Biometric kuyerwa kwemuviri uye maverengero ane chekuita nehunhu hwevanhu. Biometric authentication (kana realistic authentication) inoshandiswa musainzi yekombuta senzira yekuzivikanwa uye kutonga kwekuwana. Inoshandiswawo kuona vanhu vari mumapoka ari pasi pekutariswa.

Biometric identifiers ndiwo akasiyana, anoyerika maitiro anoshandiswa kuisa nekutsanangura vanhu. Biometric identifiers inowanzoiswa muzvikamu sehunhu hwemuviri hune hukama nechimiro chemuviri. Mienzaniso inosanganisira, asi haina kugumira kune zvigunwe, tsinga dzemuchindwe, kuzivikanwa kwechiso, DNA, kudhindwa kwemichindwe, geometry yemaoko, kuzivikanwa kweiris, retina, uye kunhuwa/kunhuwirira.

Biometric identification tekinoroji inosanganisira sainzi yekombuta, optics uye acoustics uye mamwe sainzi emuviri, biological sainzi, biosensors uye biostatistics nheyo, kuchengetedza tekinoroji, uye artificial intelligence tekinoroji uye mamwe akawanda ekutanga sainzi uye hunyanzvi hwekushandisa. Iyo yakazara multidisciplinary tekinoroji mhinduro.

Mumakore achangopfuura, nekuvandudzwa kwehungwaru hwekugadzira, tekinoroji yekuzivikanwa kwebiometric yakawedzera kukura. Parizvino, tekinoroji yekuziva kumeso ndiyo inonyanya kumiririra biometric.

Kuzivikanwa kwechiso

Maitiro ekuzivikanwa kwechiso anosanganisira kuunganidza chiso, kuona chiso, kubviswa kwechiso uye kuzivikanwa kwekufananidza chiso. Iyo yekuziva kumeso maitiro inoshandisa akasiyana matekinoroji akadai seAdaBoos algorithm, convolutional neural network uye inotsigira vector muchina mukudzidza muchina.

Nzira yekuzivikanwa kwechiso

Parizvino, matambudziko echinyakare ekuzivikanwa kwechiso anosanganisira kutenderera kwechiso, kuvharika, kufanana, nezvimwe zvakagadziridzwa zvakanyanya, izvo zvinovandudza zvakanyanya kurongeka kwekuzivikanwa kwechiso. 2D face, 3D face, multi-spectral face Imwe neimwe modhi ine akasiyana ekutora ekugadzirisa mamiriro, data kuchengetedza dhigirii uye kuvanzika senitivity, nezvimwewo, uye kuwedzerwa kwekudzidza kwakadzama kwedata hombe kunoita kuti 3D yekuzivikanwa kumeso algorithm iwedzere kukanganisa kwe2D fungidziro, Inogona kukurumidza kuziva kuzivikanwa kwemunhu, izvo zvaunza imwe budiriro yekushandiswa kwekuzivikanwa kwechiso chemativi maviri.

Panguva imwecheteyo, tekinoroji yekuona biometric parizvino iri kushandiswa seyakakosha tekinoroji yekuvandudza kuchengetedzeka kwekuzivikanwa kwechiso, iyo inogona kunyatso ramba hutsotsi hwekunyepedzera senge mafoto, mavhidhiyo, mhando dze3D, uye masks eprosthetic, uye nekuzvimiririra kuti ndiani. vashandisi vanoshanda. Parizvino, nekukurumidza kuvandudza tekinoroji yekuzivikanwa kwechiso, akawanda maapplication akadai seakangwara maturusi, mari yepamhepo, uye kubhadhara kumeso zvave kuwedzera mukurumbira, zvichiunza kukurumidza uye nyore kuhupenyu hwemunhu wese nebasa.

Palmprint kuzivikanwa

Palmprint recognition imhando nyowani yebiometric recognition tekinoroji, iyo inoshandisa palmprint yemuviri wemunhu sechinhu chinotarirwa, uye inounganidza ruzivo rwebiology kuburikidza nemultispectral imaging tekinoroji. Multi-spectral palmprint recognition inogona kutariswa semuenzaniso weiyo biometric recognition tekinoroji inosanganisa akawanda-modality uye akawanda anotangwa maitiro. Iyi tekinoroji nyowani inosanganisa izvo zvitatu zvinozivikanwa zveganda spectrum, palm print uye tsinga tsinga kupa ruzivo rwakawanda panguva imwe chete uye kuwedzera kusiyanisa kwezvinhu zvinonangwa.

Gore rino, tekinoroji yeAmazon yekucherechedzwa kwemichindwe, kodhi-inonzi Orville, yatanga kuyedza. Iyo scanner inotanga kuwana seti ye infrared polarized yepakutanga mifananidzo, inotarisa kune ekunze maficha emuchindwe, semitsetse uye kupeta; paunenge uchiwana chikamu chechipiri chemifananidzo yakasvibiswa zvakare, inotarisa pakugadzirwa kwemichindwe uye zvinhu zvemukati, zvakadai setsinga, mapfupa, zvinyoro zvinyoro, nezvimwewo. Mifananidzo yakasvibirira inotanga kugadziriswa kuti ipe seti yemifananidzo ine maoko. Iyi mifananidzo yakavhenekerwa zvakanaka, yakatariswa, uye inoratidza chanza mune chaiyo yakatarisa, mune chaiyo pose, uye yakanyorwa seruboshwe kana rudyi.

Parizvino, tekinoroji yeAmazon yekucherekedza palmprint inogona kuonesa chitupa chemunhu uye kubhadhara kwakazara mu300 milliseconds chete, uye haidi kuti vashandisi vaise maoko avo pane yekuvheneka mudziyo, kungovheyesa uye kuongorora pasina kubata. Iko kukundikana kweiyo tekinoroji ndeye 0.0001%. Panguva imwecheteyo, kucherechedzwa kwepalmprint ndeyekusimbisa kaviri mudanho rekutanga - kekutanga kuwana maitiro ekunze, uye kechipiri kuwana maitiro emukati esangano. Kuenzaniswa nehumwe biometric tekinoroji maererano nekuchengetedza, yakavandudzwa.

Pamusoro pezviri pamusoro apa biometric maficha, iris recognition tekinoroji iri zvakare kufarirwa. Iyo yenhema yekuzivikanwa mwero wekuzivikanwa kweiris yakaderera se1/1000000. Iyo inonyanya kushandisa hunhu hweiris hupenyu invariance uye mutsauko kuti uone kuzivikanwa.

Parizvino, kubvumirana muindasitiri ndeyekuti kucherechedzwa kweimwe modhi kune zvipingamupinyi mune zvese kucherechedzwa kwekuita uye chengetedzo, uye multi-modal fusion ndiyo yakakosha budiriro mukuzivikanwa kwechiso uye kunyangwe biometric kucherechedzwa-kwete chete kuburikidza ne-multi-factor Iyo nzira. kuvandudza kurongeka kwekuzivikanwa kunogonawo kuvandudza kuchinjika kwechiitiko uye kuchengetedzeka kwekuvanzika kwebiometric tekinoroji kune imwe nhanho. Kana ichienzaniswa neyechinyakare imwe-modhi algorithm, inogona kuzadzisa zvirinani mwero wezvemari-yenhema yekuzivikanwa mwero (yakaderera seimwe pamamiriyoni gumi), inovawo maitiro makuru ekuvandudzwa kwekuzivikanwa kwebiometric.

Multimodal biometric system

Multimodal biometric masisitimu anoshandisa masensa akawanda kana biometrics kuti akunde kuganhurirwa kweunimodal biometric masisitimu.Semuyenzaniso iris recognition masisitimu anogona kukanganiswa nekuchembera irises uye kuzivikanwa kwemagetsi zvigunwe kunogona kuipisisa nekusakara kana kucheka mafingerprints. Nepo unimodal biometric masisitimu akaganhurirwa nekuvimbika kwechiziviso chavo, hazvigoneke kuti akati wandei unimodal masisitimu achatambura kubva kune zvimwe chete zvisingakwanisi. Multimodal biometric masisitimu anogona kuwana maseti eruzivo kubva kune imwechete marker (kureva, akawanda mifananidzo yeiris, kana scan yemunwe mumwechete) kana ruzivo kubva kune akasiyana biometrics (inoda mafingerprint scans uye, uchishandisa izwi rekuziva, yekutaura passcode).

Multimodal biometric masisitimu anogona kusanganisa aya unimodal masisitimu akateedzana, panguva imwe chete, musanganiswa waro, kana munhevedzano, ayo anoreva kutevedzana, kufanana, hierarchical uye serial yekubatanidza modhi, zvichiteerana.

CHANCCTVyakagadzira nhevedzano yebiometric lensesyekuzivikanwa kwechiso, kucherechedzwa kwepalmprint pamwe nekuzivikanwa kweminwe uye kuzivikanwa kweiris.Semuenzaniso CH3659A i4k low distortion lens iyo yakagadzirirwa 1/1.8'' sensors. Iyo inoratidzira ese magirazi uye compact magadzirirwo ane chete 11.95mm TTL. Inotora 44 madhigirii yakatwasuka yekuona. Iyi lenzi yakanakira kuzivikanwa kwepalmprint.