Placeholder Image

字幕列表 影片播放

由 AI 自動生成
  • So have you ever wondered what it would be like to see through walls?


  • Well, researchers at Stanford University might have just cracked the code on this superpower this latest where it comes from the same team at Stanford that developed a wave based, non line of sight imaging system back in 2019.


  • This camera is capable of seeing around corners by bouncing a laser off the wall and analyzing the reflected light that returns to the sensor.


  • That 2000 and 19 goal was to improve obstacle imaging techniques such as light detection and ranging systems, or light art, which is most often used in ground underwater and aerial vehicles to better understand their immediate surroundings.


  • But one of the biggest challenges of light our systems has been trying to peer through hazy conditions like an autonomous vehicle trying to drive through fog or orbiting spacecraft.


  • Looking through the thick atmosphere of another planet, these types of conditions scatter the light way too much for light our systems, making it really tough to detect anything through it.


  • But now the Stanford team has gone one step further by creating come focal diffused tomography or C D.

    但現在斯坦福大學的團隊已經更進一步創造來焦距彌散斷層掃描或C D。

  • T which allows a camera to see right through what they call a scattering media.


  • Or they could just call it what it is.


  • It's a thin piece of phone.


  • C D.

    C D.

  • T is a process that uses a single photon avalanche style and pulsating laser systems to effectively see through a medium.


  • So what exactly does that mean?


  • Let me explain it this way.


  • With a short pulse of light, a laser illuminates a point on the surface.


  • From here, the light diffuses through the foam and onto the surface of your target object on the opposite side.


  • The light from the laser then bounces off the surface of the object and defuse this once again back through the phone and returns to the original point.


  • The laser was illuminated.


  • Researchers then used a processing algorithm, and they're collected data points to produce a hist a gram, which will show the photon count along the Y axis and the time along the X axis.

    然後,研究人員使用了一種處理算法,他們收集的數據點會產生一個hist a gram,它會顯示沿Y軸的光子數和沿X軸的時間。

  • So through these points, this helps give researchers their first glimpse of what's actually on the other side.


  • But that's just the first step.


  • In order to get a three D image, this process has to be done multiple times by scanning a syriza points on a grid through a process the team calls come focal, scanning these data points, then work together to effectively create a blurry map of the image.


  • Using this, researchers can then apply a specific set of algorithms and light filters to reconstruct the completed three D image of the target object In the best part about it is that although this might seem like a really, really long process, this team is actually able to completely reconstruct an image and as little as 0.3 seconds.

    利用這一點,研究人員就可以應用一組特定的算法和光濾鏡來重建目標物體的完成的三D影像 在最好的部分是,雖然這可能看起來是一個非常非常漫長的過程,但這個團隊實際上能夠完全重建一個影像,而且只要0.3秒。

  • So how can CDT help us in the future?


  • The research suggests that this form of non invasive three D imaging through a medium might just revolutionized the way we approach autonomous vehicle navigation in bad weather like fog and heavy snowstorms.


  • But if we think even bigger, this could probably also helped first responders handle emergencies like a firefighter and a burning building.


  • But until then, I'll settle for using this newfound power to be the hide and go seek master.


  • To learn more about the team's research from last year, check it out here.


  • Let us know if you like this video down the comments below and make sure to subscribe.


  • Thanks so much for watching, and I'll see you next time.


So have you ever wondered what it would be like to see through walls?


由 AI 自動生成

單字即點即查 點擊單字可以查詢單字解釋