\\/uploads\\\/2025\\\/05\\\/CMTNet-Redefines-Precision-Agriculture-By-Outperforming-Traditional-Crop-Classification.png?fit=3600%2C2025&ssl=1\",\"contentUrl\":\"https:\\\/\\\/i0.wp.com\\\/geopard.tech\\\/wp-content\\\/uploads\\\/2025\\\/05\\\/CMTNet-Redefines-Precision-Agriculture-By-Outperforming-Traditional-Crop-Classification.png?fit=3600%2C2025&ssl=1\",\"width\":3600,\"height\":2025},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\\\/\\\/geopard.tech\\\/blog\\\/cmtnet-redefines-precision-agriculture-by-outperforming-traditional-crop-classification\\\/#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"Home\",\"item\":\"https:\\\/\\\/geopard.tech\\\/\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"CMTNet Redefines Precision Agriculture By Outperforming Traditional Crop Classification\"}]},{\"@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\":\"lv-LV\"},{\"@type\":\"Organization\",\"@id\":\"https:\\\/\\\/geopard.tech\\\/#organization\",\"name\":\"GeoPard Agriculture\",\"alternateName\":\"GeoPard\",\"url\":\"https:\\\/\\\/geopard.tech\\\/\",\"logo\":{\"@type\":\"ImageObject\",\"inLanguage\":\"lv-LV\",\"@id\":\"https:\\\/\\\/geopard.tech\\\/#\\\/schema\\\/logo\\\/image\\\/\",\"url\":\"https:\\\/\\\/i0.wp.com\\\/geopard.tech\\\/wp-content\\\/uploads\\\/2026\\\/07\\\/geopard_logo_transparentbackground.png?fit=512%2C68&ssl=1\",\"contentUrl\":\"https:\\\/\\\/i0.wp.com\\\/geopard.tech\\\/wp-content\\\/uploads\\\/2026\\\/07\\\/geopard_logo_transparentbackground.png?fit=512%2C68&ssl=1\",\"width\":512,\"height\":68,\"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\\\/\",\"https:\\\/\\\/www.youtube.com\\\/channel\\\/UCiaPGLAhRPNh-s85dXdC-Sw\",\"https:\\\/\\\/www.g2.com\\\/products\\\/geopard-agriculture-precision-farming-software\\\/reviews\"]},{\"@type\":\"Person\",\"@id\":\"https:\\\/\\\/geopard.tech\\\/#\\\/schema\\\/person\\\/dd217733c742620adc57befbbcd84a8a\",\"name\":\"dementievgeopard\",\"image\":{\"@type\":\"ImageObject\",\"inLanguage\":\"lv-LV\",\"@id\":\"https:\\\/\\\/secure.gravatar.com\\\/avatar\\\/7c09adac9c20b00454199df5a41ea67c812a6c97e948bf109899836ddde31354?s=96&d=identicon&r=g\",\"url\":\"https:\\\/\\\/secure.gravatar.com\\\/avatar\\\/7c09adac9c20b00454199df5a41ea67c812a6c97e948bf109899836ddde31354?s=96&d=identicon&r=g\",\"contentUrl\":\"https:\\\/\\\/secure.gravatar.com\\\/avatar\\\/7c09adac9c20b00454199df5a41ea67c812a6c97e948bf109899836ddde31354?s=96&d=identicon&r=g\",\"caption\":\"dementievgeopard\"}}]}<\/script>\n<!-- \/ Yoast SEO plugin. -->","yoast_head_json":{"title":"CMTNet p\u0101rdefin\u0113 prec\u012bzo lauksaimniec\u012bbu, p\u0101rsp\u0113jot tradicion\u0101lo kult\u016braugu klasifik\u0101ciju \u2014 GeoPard Agriculture","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\/lv\/emuars\/cmtnet-no-jauna-define-precizo-lauksaimniecibu-parspejot-tradicionalo-kulturaugu-klasifikaciju\/","og_locale":"lv_LV","og_type":"article","og_title":"CMTNet Redefines Precision Agriculture By Outperforming Traditional Crop Classification - GeoPard Agriculture","og_description":"Accurate crop classification is essential for modern precision agriculture, enabling farmers to monitor crop health, predict yields, and allocate resources efficiently. Traditional methods, however, often...","og_url":"https:\/\/geopard.tech\/lv\/emuars\/cmtnet-no-jauna-define-precizo-lauksaimniecibu-parspejot-tradicionalo-kulturaugu-klasifikaciju\/","og_site_name":"GeoPard - Precision agriculture Mapping software","article_publisher":"https:\/\/www.facebook.com\/geopardAgriculture\/","article_published_time":"2025-05-04T19:00:25+00:00","og_image":[{"width":1024,"height":576,"url":"https:\/\/geopard.tech\/wp-content\/uploads\/2025\/05\/CMTNet-Redefines-Precision-Agriculture-By-Outperforming-Traditional-Crop-Classification-1024x576.png","type":"image\/png"}],"author":"dementievgeopard","twitter_card":"summary_large_image","twitter_creator":"@geopardagri","twitter_site":"@geopardagri","twitter_misc":{"Written by":"dementievgeopard","Est. reading time":"10 min\u016b\u0161u"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"Article","@id":"https:\/\/geopard.tech\/blog\/cmtnet-redefines-precision-agriculture-by-outperforming-traditional-crop-classification\/#article","isPartOf":{"@id":"https:\/\/geopard.tech\/blog\/cmtnet-redefines-precision-agriculture-by-outperforming-traditional-crop-classification\/"},"author":{"name":"dementievgeopard","@id":"https:\/\/geopard.tech\/#\/schema\/person\/dd217733c742620adc57befbbcd84a8a"},"headline":"CMTNet Redefines Precision Agriculture By Outperforming Traditional Crop Classification","datePublished":"2025-05-04T19:00:25+00:00","mainEntityOfPage":{"@id":"https:\/\/geopard.tech\/blog\/cmtnet-redefines-precision-agriculture-by-outperforming-traditional-crop-classification\/"},"wordCount":2064,"commentCount":0,"publisher":{"@id":"https:\/\/geopard.tech\/#organization"},"image":{"@id":"https:\/\/geopard.tech\/blog\/cmtnet-redefines-precision-agriculture-by-outperforming-traditional-crop-classification\/#primaryimage"},"thumbnailUrl":"https:\/\/i0.wp.com\/geopard.tech\/wp-content\/uploads\/2025\/05\/CMTNet-Redefines-Precision-Agriculture-By-Outperforming-Traditional-Crop-Classification.png?fit=3600%2C2025&ssl=1","articleSection":["Precision Farming","Crop monitoring"],"inLanguage":"lv-LV","potentialAction":[{"@type":"CommentAction","name":"Comment","target":["https:\/\/geopard.tech\/blog\/cmtnet-redefines-precision-agriculture-by-outperforming-traditional-crop-classification\/#respond"]}]},{"@type":"WebPage","@id":"https:\/\/geopard.tech\/blog\/cmtnet-redefines-precision-agriculture-by-outperforming-traditional-crop-classification\/","url":"https:\/\/geopard.tech\/blog\/cmtnet-redefines-precision-agriculture-by-outperforming-traditional-crop-classification\/","name":"CMTNet p\u0101rdefin\u0113 prec\u012bzo lauksaimniec\u012bbu, p\u0101rsp\u0113jot tradicion\u0101lo kult\u016braugu klasifik\u0101ciju \u2014 GeoPard Agriculture","isPartOf":{"@id":"https:\/\/geopard.tech\/#website"},"primaryImageOfPage":{"@id":"https:\/\/geopard.tech\/blog\/cmtnet-redefines-precision-agriculture-by-outperforming-traditional-crop-classification\/#primaryimage"},"image":{"@id":"https:\/\/geopard.tech\/blog\/cmtnet-redefines-precision-agriculture-by-outperforming-traditional-crop-classification\/#primaryimage"},"thumbnailUrl":"https:\/\/i0.wp.com\/geopard.tech\/wp-content\/uploads\/2025\/05\/CMTNet-Redefines-Precision-Agriculture-By-Outperforming-Traditional-Crop-Classification.png?fit=3600%2C2025&ssl=1","datePublished":"2025-05-04T19:00:25+00:00","breadcrumb":{"@id":"https:\/\/geopard.tech\/blog\/cmtnet-redefines-precision-agriculture-by-outperforming-traditional-crop-classification\/#breadcrumb"},"inLanguage":"lv-LV","potentialAction":[{"@type":"ReadAction","target":["https:\/\/geopard.tech\/blog\/cmtnet-redefines-precision-agriculture-by-outperforming-traditional-crop-classification\/"]}]},{"@type":"ImageObject","inLanguage":"lv-LV","@id":"https:\/\/geopard.tech\/blog\/cmtnet-redefines-precision-agriculture-by-outperforming-traditional-crop-classification\/#primaryimage","url":"https:\/\/i0.wp.com\/geopard.tech\/wp-content\/uploads\/2025\/05\/CMTNet-Redefines-Precision-Agriculture-By-Outperforming-Traditional-Crop-Classification.png?fit=3600%2C2025&ssl=1","contentUrl":"https:\/\/i0.wp.com\/geopard.tech\/wp-content\/uploads\/2025\/05\/CMTNet-Redefines-Precision-Agriculture-By-Outperforming-Traditional-Crop-Classification.png?fit=3600%2C2025&ssl=1","width":3600,"height":2025},{"@type":"BreadcrumbList","@id":"https:\/\/geopard.tech\/blog\/cmtnet-redefines-precision-agriculture-by-outperforming-traditional-crop-classification\/#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https:\/\/geopard.tech\/"},{"@type":"ListItem","position":2,"name":"CMTNet Redefines Precision Agriculture By Outperforming Traditional Crop Classification"}]},{"@type":"WebSite","@id":"https:\/\/geopard.tech\/#website","url":"https:\/\/geopard.tech\/","name":"GeoPard - Prec\u012bzijas lauksaimniec\u012bbas programmat\u016bra","description":"Prec\u012bz\u0101s lauksaimniec\u012bbas kart\u0113\u0161anas programmat\u016bra","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":"lv-LV"},{"@type":"Organization","@id":"https:\/\/geopard.tech\/#organization","name":"GeoPard Lauksaimniec\u012bba","alternateName":"GeoPard","url":"https:\/\/geopard.tech\/","logo":{"@type":"ImageObject","inLanguage":"lv-LV","@id":"https:\/\/geopard.tech\/#\/schema\/logo\/image\/","url":"https:\/\/i0.wp.com\/geopard.tech\/wp-content\/uploads\/2026\/07\/geopard_logo_transparentbackground.png?fit=512%2C68&ssl=1","contentUrl":"https:\/\/i0.wp.com\/geopard.tech\/wp-content\/uploads\/2026\/07\/geopard_logo_transparentbackground.png?fit=512%2C68&ssl=1","width":512,"height":68,"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\/","https:\/\/www.youtube.com\/channel\/UCiaPGLAhRPNh-s85dXdC-Sw","https:\/\/www.g2.com\/products\/geopard-agriculture-precision-farming-software\/reviews"]},{"@type":"Person","@id":"https:\/\/geopard.tech\/#\/schema\/person\/dd217733c742620adc57befbbcd84a8a","name":"dementievgeopard","image":{"@type":"ImageObject","inLanguage":"lv-LV","@id":"https:\/\/secure.gravatar.com\/avatar\/7c09adac9c20b00454199df5a41ea67c812a6c97e948bf109899836ddde31354?s=96&d=identicon&r=g","url":"https:\/\/secure.gravatar.com\/avatar\/7c09adac9c20b00454199df5a41ea67c812a6c97e948bf109899836ddde31354?s=96&d=identicon&r=g","contentUrl":"https:\/\/secure.gravatar.com\/avatar\/7c09adac9c20b00454199df5a41ea67c812a6c97e948bf109899836ddde31354?s=96&d=identicon&r=g","caption":"dementievgeopard"}}]}},"jetpack_publicize_connections":[],"jetpack_likes_enabled":true,"jetpack_sharing_enabled":true,"jetpack_shortlink":"https:\/\/wp.me\/pdiCPa-30h","jetpack_featured_media_url":"https:\/\/i0.wp.com\/geopard.tech\/wp-content\/uploads\/2025\/05\/CMTNet-Redefines-Precision-Agriculture-By-Outperforming-Traditional-Crop-Classification.png?fit=3600%2C2025&ssl=1","_links":{"self":[{"href":"https:\/\/geopard.tech\/lv\/wp-json\/wp\/v2\/posts\/11549","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/geopard.tech\/lv\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/geopard.tech\/lv\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/geopard.tech\/lv\/wp-json\/wp\/v2\/users\/210157960"}],"replies":[{"embeddable":true,"href":"https:\/\/geopard.tech\/lv\/wp-json\/wp\/v2\/comments?post=11549"}],"version-history":[{"count":0,"href":"https:\/\/geopard.tech\/lv\/wp-json\/wp\/v2\/posts\/11549\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/geopard.tech\/lv\/wp-json\/wp\/v2\/media\/11556"}],"wp:attachment":[{"href":"https:\/\/geopard.tech\/lv\/wp-json\/wp\/v2\/media?parent=11549"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/geopard.tech\/lv\/wp-json\/wp\/v2\/categories?post=11549"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/geopard.tech\/lv\/wp-json\/wp\/v2\/tags?post=11549"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}