1. Gabatarwa

Na'urori na Auna Lokacin Tafiya (ToF) na Kamara sun kawo sauyi mai girma ga fahimtar 3D ta hanyar samar da bayanan zurfin kowane pixel ta hanyar hasken da ake kunna. Wannan takarda tana magance wani gibi mai mahimmanci: buƙatar ingantaccen tsarin siminti don hasashen aikin na'urar, fahimtar hadaddun abubuwan gani, da kuma jagorantar ƙirar kayan aiki kafin yin ƙirar jiki mai tsada. Marubutan sun ba da shawarar wata hanya wacce ta wuce ƙirar ƙira mai sauƙi don ɗaukar cikakkun hanyoyin mu'amalar gani na zahiri.

2. Ka'idojin Auna Lokacin Tafiya (ToF)

Na'urorin ToF suna auna nisa ta hanyar lissafin lokacin tafiya na haske. Manyan fasahohi guda biyu ne suka mamaye:

2.1 Auna Lokacin Tafiya Kai Tsaye (D-ToF)

Yana auna jinkirin lokacin ƙaramin bugun haske kai tsaye. Yana ba da daidaito mai girma amma yana fama da ƙarancin sigina zuwa amo (SNR) saboda buƙatar na'urorin lantarki masu saurin GHz da ƙananan lokutan haɗawa (misali, 10 ns don 1.5 m).

2.2 Auna Lokacin Tafiya na Tushen Haɗin Kai (C-ToF/P-ToF)

Hanyar da ta fi yadu a cikin na'urorin masu amfani. Tana amfani da hasken igiyoyin ruwa mai daidaitaccen girma (AMCW). Ana samun nisa daga canjin lokaci ($\phi$) tsakanin siginonin da aka fitar da waɗanda aka karɓa. Ana lissafin zurfin ($d$) kamar haka: $d = \frac{c \cdot \phi}{4 \pi f_{mod}}$, inda $c$ shine saurin haske kuma $f_{mod}$ shine mitar daidaitawa. Wannan hanyar tana aiki a cikin kewayon MHz, tana sauƙaƙe buƙatun lantarki amma tana haifar da shubuha a nisassun da suka wuce tsawon igiyar daidaitawa.

3. Hanyar Siminti da Ake Shawarwari

Babban gudummawar shine hanyar siminti wacce ta ɗauki tsawon hanyar gani a matsayin babban ma'auni don lissafin zurfin.

3.1 Hanyar Tsawon Hanyar Gani ta Tushen Binciken Hasken (Raytracing)

Maimakon yin siminti na siginonin lantarki, hanyar tana bin diddigin kowane haske daga tushen (misali, VCSEL), ta cikin wurin (ciki har da jujjuyawar da yawa, watsawa, da wucewa ta ciki), har zuwa cikin ruwan tabarau na na'urar. Ana lissafin jimillar tsawon hanyar gani (OPL) na kowane haske kamar haka $OPL = \int n(s) \, ds$, inda $n$ shine ma'auni na karyewar haske kuma $s$ shine hanyar lissafi. Wannan OPL yana da alaƙa kai tsaye da lokacin tafiya.

3.2 Aiwarta a cikin Zemax OpticStudio da Python

Ana yin siminti na yaduwar gani da tasirin ruwan tabarau (karkacewa, ɓarna) a cikin Zemax OpticStudio. Sakamakon, ciki har da bayanan haske da OPL, ana fitar da su kuma a sarrafa su a cikin yanayin Python. Python yana kula da siffar wurin, kaddarorin kayan aiki, ƙirar na'urar (misali, amsawar pixel na PMD), da lissafin haɗin kai/ zurfin ƙarshe, yana ƙirƙirar tsarin aiki mai sassauƙa da faɗaɗawa.

3.3 Tasirin Gani da Ake Taimakawa

  • Kutsawa ta Hanyoyi da Yawa (MPI): Yana yin siminti na haskoki waɗanda ke billa tsakanin abubuwa da yawa kafin su isa na'urar, babban tushen kuskuren zurfin.
  • Kayan da Hasken ke Wucewa: Yana ƙirƙira watsawar ƙarƙashin ƙasa a cikin abubuwa kamar robobi ko fata.
  • Kurakuran Ruwan Tabarau: Ya haɗa da karkacewar ruwan tabarau na zahiri wanda ke shafa siginonin gani a cikin pixels.
  • Tushen Hasken da Yawa & Faɗaɗaɗɗu: Yana ƙirƙira daidai tsarin haske mai sarƙaƙiya, ba kawai tushen ma'ana ba.

Mahimman Ƙarfin Siminti

Jujjuyawar hanyoyi da yawa, Watsawar ƙarƙashin ƙasa, Karkacewar ruwan tabarau, Hasken da ya ƙunshi sarƙaƙiya

Kayan Aikin Aiwarta

Zemax OpticStudio (Gani), Python (Sarrafawa & Bincike)

4. Cikakkun Bayanai na Fasaha & Tushen Lissafi

Ana samun ƙimar zurfin $z$ don pixel na ToF na tushen haɗin kai daga canjin lokaci na samfuran haɗin kai guda huɗu ($A_0$, $A_1$, $A_2$, $A_3$), waɗanda galibi ake samu tare da canjin lokaci na digiri 90:

$$\phi = \arctan\left(\frac{A_3 - A_1}{A_0 - A_2}\right)$$

$$z = \frac{c}{4\pi f_{mod}} \phi$$

Simintin yana haifar da waɗannan samfuran haɗin kai $A_i$ ta hanyar haɗa ƙarfin gani mai shigowa, wanda aka daidaita ta hanyar jinkirin hanyar gani da aka yi siminti, a kan lokacin haɗawa na pixel. Ƙarfin gani don tarin hasken da ya kai pixel ana auna shi da ƙarfinsa da tsawon hanyarsa da aka yi siminti.

5. Sakamakon Gwaji & Nunawa

Takardar ta nuna hanyar a kan wani wurin gwaji na 3D mai sauƙi. Duk da yake ba a bayyana takamaiman ma'auni na kuskure a cikin abin da aka ba da ba, nunawar mai yiwuwa tana nuna:

  • Gaskiya ta ƙasa vs Taswirar Zurfin da aka yi Siminti: Kwatancen gani da ƙima wanda ke nuna daidaiton simintin wajen sake samar da ƙimar zurfin.
  • Hoton Abubuwan Ƙira: Hotuna suna nuna inda kutsawa ta hanyoyi da yawa (MPI) ke haifar da aunan zurfin kuskure (misali, kurakuran zurfi a kusurwoyi ko bayan abubuwan da haske ke wucewa).
  • Tasirin Karkacewar Ruwan Tabarau: Yana kwatanta yadda ruwan tabarau mara kyau ke shafa gefen zurfi da rage ingantaccen ƙuduri.

Ma'anar Ginshiƙi: Nuni mai nasara zai nuna babban haɗin kai tsakanin kurakuran zurfin da aka yi siminti da waɗanda aka auna daga na'urar ta jiki da ke kallon wuri ɗaya, yana tabbatar da ikon hasashen ƙirar don yanayin gani mai matsala.

6. Tsarin Bincike: Cikakkiyar Fahimta & Tsarin Ma'ana

Cikakkiyar Fahimta: Babban nasarar takardar ba sabon algorithm ba ne, amma canjin falsafa a cikin simintin ToF. Maimakon ɗaukar na'urar a matsayin akwati baƙar fata tare da aikin fitar da zurfin da ya dace, sun ƙirƙira ta a matsayin tsarin gani na zahiri da farko. Hanyar "tsawon hanyar gani a matsayin babban ma'auni" tana tilasta simintin ya mutunta dokokin gani na lissafi, yana mai da shi kayan aiki na farko maimakon ƙirar da aka dace. Wannan yana kama da sauyi daga sarrafa hoto na zahiri zuwa ƙirar tushen kimiyyar lissafi a cikin zane na kwamfuta.

Tsarin Ma'ana: Hujjar marubutan tana da tsari: 1) Gano cewa tasirin gani na zahiri (MPI, watsawa) sune manyan masu iyakance daidaiton ToF. 2) Yi hujjar cewa ƙirar lantarki ko sauƙaƙaƙƙun gani na yanzu ba za su iya ɗaukar waɗannan ba. 3) Ba da shawarar tsarin binciken haske a matsayin mafi ƙarancin rikitarwa wanda zai iya ɗaukar su. 4) Tabbatarwa ta hanyar nuna cewa yana iya yin siminti da ainihin tasirin da ke addabar na'urori na zahiri. Ma'anar tana da ƙarfi saboda tana kai hari ga matsalar a tushenta.

7. Ƙarfafawa, Kurakurai & Shawarwari masu Amfani

Ƙarfafawa:

  • Ikon Hasashen Abubuwan Ƙira masu Muni: Wannan shine siffarsa mai kashewa. Ta hanyar ɗaukar MPI da watsawa, zai iya hasashen kurakuran zurfi a cikin wurare masu sarƙaƙiya (misali, kusurwoyin cikin gida, cikin motoci) kafin gina na'urar, yana adana miliyoyin a cikin sake maimaitawa na ƙira.
  • Rashin Dogaro da Kayan Aiki: Yin amfani da Zemax da Python yana sa ya zama mai sauƙi. Ana iya canza ra'ayin zuwa Blender/Cycles ko NVIDIA OptiX don saurin binciken haske mai saurin GPU.
  • Tushen Horon AI: Zai iya samar da babban, cikakken bayanan da aka yiwa lakabi na taswirorin zurfi tare da taswirorin kuskure masu dacewa—zinariyar ƙura don horar da ƙirar AI don gyara kurakuran ToF, kama da yadda hanyoyin sadarwa irin na CycleGAN suke koyon fassarar yanki.

Kurakurai masu Kyau & Rashin Bayyanawa:

  • Akwatin Baƙi na Farashin Lissafi: Takardar tana shiru sosai akan lokacin gudu. Binciken haske na wurare masu sarƙaƙiya tare da miliyoyin haskoki a kowane firam yana da sauri sosai. Ba tare da ingantaccen ingantawa ko kiyasi ba, wannan kayan aikin bincike ne, ba kayan aikin ƙira ba.
  • Ƙirar Amo ba ta da Ma'ana: Sun ambaci amo amma ba su haɗa cikakkiyar ƙirar amo na na'urar ba (amo na harbi, amo na karantawa, igiyar ruwa mai duhu). Wannan babban gajeriyar hanya ce; amo shine abin da ke sa MPI da matsalolin ƙarancin sigina su zama bala'i.
  • Tabbatarwa Ba ta da Nauyi: "Wurin gwaji na 3D mai sauƙi" bai isa ba. Ina kwatancen ƙididdiga da babban ma'auni kamar na'urar sikanar Laser don wuri mai daidaitaccen tsari, mai sarƙaƙiya?

Shawarwari masu Amfani:

  1. Ga Masu Bincike: Yi amfani da wannan tsarin don samar da "taswirorin kuskure" don sabbin wurare. Mayar da hankali kan amfani da sakamakon don horar da hanyoyin sadarwar jijiyoyi masu sauƙi waɗanda za su iya gyara waɗannan kurakuran a lokacin gaskiya akan na'urar, suna matsar da aiki mai nauyi daga lokacin siminti zuwa lokacin ƙididdiga.
  2. Ga Injiniyoyi: Haɗa sauƙaƙaƙƙen sigar wannan ƙirar, mai iya aiki a lokacin gaskiya, cikin software na ƙirar na'urar. Yi amfani da shi don gudanar da saurin binciken "menene idan" akan ƙirar ruwan tabarau da tsarin haske don rage yuwuwar kamuwa da MPI tun daga farko.
  3. Takarda ta Gaba da za a Rubuta: "Na'urar Siminti na ToF mai Bambancewa don Ingantaccen Ingantawa har ƙarshe." Haɗa wannan hanyar binciken haske tare da dabarun zane mai bambancewa. Wannan zai ba ku damar ba kawai yin siminti na kurakurai ba, amma don inganta kayan aikin na'urar (siffar ruwan tabarau, tsarin daidaitawa) kai tsaye ta hanyar komawa baya ta cikin siminti don rage aikin asarar kuskuren zurfin.

8. Hasashen Aikace-aikace & Hanyoyin Gaba

Tsarin simintin yana buɗe kofofi a cikin mahimman fagage da yawa:

  • LiDAR/ToF na Motoci: Yin siminti na fahimtar zurfi a cikin yanayi mara kyau (ruwan sama, hazo, kutsawar motoci da yawa) don haɓaka ingantattun algorithms don motoci masu cin gashin kansu.
  • Binciken Halittu & Kula da Lafiya: Ƙirƙirar mu'amalar haske tare da naman ɗan adam don aikace-aikace kamar hoton jijiyoyin jini, saka idanu kan numfashi, ko gano bugun zuciya ba tare da taɓawa ba, inda watsawar ƙarƙashin ƙasa ta mamaye.
  • Ƙarfafa/Ƙirar Gaskiya (AR/VR): Inganta na'urorin bin diddigin ciki-waje don aiki a cikin wurare daban-daban na gida cike da jujjuyawar hanyoyi da yawa.
  • Ma'aunin Masana'antu: Ƙirƙirar tsarin ToF don daidaitaccen ma'auni na sassa na masana'antu masu sarƙaƙiya, masu sheki, ko masu wucewar haske.

Bincike na Gaba dole ne ya mayar da hankali kan gina gada zuwa aikin lokacin gaskiya ta hanyar samfurin mahimmanci (ba da fifiko ga haskoki masu yuwuwar haifar da MPI) da ƙirar ƙira masu rage kimiyyar lissafi, da kuma kan haɗin kai mai ƙarfi tare da cikakkiyar simintin amo na lantarki.

9. Nassoshi

  1. Baumgart, M., Druml, N., & Consani, C. (2018). Hanyar da ke ba da damar yin siminti da cikakken bincike na tasirin gani a cikin na'urori na auna lokacin tafiya na kamara. The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, XLII-2, 83-90.
  2. Lange, R. (2000). Auna nisa na 3D Time-of-Flight tare da na'urorin hoto na musamman a cikin fasahar CMOS/CCD. PhD Thesis, Jami'ar Siegen.
  3. Schwarte, R., et al. (1997). Sabuwar na'urar haɗawa da haɗin kai ta lantarki-gani: wurare da aikace-aikace na na'urar haɗawa ta gani (PMD). Proc. SPIE, 3100.
  4. Jarabo, A., et al. (2017). Tsarin don Zane mai Wucewa. ACM Computing Surveys. (Tushen waje akan hoton wucewa)
  5. Remondino, F., & Stoppa, D. (Eds.). (2013). Kyamarorin Taswira na Nisa na TOF. Springer. (Littafi mai mahimmanci na waje akan ToF)
  6. Zhu, J.Y., et al. (2017). Fassarar Hoton zuwa Hoton mara Haɗin gwiwa ta amfani da Hanyoyin Sadarwa masu Adawa da Zagayowar. IEEE ICCV. (Nassoshi na CycleGAN don ra'ayin gyaran kuskuren na AI)