1. Who is responsible for combining raw sensor data in image fusion at the most basic level?
(A) Feature level fusion
(B) Pixel level fusion
(C) Decision level fusion
(D) Context level fusion
2. Which level of image fusion combines extracted features such as edges or regions from different images?
(A) Pixel level fusion
(B) Data level fusion
(C) Feature level fusion
(D) Decision level fusion
3. Which level of fusion uses the outputs of classifiers from different sources?
(A) Feature level
(B) Pixel level
(C) Decision level
(D) Data level
4. Which method generally results in the highest spatial resolution in image fusion?
(A) Feature level
(B) Decision level
(C) Pixel level
(D) Object level
5. Which fusion level is most appropriate for medical image classification based on multiple imaging modalities?
(A) Pixel level
(B) Feature level
(C) Decision level
(D) Context level
6. Which of the following is an advantage of pixel-level image fusion?
(A) Robust against noise
(B) Preserves raw data
(C) Requires fewer computations
(D) Uses semantic interpretation
7. Which is the most computationally intensive image fusion level?
(A) Decision level
(B) Pixel level
(C) Object level
(D) Rule-based level
8. What is a primary disadvantage of decision-level fusion?
(A) Reduces semantic information
(B) Requires high-quality input
(C) Loss of detailed image information
(D) High memory consumption
9. Which technique uses transforms like wavelets in pixel-level image fusion?
(A) Discrete Fourier Transform
(B) Principal Component Analysis
(C) Discrete Wavelet Transform
(D) Linear Discriminant Analysis
10. What is the main advantage of feature-level fusion?
(A) Simple and fast
(B) Combines rich information
(C) Avoids classifier usage
(D) Works without feature extraction
11. In decision-level fusion, which type of model is commonly used?
(A) Convolutional Neural Networks
(B) Support Vector Machines
(C) Majority Voting
(D) Histogram Equalization
12. Which of the following is typically used in pixel-level fusion of multispectral images?
(A) Decision trees
(B) PCA
(C) Genetic algorithms
(D) Deep learning
13. What does pixel-level fusion directly combine?
(A) Final decisions
(B) Classification labels
(C) Raw intensity values
(D) Probability scores
14. Which level of fusion is best for applications needing interpretability?
(A) Pixel level
(B) Feature level
(C) Decision level
(D) Data level
15. Which type of fusion improves classification accuracy by combining predictions from multiple classifiers?
(A) Feature fusion
(B) Score fusion
(C) Decision fusion
(D) Kernel fusion
16. In feature-level fusion, what is the typical first step?
(A) Image normalization
(B) Image thresholding
(C) Feature extraction
(D) Image compression
17. What type of fusion would be most appropriate when input images have very different resolutions?
(A) Pixel level
(B) Feature level
(C) Decision level
(D) Histogram level
18. Which of these techniques is used for dimensionality reduction in feature-level fusion?
(A) PCA
(B) Median filter
(C) Morphological operations
(D) Gaussian smoothing
19. Which image fusion level is considered least sensitive to misregistration errors?
(A) Pixel level
(B) Feature level
(C) Decision level
(D) Hybrid level
20. Which of the following methods can be applied for pixel-level fusion in thermal and visible image fusion?
(A) Histogram specification
(B) DWT
(C) Otsu’s method
(D) Hough transform
21. What is the key benefit of decision-level fusion in surveillance applications?
(A) Real-time feature extraction
(B) Enhanced spatial resolution
(C) Increased classification robustness
(D) Data format unification
22. Which level of fusion is most affected by noise and misalignment?
(A) Feature level
(B) Pixel level
(C) Decision level
(D) Contextual level
23. Which method is commonly used to align images before pixel-level fusion?
(A) Histogram equalization
(B) Image registration
(C) Gaussian filtering
(D) Edge detection
24. Which fusion level best supports integration of heterogeneous data types?
(A) Pixel level
(B) Feature level
(C) Decision level
(D) Noise level
25. What type of learning model can be used for feature-level fusion?
(A) Linear regression
(B) Decision trees
(C) Neural networks
(D) K-means
26. Which of the following describes a major challenge in feature-level fusion?
(A) Low resolution
(B) Time synchronization
(C) Feature compatibility
(D) Sensor cost
27. What is the output of pixel-level image fusion?
(A) Final labels
(B) Class probabilities
(C) Combined image
(D) Confidence scores
28. Which level of fusion would be best for combining RGB and thermal images for object detection?
(A) Pixel level
(B) Feature level
(C) Decision level
(D) Sensor level
29. Which one is not a benefit of image fusion?
(A) Improved visualization
(B) Enhanced decision making
(C) Increased storage requirements
(D) Better spatial or spectral resolution
30. Which technique is not typically involved in pixel-level fusion?
(A) Intensity averaging
(B) Maximum selection
(C) Majority voting
(D) PCA
31. What is required before performing pixel-level fusion on multiple images?
(A) Feature extraction
(B) Image registration
(C) Thresholding
(D) Decision model training
32. Which method is useful for reducing redundancy in feature-level fusion?
(A) Clustering
(B) Wavelet transform
(C) PCA
(D) Dilation
33. Which decision-level fusion technique combines results based on probability distributions?
(A) Max rule
(B) Bayesian inference
(C) Logical AND
(D) Image subtraction
34. Which technique is often used to evaluate the effectiveness of image fusion?
(A) Entropy
(B) Laplacian filter
(C) Gradient descent
(D) Low-pass filtering
35. In which application is decision-level fusion commonly used?
(A) Remote sensing
(B) Image compression
(C) Histogram equalization
(D) Texture analysis
36. Which of the following uses raw pixel intensities from multiple sensors?
(A) Feature-level fusion
(B) Decision-level fusion
(C) Pixel-level fusion
(D) Context-aware fusion
37. Which is a common tool for feature extraction before fusion?
(A) DCT
(B) Edge detectors
(C) Histogram equalization
(D) Interpolation
38. Which metric is used to assess quality of fused images?
(A) PSNR
(B) ORB
(C) SIFT
(D) HOG
39. Which fusion method is best suited for compressing complementary features from different sensors?
(A) Pixel-based fusion
(B) Feature-based fusion
(C) Histogram-based fusion
(D) Resolution fusion
40. Which of the following improves classification by combining multiple predictions?
(A) Entropy mapping
(B) PCA
(C) Voting schemes
(D) Wavelet transforms
41. Which of the following fusion levels can be used without needing spatial alignment?
(A) Pixel level
(B) Feature level
(C) Decision level
(D) Resolution level
42. Which feature-level fusion technique transforms images into uncorrelated components?
(A) Median filter
(B) PCA
(C) Max pooling
(D) Cross-entropy
43. Which level of fusion is typically used when raw image quality is poor but high-level decisions are possible?
(A) Pixel level
(B) Feature level
(C) Decision level
(D) Frequency level
44. What is a common problem in pixel-level fusion?
(A) Lack of training data
(B) Information redundancy
(C) Classification accuracy
(D) Edge detection errors
45. Which fusion technique is likely to be used in autonomous vehicles?
(A) Decision fusion
(B) Intensity subtraction
(C) Adaptive histogram equalization
(D) Bit plane slicing
46. Which method is likely to be used in real-time applications requiring speed over precision?
(A) Pixel fusion
(B) Feature fusion
(C) Decision fusion
(D) Object fusion
47. Which feature extraction method is best suited for texture-based fusion?
(A) Wavelet transform
(B) Sobel filter
(C) Histogram equalization
(D) Median filtering
48. Which sensor fusion level is often used in biometric systems like face and fingerprint fusion?
(A) Pixel level
(B) Feature level
(C) Decision level
(D) Time level
49. Which fusion strategy uses different combinations of fusion levels?
(A) Hybrid fusion
(B) Linear fusion
(C) Parallel fusion
(D) Multi-band fusion
50. Which fusion level requires the least data storage?
(A) Feature level
(B) Decision level
(C) Pixel level
(D) Frame level
