引導式測驗 · 中等 等級

影像去霧與去雨測驗

回顧霧霾散射與雨條紋、暗通道假設、擦除偽影、合成到真實的傳輸和任務驗證。

相關引導路徑影像去霧和去雨
問題 1 的 8

Which degradation is most directly addressed by dehazing?

預覽每個問題8 問題

當您參加測驗時,就會出現答案和解釋。

  1. Which degradation is most directly addressed by dehazing?
    • Loss of exact file integrity
    • A contrast-reducing atmospheric veil with distance
    • Wrong category labels in the dataset
    • Purely geometric camera rotation
  2. Why is removing rain streaks not the same as recovering scene detail hidden by an opaque droplet?
    • Every droplet is only a transparent streak
    • A model can measure exact hidden text
    • The sensor always saves an unobscured reference
    • The hidden detail may not be measured in that frame
  3. What kind of assumption underlies the dark channel prior method?
    • Statistics of local color channels in haze-free outdoor images
    • Every pixel has the same scene depth
    • Rain streaks always point vertically
    • All bright surfaces are physically black
  4. A deraining model deletes a thin wire. What likely went wrong?
    • The wire changed camera intrinsics
    • Atmospheric scattering alone removed the line
    • It treated a real line-like structure as rain
    • The file format cannot store narrow structures
  5. Why may success on synthetic rain not transfer to a wet windshield?
    • Synthetic streaks cover all real optics
    • Real droplets, optics and motion can differ from simulated streaks
    • Windshields cannot collect droplets
    • Simulation guarantees field generalization
  6. A restored landscape looks attractive. What else should a road system measure?
    • Downstream sign and obstacle errors in real weather
    • Only the aesthetic preference score
    • Only the restored image’s brightness
    • Only the method’s parameter count
  7. Why preserve the original weather-degraded image?
    • It makes the restoration mathematically invertible
    • It proves all hidden colors were exact
    • It eliminates uncertainty about occluded objects
    • It distinguishes captured evidence from estimated restoration
  8. When visibility remains too poor for a safety decision, what should the system do?
    • Treat restoration as certain perception
    • Sharpen until a hazard looks absent
    • Use a defined degraded-operation or fallback behavior
    • Assume low visibility means an empty road