isiZulu
Ingabe Ama-Module weKhamera Angasebenza Ngaphandle Kwelensi? Ukushintsha Okuphumelelayo Ekuthwebuleni Izithombe Ngokomshini
Kwadalwa ngo 2025.11.21
Ngeminyaka engamashumi, sithathe lokhu njengokujwayelekile ukuthi amakhamera adinga amalensi—njengoba sithatha ukuthi izimoto zidinga amajini noma amafoni adinga izikrini. Amalensi isikhathi eside abe "amehlo" kaamamojula ekhamera, ukugoba ukukhanya ukuze kugxilwe izithombe ezicacile kumasensori. Kodwa kwenzekani uma le mfundiso eyisisekelo ingasasebenzi? Namuhla, intuthuko ekwenzeni izithombe, ama-algorithms e-AI, kanye nokwakhiwa kwemicro kubonisa ukuthi ama-module wekhamera angasebenza ngaphandle kwezibuko—evula iminyango emisha kumadivayisi amancane, aphansi, futhi anokusebenza okuningi aphinde ahlukanise lokho ubuchwepheshe bezithombe bungakwenza.
I-Lensless Revolution: Indlela Yokusebenza Kwayo Yonke
Amamojula ekhamera angenalensi awasuswa nje ilensi—aphinde acabangele inqubo yonke yokuthwebula. Amakhamera ajwayelekile athembele kumalensi we-optical ukuze aguqule ukukhanya futhi akhe isithombe esiqondile kum sensor. Izinhlelo ezingekho nelensi zishintsha le msebenzi wokugxila ngokomzimba nge "computational focusing": ziqopha idatha yokukhanya eluhlaza ngezakhiwo ze-optical ezihlukile futhi zisebenzisa ama-algorithms ukuze ziqophe izithombe ezicacile, ezisebenzisekayo. Nansi imishini emithathu yokuphumelela eyenza lokhu kungenzeke:
1. Ukufaka Isithombe seMaski Ekwenziwe Ngendlela: Ukukhanyisa Okushintshashintshayo
Umsebenzi ophawulekayo kwezobuchwepheshe obungenalensi uvela kubacwaningi eNyuvesi yaseNanjing yeSayensi neTekhnoloji, abathuthukise uhlelo lwe-Lensless Imaging with a Programmable Fresnel Zone Aperture (LIP). Esikhundleni se-lensi, i-LIP isebenzisa imaski ehlelekayo ekhombisa ama-Fresnel Zone Apertures (FZA) aguquguqukayo—ama-patterns aguqula ukukhanya ukuze abambe ulwazi lwezindawo nezinga.
Uhlelo lusebenza ezinyathelweni ezimbili ezisemqoka: kuqala, i-mask eprogrammable ishiftela i-FZA ukuze iqoqe amaphuzu amaningi edatha ye-sub-aperture light field. Bese, i-algorithm ye-parallel fusion ihlanganisa la maphuzu edatha emkhakheni wezwi ukuze ibuyisele izithombe eziphezulu. Umphumela? Ukukhuphuka kwe-resolution ngo-2.5x kanye nokuthuthukiswa kwe-signal-to-noise ratio ngo-3 dB uma kuqhathaniswa nezinhlelo ezijwayelekile ze-static mask lensless. Emodenini ye-dynamic, ifinyelela ku-15 fps—ishesha ngokwanele ukuze kuqondwe izenzo ngesikhathi sangempela nokuxhumana komuntu-nekhompyutha—ngenkathi yehlisa usayizi we-module yekhamera ngo-90%.
2. Ukukhanya Ukubonisa Imifanekiso: Ingilazi njenge "Lens Engabonakali"
University of Utah’s Rajesh Menon uthathe indlela ehlukile: esebenzisa ukukhanya okubonakalayo ngaphakathi kwepulasitiki ukuze kuthathelwe indawo amalensi ajwayelekile. Iningi lokukhanya lidlula kupulasitiki, kodwa ingxenye encane ibuyisela emuva ezindaweni zayo zangaphakathi. Iqembu likaMenon lishayele i-CMOS sensor emaphethelweni epaneli ye-acrylic futhi laphendula eminye imiphetho ye-panel nge-tape ebonakalayo ukuze ibambe leli khanya eliphumayo.
Lapho ukukhanya kuthinta ingilazi, umshini wokuhlola uthola izimpawu ezibuyisiwe, futhi ama-algorithms okufunda ngomshini ahumusha le datha abe yimifanekiso. Ubuhlakani balolu hlelo bukhona ekwakheni kwayo okulula: ingilazi uqobo isebenza njengento yokubona, ikhipha isidingo sokuba nezibuko ezijolile. Ngenkathi imifanekiso eluhlaza ibonakala ingacacile emehlweni abantu, contains data eyanele ukuze amakompyutha akhiphe ulwazi olubalulekile—kulungele izinhlelo lapho imishini, hhayi abantu, "ababukeli."
3. AmaMicrolens Arrays: Abalaleli Bokukhanya Abancane
Abacwaningi eNyuvesi yaseCalifornia, eDavis, bakhe imodyuli engenalensi usebenzisa i-microlens array encane. Ngokwehlukana nelensi elikhulu elilodwa, le array isebenzisa ama-microlens angu-37 amancane (ane-12mm ububanzi) ukuze athathe ukukhanya ezikhungweni eziningi. I-microlens ngayinye isebenza njengendawo yokubuka eyodwa, iqoqela ulwazi lobukhulu olwakhiwa ama-algorithms e-AI abe izithombe ze-3D ngesikhathi sangempela.
Le teknoloji ixazulula umkhawulo omkhulu wamakhamera e-3D ajwayelekile: isebenza ngokuqhafaza okukodwa futhi igwema ukuhlela okunzima. I-array elula, eshintshashintshayo ilungele ama-robot, ukuhlolwa kwezimboni, kanye nezinhlelo ze-VR/AR—lapho usayizi nesivinini kubaluleke kakhulu kunezithombe eziphelele zokuphila.
Izicelo Zangempela: Lapho Amakhamera Anganemibono Khanya
Amamojula ekhamera angenalensi awawona ama-experiment ase-lab kuphela—sevele athola izinzuzo ezisebenzayo emikhakheni ehlukene, aqhutshwa yizinzuzo zawo ezinkulu: usayizi omncane, izindleko eziphansi, nokuhlala isikhathi eside. Nansi imikhakha ethuthukisiwe:
VR/AR kanye neTekhnoloji Yokugqoka
Ibhodlela elikhulu kwi-VR/AR amadivayisi yisikhala—ukwengeza ikhamera yendabuko yokulandela amehlo noma ukulawula izenzo kwandisa ubukhulu bezinhlangothi. Izimodyuli ezinganemibono zixazulula lokhu: Uhlelo lwe-Menon olwakhiwe ngengilazi luhambisana kahle ne-VR/AR lenses ukuze kulandeleke ukuhamba kwamehlo, kanti ukuncipha kwesisindo kwe-LIP module yeNyuvesi yaseNanjing okungu-90% kwenza kube kuhle kakhulu kuma-wearables alula. Lezi zimodyuli zengeza amandla okuthwebula ngaphandle kokuphula induduzo noma ukuklama.
Imaging yezokwelapha
Izinsimbi zokuhlola ezijwayelekile zisebenzisa ama-lenses amade, aqinile angase abe nokungakhululeki kubaguli. Imodyuli engena-lens ivumela ama-endoscopes aphansi kakhulu, aguquguqukayo angakwazi ukuhamba ezindaweni ezicindezelwe emzimbeni. Usayizi wabo omncane uphinde unciphise ingozi yokulimala kwezicubu, ngenkathi ukwakhiwa kwekhompyutha kugcina ukujula kwesithombe okudingekayo odokotela ukuze benze izinqumo ezichanile.
Ubumfihlo kanye Nokubhekwa
Amakhamera angenalensi anikeza inzuzo yokufihla: angahlanganiswa ezindongeni, ezindongeni, noma ezintweni ezijwayelekile ngaphandle kokutholwa. Amabhendi afana neHikvision aqale "amakhamera angabonakali" asebenzisa ubuchwepheshe bangenalensi, abahlanganisa ezindaweni ngenkathi bethwebula ukuhamba nokusebenza. Ukuhlala kwabo—ngaphandle kwezinto eziphukile zelesi—kuphinde kwenze kube kuhle ezimeni ezinzima zangaphandle.
Izimoto kanye Nezobuchwepheshe BeRobot
Izimoto ezizishayelayo kanye namarobhothi adinga izinhlelo zokuthwebula ezincane, ezithembekile ukuze zikwazi ukuhamba. Imodyuli ezinganeni zifaka ezindaweni ezicindezelwe kumashalofu ezimoto noma ezandleni zamabhothi, kanti ububanzi bazo obungapheli (umphumela ohlangene nokungabi nokugxila komzimba) busiza ekutholeni izinto ezikude ngezinga elihlukahlukene. Imodyuli ze-microlens ezikwazi ukuhlinzeka nge-3D ziwusizo ikakhulukazi ekuphathweni kwamabhothi, zivumela imishini "ukubona" ukuma kwezinto ezithwalayo.
Ukukhula Kwezimakethe: Izibalo Emva Kwenkululeko
Imaketheka ye-khamera engenalensi iyakhula njengoba lezi zinhlelo zithola ukuthandwa. Ngo-2020, usayizi wemakethe yomhlaba wonke wawungu-25 billion, futhi kulindeleke ukuthi ifinyelele ku-60 billion ngo-2025—ikhula nge-CAGR engaphezu kuka-18%. E-China kuphela, imakethe ilindeleke ukuthi ifinyelele ku-$21 billion (150 billion RMB) ngo-2025, iqhutshwa ukufunwa kwezinto zokusetshenziswa zabathengi nezinsiza zezokwelapha.
Abadlali abakhulu njenge-Hitachi, i-Teledyne Princeton Instruments, kanye ne-Huawei batshala imali enkulu kule teknoloji. Ngisho nezinkampani ezinkulu zekhono lokuthwebula njenge-Canon ne-Sony zihlola imiklamo enganemibono ukuze zihlale ziqhudelana emakethe ye-wearable ne-IoT. Umgibeli? Izindleko: ukususa imibono kukhipha enye yezingxenye ezibizayo kakhulu zamamojula okuthwebula, okwenza ukuthwebula kube nokufinyelela kumadivayisi amaningi.
Izinselelo Nezindlela Ezizayo
Amamojula ekhamera angenalensi awaphelele—kusukela manje. Abhekene nezinselelo ezintathu ezibalulekile ezibhekene nocwaningo olukhuluma ngazo:
Okokuqala, ukusebenza kokukhanya okuphansi. Ngaphandle kwelensi yokugxila ekukhanyeni, izinhlelo ezingelensi ziba nzima ezimeni zokukhanya okuphansi, okuholela ezithombeni ezinezwi. Intuthuko yakamuva ku-AI yokususa umsindo, njengendlela ye-EPFL ehlanganisa ukukhanya kwangaphandle, ithuthukisa ukusebenza, kodwa kudingeka umsebenzi owengeziwe ezindaweni zokukhanya okuphansi njengezokubhekwa ebusuku.
Okwesibili, imikhawulo ye-high-resolution. Ngenkathi ubuchwepheshe be-LIP buphumelela ekutholeni izinzuzo ezinhle zokuhluka, ama-modules angenalensi awakwazi ukufanisa imininingwane yamalensi e-DSLR aphezulu. Kubuciko bokuthwebula, lokhu kusho ukuthi kungenzeka bangakwazi ukufaka esikhundleni samakhamera ajwayelekile kungekudala—kodwa kumabonakude wemishini kanye nezithombe eziyisisekelo, ukuhluka sekukwanele.
Okwesithathu, ubunzima be-algorithm. Ukuthwebula ngaphandle kwezibuko kusekelwe kumaprocessor anamandla ukuze kuqhutshekwe nezinhlelo zokwakha. Kuma-device anamandla aphansi afana nezinsiza ze-IoT, lokhu kungadla amabhethri. Amanethiwekhi e-neural ahloliwe kanye nezinsiza ezisebenza kahle kakhulu zixazulula lokhu, kodwa ukusebenza kahle kwamandla kuqhubeka nokuba yinto ebalulekile.
Ikusasa libukeka lihle, nokho. Njengoba ama-algorithms e-AI eba namandla kakhulu futhi ukwenziwa kwemicro kubiza kancane, ama-modules angenalensi azophumelela. Abacwaningi sebesiqalile ukuhlola izithombe eziningi—kuhlanganisa izinhlelo ezingalensi nezokuhlola i-polarization noma i-spectral ukuze kwenziwe ukuhlolwa kwezokwelapha nokuhlaziywa kwezinto. Siphinde sibona ukuhlanganiswa ne-5G, okuvumela ukuvuselelwa kwezithombe ngesikhathi sangempela kumaseva e-cloud esikhundleni sezinsiza zendawo.
Isiphetho: Ukuphela KweSikhathi SeLens?
Ngakho, ingabe amamojula ekhamera angasebenza ngaphandle kwesibuko? Impendulo iyavuma—futhi asevele adlula izinhlelo ezijwayelekile ezisekelwe kumasibuko ezindaweni ezibalulekile. Ubuchwepheshe obungenasibuko abukhona nje kuphela njengokuhlukile; kuyashintsha indlela esicabanga ngayo, kugxilwe emsebenzini, usayizi, kanye nezindleko kunokwethembeka kwesithombe okuphelele.
Kubathengi, lokhu kusho amadivayisi amancane, aphumelelayo anemifanekiso efakiwe—kusukela kumawatch smart alandelela impilo nge-sensors encane ngaphandle kwezibuko kuya kumakhanda e-VR azizwa elula kakhulu. Kubhizinisi, kusho izixazululo zokufaka imifanekiso ezihambisana nalapho amakhamera ajwayelekile engakwazi khona, kusukela ngaphakathi emzimbeni womuntu kuya ezindaweni ezicindezelayo zemoto ezizimele.
I-lens ayizukuphela—ukuthwebula izithombe ezisezingeni eliphezulu kanye ne-videography yobuchwepheshe kuzohlala kuncike kuma-lens anembile eminyakeni ezayo. Kodwa-ke, kumabhiliyoni wezinsiza zokuthwebula ezidinga izithombe ezingezona ezemuseum, ama-modules angena-lens ayathatha indawo. Njengoba ukwakhiwa kwezithombe okuhlanganisiwe kuqhubeka nokuthuthuka, sizosuka masinyane sibuze "Ingabe amakhamera angasebenza ngaphandle kwama-lens?" bese siqala ukuzibuza ukuthi kungani sasingakaze sibe nama-lens ekuqaleni.
Uxhumane
Sicela uxhumane nathi uhambele
WhatsApp
WeChat