Quiz połączony z przewodnikiem · Średni Poziom

Quiz dotyczący wskaźników jakości obrazu PSNR i SSIM

Porównaj PSNR i SSIM, oblicz, co każdy z nich reprezentuje i zidentyfikuj zakres danych, wyrównanie, agregację i ograniczenia wynikające z oceny człowieka.

Powiązane ścieżki prowadząceMetryki jakości obrazu Psnr Ssim
Pytanie 1 z 8

What error quantity underlies the standard PSNR calculation?

Podgląd każdego pytania8 pytania

Odpowiedzi i wyjaśnienia pojawiają się w miarę rozwiązywania quizu.

  1. What error quantity underlies the standard PSNR calculation?
    • Średni bezwzględny błąd piksela
    • The maximum absolute pixel error
    • Średni kwadratowy błąd piksela
    • The fraction of pixels that match exactly
  2. What does higher PSNR indicate under the same data-range convention?
    • Higher annotation agreement
    • Lower mean squared reconstruction error
    • More accurate object detection
    • Better color calibration in all cases
  3. Which local properties does SSIM compare?
    • Only the global mean pixel value
    • Only independent squared differences at each pixel
    • Only geometric edge coordinates
    • Luminance, contrast and structure
  4. Why must PSNR's data range be explicit when comparing implementations?
    • Its peak value depends on the image representation
    • It determines the SSIM window size
    • It removes spatial alignment needs
    • It decides which image is the reference
  5. Why can a shifted edge reduce PSNR even when it looks visually similar?
    • Matching resolution makes spatial alignment irrelevant
    • PSNR automatically corrects every spatial offset
    • Pixelwise differences penalize misalignment
    • PSNR compares object identity without considering pixel positions
  6. Does a high SSIM guarantee that an image is useful for a specific task?
    • Yes; SSIM validates every downstream decision
    • No; task-specific details may still be lost
    • Yes, if the image is grayscale
    • No; SSIM has no numerical meaning
  7. What should be held consistent for a fair PSNR/SSIM comparison?
    • Using the same nominal resolution is enough even if images are misaligned
    • Using the same file format fixes dynamic range differences
    • Averaging scores per image and over pixels always gives the same result
    • Alignment, range, crop, color handling and aggregation
  8. What does PSNR report when mean squared error is zero under its idealized formula?
    • An infinite value
    • Zero
    • Exactly one decibel
    • A value equal to the image's peak pixel intensity