Ntụle omimi Stereo
Ntụle omimi stereo na-eweghachite ebe ihe dị anya site n'ịtụle nlele igwefoto abụọ dị ntakịrị, dịka anya gị abụọ si eme.
Nchịkọta
It turns flat images into 3D distance maps that robots, cars, and phones rely on to understand space.
Ime miri emi
Ntụle omimi stereo na-eji igwefoto abụọ n'ebe dị anya (ihe ndabere). Otu ebe na ụwa ala na ubé dị iche iche kehoraizin ọnọdụ na aka ekpe na aka nri oyiyi, na mgbanwe a na-akpọ ọdịiche. Ihe ndị dị nso na-agbanwe nke ukwuu; ndị dị anya na-akpụ akpụ. A na-agbakọ omimi dị ka (focal long x baseline) / dị iche, ya mere omimi na nghọtahie nwere njikọ chiri anya. Akụkụ siri ike bụ ijikọ pikselụ n'etiti onyonyo abụọ a, ọkachasị na mgbidi dị larịị, usoro na-emegharịgharị, ma ọ bụ ihu na-egosipụta ebe ọtụtụ pikselụ yiri otu. Ụzọ kpochapụwo dị ka Semi-Global Matching scan n'akụkụ scanlines, ebe netwọk miri emi nke oge a dị ka PSMNet na RAFT-Stereo na-amụta atụmatụ bara ụba ma na-emeziwanye ọdịiche na-aga n'ihu, na-emepụta nnukwu omimi zuru oke ọbụna na mpaghara aghụghọ.
Nghọta nka nka
A na-ebu ụzọ mezie ihe oyiyi abụọ a ka isi ihe dabara adaba dabere n'otu ahịrị kwụ ọtọ, na-ebelata ọchụchọ ahụ n'otu akụkụ. A na-ewu olu ọnụ ahịa site n'ịnwale ndịiche nke onye ọ bụla na-eme ntuli aka maka pikselụ ọ bụla, na-atụ etu njirimara aka ekpe na aka nri si kwekọrịta. Netwọk na-ejikọta olu a na mgbanwe 3D ma ọ bụ mmelite na-emegharị ugboro ugboro, wee were nro-argmin maka ndịiche iji nweta nkenke nke sub-pixel. Mmekọrịta na-agbanwe agbanwe n'etiti ọdịiche na ịdị omimi pụtara na omimi dị anya na-eme mkpọtụ karịa omimi dị nso.
Mmetụta atụmatụ
Ọsọ na ọnụ ọgụgụ
Visual AI nwere ike megharịa nyocha, nchọpụta na mkpado ọrụ n'ọtụtụ.
Mee nhọrọ
Otu ndị na-emepụta ihe nwere ike imepụta echiche ngwa ngwa site na ngbanwe akwụkwọ ntuziaka ole na ole.
Team na usoro ọrụ
Ọrụ nwere ike iji onyonyo na akara vidiyo siri ike ịhazi.
Ọdịnihu nke Atụmatụ omimi Stereo
Na-atụ anya njikọ siri ike nke stereo na LiDAR, radar na akara monocular ka sistemụ na-emebi nke ọma mgbe otu ihe mmetụta dara. Ntugharị dabere na ndakọrịta na ọzụzụ na-elekọta onwe ya (ịmụta site na vidiyo raw na-enweghị omimi ala-eziokwu) na-ebelata mkpa maka data akpọrọ aha dị oke ọnụ. Ịrụ ọrụ na ngwaọrụ na-akawanye ngwa ngwa, na-etinye stereo oge na drones, iko AR, na robots dị ọnụ ala. Igwefoto mmemme na ụkpụrụ arụ ọrụ amụtara na-ekwe nkwa omimi a pụrụ ịdabere na ya ọbụlagodi na obere ọkụ, nhụsianya ngagharị na ihe nkiri enweghị textures na-emeri usoro nke taa.
Mmejuputa n'ezie n'ụwa
Sistemu ịnya ụgbọ ala na onye ọkwọ ụgbọ ala na-eji igwefoto stereo iji tụọ anya n'ụgbọ ala, ndị ji ụkwụ aga, na mgbochi maka breeki na idobe okporo ụzọ.
Ụlọ nkwakọba ihe na ígwè ọrụ ugbo na-arụ maapụ 3D iji ghọta ihe, zere ihe mgbochi, ma were mkpụrụ n'ime omimi ziri ezi.
Igwe isi AR/VR dị ka ngwaọrụ passthrough na-eme atụmatụ geometry ime ụlọ ka ihe mebere nke ọma na-anọdụ ala nke ọma n'elu ala.
Mars rovers (dịka ọmụmaatụ, nnọgidesi ike) na-eji igwefoto igodo stereo iji hazie ụzọ dị nchebe n'elu ala okwute na-enweghị GPS.
Ihe ize ndụ & okporo ụzọ nche
Ikike onyonyo na nkwenye nwere ike bụrụ ihe egwu dị n'iwu ma ọ bụrụ na edoghị anya.
Ọrụ nlereanya nwere ike ịdịgasị iche n'ofe ọkụ, igwe mmadụ, na gburugburu.
Enwere ike ghara ịhụ ihe dị mma ma ọ bụrụ na enyochaghị oke ntụkwasị obi.
Map mmejuputa
Kọwaa ụkpụrụ nnabata maka nkenke, icheta, na ụgwọ njehie.
Nwalee na data dabara na ọnọdụ mmepụta n'ezie.
Tinye nyocha mmadụ maka obere obi ike ma ọ bụ amụma mmetụta dị elu.
Sochie ihe nlere anya wee megharịa ka emechara mgbanwe igwefoto ma ọ bụ dataset.
Nọgide na-eme nchọpụta
Free newsletter
Get the daily AI briefing
Three verified AI stories every weekday morning, written in plain English. Free forever, no ads.
One email each weekday. Unsubscribe in one click. We never sell or share your address.
Test yourself
Take the Stereo Depth Estimation quiz
Instant feedback on every answer, and a shareable certificate with a verifiable ID once you pass a course.
Support free AI education. AI Understanding is a 501(c)(3) nonprofit — no ads, no paywall, ever. Make a donation
Ntuziaka na-esote
Ntụle omimi nke Marigold Diffusion
Ajụjụ a na-ajụkarị
What is Stereo Depth Estimation?
Ntụle omimi stereo na-eweghachite ebe ihe dị anya site n'ịtụle nlele igwefoto abụọ dị ntakịrị, dịka anya gị abụọ si eme. Ọ na-atụgharị onyonyo dị larịị ka ọ bụrụ maapụ anya 3D nke robots, ụgbọ ala, na ekwentị na-adabere na ya ịghọta oghere.
Kedu ihe bụ 'ọdịiche' na ọhụụ stereo?
Ọdịiche bụ ka isi ihe kwekọrọ na-aga n'ụzọ kwụ ọtọ n'etiti echiche abụọ ahụ emeziri; nnukwu ọdịiche pụtara na ihe ahụ dị nso.
Kedu ka omimi siri metụta ndịiche?
Omimi = (ntụkwasị ụkwụ n'ogologo x ntọala) / dị iche iche, ka ndịiche na-ebelata, omimi atụmatụ na-eto eto. Ihe ndị dị anya nwere ntakịrị ọdịiche.
Kedu ihe kpatara eji 'dozie' foto tupu ọ dabara?
Ndozigharị ihe onyonyo abụọ a ka etinyere n'otu ahịrị kwụ ọtọ, na-atụgharị ọchụchọ 2D ka ọ bụrụ ọchụchọ 1D ngwa ngwa.
Kedu ọnọdụ kacha sie ike maka stereo dakọtara?
Elu enweghị ederede ma ọ bụ ugboro ugboro na-eme ka ọtụtụ pikselụ yie otu, ya mere algọridim enweghị ike iji obi ike chọta egwuregwu ziri ezi.
Kedu ihe 'baseline' na-ezo aka?
Ihe ndabere bụ nkewa n'etiti etiti igwefoto abụọ; ntọala sara mbara na-eme ka izi ezi maka ihe ndị dị anya.