Computer Vision Interview 20 essential Q&A Updated 2026
thresholding

Image Thresholding: 20 Essential Q&A

Global vs adaptive methods, Otsu, and when simple thresholding fails.

~10 min read 20 questions Beginner
OtsuadaptivebinaryOpenCV
1 What is image thresholding? ⚡ easy
Answer: Classifying pixels as foreground vs background by comparing intensity to one or more thresholds—produces binary or multi-label masks.
2 What is global thresholding? ⚡ easy
Answer: Single threshold T for the whole image: pixel → foreground if I > T (or < depending on type). Fast but fails under uneven lighting or overlapping histograms.
3 When use inverse binary threshold? ⚡ easy
Answer: When objects are darker than background (or you want white objects on black mask). Complement of standard binary THRESH_BINARY.
4 How does Otsu choose T? 📊 medium
Answer: Assumes roughly bimodal histogram; picks T that minimizes intra-class variance (equivalently maximizes between-class variance). Automatic, no manual T.
_, bw = cv2.threshold(gray, 0, 255, cv2.THRESH_BINARY + cv2.THRESH_OTSU)
5 When does Otsu fail? 📊 medium
Answer: Unimodal or flat histograms, uneven illumination, low contrast, or when foreground fraction is tiny—histogram may not have clear valleys.
6 What is adaptive mean threshold? 📊 medium
Answer: For each pixel, threshold = mean of local neighborhood − C. Handles varying illumination; needs block size larger than foreground features.
7 Adaptive threshold with Gaussian weights? 📊 medium
Answer: Local threshold from Gaussian-weighted mean instead of flat mean—smoother local estimates, slightly better on gradual shading.
8 What is block size in adaptive threshold? ⚡ easy
Answer: Odd window size defining local neighborhood. Too small: noisy mask; too large: loses detail near object boundaries.
9 What is Sauvola / Niblack? 🔥 hard
Answer: Local methods using mean and standard deviation to set threshold—good for document and degraded scans with uneven background.
10 Fix uneven lighting before threshold? ⚡ easy
Answer: Homomorphic filtering, background normalization, CLAHE on luminance, or large-kernel low-pass estimate of illumination to flatten.
11 Threshold colored objects? 📊 medium
Answer: Often convert to HSV and threshold H/S/V ranges (e.g. colored ball)—more robust than RGB for hue-based objects under some lighting.
12 What is multi-level thresholding? 📊 medium
Answer: Several thresholds to get multiple classes (e.g. tissue types). Extension of Otsu exists (multi-Otsu) but cost grows with levels.
13 How pick T without Otsu? 📊 medium
Answer: Manual inspection, ROC on validation set, entropy methods, or trial with domain constraints (known object brightness).
14 Common approach for scanned documents? ⚡ easy
Answer: Adaptive threshold or Sauvola-class; deskew/denoise first; morphology to clean speckles—DL methods also used for hard cases.
15 Why binary masks have holes / noise? ⚡ easy
Answer: Sensor noise, shadows, partial overlap of histograms—use morphology, median blur pre-threshold, or adaptive methods.
16 Apply morphology after thresholding? 📊 medium
Answer: Yes—opening removes pepper noise, closing fills small holes in foreground; preserves label if structuring element smaller than features.
17 Do deep nets replace thresholding? 📊 medium
Answer: For complex scenes, semantic segmentation wins; classical thresholding remains fast for controlled lighting, industrial vision, and documents.
18 What is soft thresholding (wavelets)? 🔥 hard
Answer: Shrinks coefficients toward zero—used in denoising, not classical image binarization. Mention only if interviewer asks denoising context.
19 Threshold on float vs uint8? ⚡ easy
Answer: Same logic but ensure consistent range ([0,1] vs 0–255). Always know your image dtype before comparing to T.
20 Typical order: blur → threshold → morph? 📊 medium
Answer: Often denoise/blur lightly → threshold → morphological cleanup—order depends on whether blur destroys thin structures.

Thresholding Cheat Sheet

Global
  • Single T
  • Otsu if bimodal
Adaptive
  • Local mean/Gauss
  • Block size + C
Docs
  • Sauvola family
  • Illumination first

💡 Pro tip: Say when global fails (lighting) and name one adaptive fix.

Full tutorial track

Go deeper with the matching tutorial chapter and code examples.