引導式測驗 · 中等 等級
低光图像增强及其局限性测验
測試亮度與證據、雜訊放大、色調曲線、連拍對齊、細節遺失、任務評估和保留原件。
問題 1 的 8
Why can simply brightening a dark image reveal speckle along with objects?
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。
- Why can simply brightening a dark image reveal speckle along with objects?
- It adds photons to the stored file
- Sensor noise disappears when gain rises
- The operation amplifies recorded noise as well as useful signal
- Brightness verifies every recovered feature
- What limitation remains when a feature was never captured above sensor noise?
- Enhancement cannot directly recover that missing measurement
- A tone curve can measure unseen scene detail
- Denoising provides a new independent exposure
- A brighter screen recaptures the scene
- Which description fits the cited Zero-DCE approach?
- It needs a paired bright target for each training image
- It changes the physical aperture after capture
- It guarantees a forensic reconstruction
- It estimates image-specific adjustment curves with non-reference losses
- A burst-enhancement method ghosts a moving hand. What likely failed?
- A single-image adjustment requires burst alignment
- Frame alignment or merging could not handle the motion
- Motion cannot affect multi-frame processing
- The method captured only one exposure
- Why evaluate a low-light enhancer with a downstream text-recognition test?
- Visual brightness guarantees OCR accuracy
- A no-reference aesthetic score proves legibility
- Brighter appearance does not guarantee more correct task decisions
- One good preview represents every task
- A denoiser removes a faint line. Which failure does this illustrate?
- A real weak feature may be mistaken for noise
- Denoising always preserves every real faint feature
- Sharpening and denoising are identical
- A lost line must have been noise
- Why preserve the original alongside an enhanced scientific image?
- The original is always higher quality to view
- Enhancement is always perfectly reversible
- Documenting processing alone replaces source evidence
- Reviewers can separate recorded evidence from later processing
- A frame is too noisy for an important identification. What response fits the guide?
- Force the top prediction as certain
- Recapture if possible or report uncertainty
- Sharpen until a plausible identity appears
- Discard the source and keep only enhancement